技术领域Technical Field
本发明涉及工业生产监控技术领域，尤其涉及一种基于人工智能的工业数据品质监控与提升系统。The present invention relates to the technical field of industrial production monitoring, and in particular to an industrial data quality monitoring and improvement system based on artificial intelligence.
背景技术Background technique
自动化和信息化已成为现代工业生产的标志。在这个过程中，生产线上的传感器、控制系统和智能设备产生了海量的数据。这些数据包括但不限于机器性能参数、生产过程监控数据、产品质量指标等，它们是实现高效生产管理和质量控制的关键。Automation and informatization have become the hallmarks of modern industrial production. In this process, sensors, control systems and intelligent devices on the production line generate massive amounts of data. These data include, but are not limited to, machine performance parameters, production process monitoring data, product quality indicators, etc. They are the key to achieving efficient production management and quality control.
然而，传统的数据监控和分析方法存在诸多不足。首先，它们通常依赖于人工操作，这不仅耗时耗力，而且由于人为因素，容易产生误差。其次，传统方法在处理大规模数据时效率低下，难以实现实时监控和快速响应。此外。传统的数据质量控制方法主要依赖专家经验规则进行数据校验，但该方法存在以下几个明显不足：规则设置所需的工作量较大，且难以覆盖所有可能出现的异常情况；无法识别新型异常，只能检测已知的异常模式；各规则独立设置，无法考虑数据之间的相关性；数据质量分析和提升缺乏系统性和智能化支持。However, traditional data monitoring and analysis methods have many shortcomings. First, they usually rely on manual operations, which is not only time-consuming and labor-intensive, but also prone to errors due to human factors. Secondly, traditional methods are inefficient when processing large-scale data, and it is difficult to achieve real-time monitoring and rapid response. In addition. Traditional data quality control methods mainly rely on expert experience rules for data verification, but this method has the following obvious shortcomings: the workload required for rule setting is large, and it is difficult to cover all possible abnormal situations; it is impossible to identify new anomalies, and can only detect known abnormal patterns; each rule is set independently, and the correlation between data cannot be considered; data quality analysis and improvement lack systematic and intelligent support.
发明内容Summary of the invention
基于此，本发明有必要提供一种基于人工智能的工业数据品质监控与提升系统，以解决至少一个上述技术问题。Based on this, it is necessary for the present invention to provide an industrial data quality monitoring and improvement system based on artificial intelligence to solve at least one of the above technical problems.
为实现上述目的，一种基于人工智能的工业数据品质监控与提升系统，所述系统包括以下模块：To achieve the above objectives, an industrial data quality monitoring and improvement system based on artificial intelligence is provided, the system comprising the following modules:
工业数据采集模块S1，用于获取车间设备空间布局数据；根据车间设备空间布局数据对目标生产车间进行高密度感知工业数据采集，得到异构多源工业数据集，其中异构多源工业数据集包括产品生产数据、设备性能参数与产品质量指标；The industrial data collection module S1 is used to obtain the spatial layout data of workshop equipment; based on the spatial layout data of workshop equipment, high-density sensing industrial data collection is performed on the target production workshop to obtain a heterogeneous multi-source industrial data set, where the heterogeneous multi-source industrial data set includes product production data, equipment performance parameters and product quality indicators;
数据品质影响因子挖掘模块S2，用于获取历史产品生产数据与历史设备性能参数；对历史产品生产数据进行低品质数据检测，得到历史低品质生产数据；基于历史产品生产数据与历史设备性能参数进行数据品质影响因子挖掘，得到产品生产数据品质影响因子集；根据产品生产数据品质影响因子集与历史低品质生产数据进行数据品质临界值范围计算，得到生产数据可接受品质区间；The data quality influencing factor mining module S2 is used to obtain historical product production data and historical equipment performance parameters; perform low-quality data detection on historical product production data to obtain historical low-quality production data; perform data quality influencing factor mining based on historical product production data and historical equipment performance parameters to obtain a product production data quality influencing factor set; calculate the data quality critical value range based on the product production data quality influencing factor set and historical low-quality production data to obtain an acceptable quality range for production data;
数据品质评估模块S3，用于根据产品生产数据品质影响因子集对产品生产数据进行数据品质量化评估，得到产品生产数据品质指数；将产品生产数据品质指数与生产数据可接受品质区间进行比较；The data quality assessment module S3 is used to perform a quantitative assessment on the data quality of the product production data according to the product production data quality influencing factor set to obtain the product production data quality index; and compare the product production data quality index with the acceptable quality range of the production data;
数据品质提升决策模块S4，用于当产品生产数据品质指数小于生产数据可接受品质区间下限时，对目标生产车间进行智能生产停线作业，并对产品生产数据进行数据异常根源追溯，得到生产数据异常诱因数据；基于历史产品生产数据、历史低品质生产数据与生产数据异常诱因数据对目标生产车间进行数据质量提升优化决策，得到生产数据质量增强策略；根据生产数据质量增强策略对目标生产车间进行策略深度融合实施；The data quality improvement decision module S4 is used to perform intelligent production line stop operation on the target production workshop when the product production data quality index is less than the lower limit of the acceptable quality range of the production data, and to trace the root cause of data anomaly on the product production data to obtain the production data anomaly inducement data; based on the historical product production data, the historical low-quality production data and the production data anomaly inducement data, the data quality improvement optimization decision is made on the target production workshop to obtain the production data quality enhancement strategy; according to the production data quality enhancement strategy, the strategy is deeply integrated and implemented on the target production workshop;
生产过程知识图谱构建模块S5，对产品生产数据品质指数、生产数据异常诱因数据与生产数据质量增强策略进行因果关系动态知识图谱构建，得到动态数据品质因果知识图谱；The production process knowledge graph construction module S5 constructs a causal relationship dynamic knowledge graph for product production data quality index, production data abnormality inducement data and production data quality enhancement strategy to obtain a dynamic data quality causal knowledge graph;
生产数据态势可视化模块S6，用于当产品生产数据品质指数等于或大于生产数据可接受品质区间数据时，对生产数据设备性能参数与产品质量指标进行时空标注，得到实时生产环境状态画像数据；对动态数据品质因果知识图谱与实时生产环境状态画像数据进行多维数据融合，得到生产质量可视化态势图；将生产质量可视化态势图上传至预设的工业数据监控平台。The production data situation visualization module S6 is used to perform spatiotemporal annotation on the production data equipment performance parameters and product quality indicators when the product production data quality index is equal to or greater than the acceptable quality interval data of the production data, so as to obtain real-time production environment status portrait data; perform multi-dimensional data fusion on the dynamic data quality causal knowledge graph and the real-time production environment status portrait data to obtain a production quality visualization situation map; and upload the production quality visualization situation map to the preset industrial data monitoring platform.
本发明中，工业数据采集模块通过高密度感知工业数据采集，系统能够及时获取车间设备的实时数据，包括产品生产数据、设备性能参数和产品质量指标。这可以提供准确的、实时的生产数据，消除了传统依赖人工操作的缺点，提高了数据获取的效率和准确性。这样的实时监测和采集能够帮助解决传统方法在处理大规模数据时效率低下的问题，并实现对生产线的实时监控和快速响应。该模块通过获取异构多源工业数据集，包括来自不同设备和系统的数据。这样的数据集成能够提供全面的数据视角，使得后续的数据分析和决策可以综合考虑不同数据源的信息，更全面地了解整个生产线上的相关数据。数据品质影响因子挖掘模块通过对历史数据进行分析和建模，系统能够挖掘出数据品质的影响因子集。这些影响因子可以用于后续的数据品质量化评估和提升决策，为评估数据品质提供了依据和参考。基于历史数据，系统可以计算生产数据的可接受品质区间。这个区间可以用作衡量当前数据品质的标准，帮助判断数据是否符合预设的品质要求。该模块能够更全面、系统地考虑数据之间的相关性，并能够识别新型异常，不仅减少了人工规则设置的工作量，还提高了数据质量分析的准确性和智能化水平。数据品质评估模块通过对产品生产数据进行数据品质量化评估，并与生产数据可接受品质区间进行比较，实现了对产品生产数据的自动化评估。通过自动化评估，系统能够快速而准确地对产品生产数据的品质进行量化评估。相比于人工评估，自动化评估可以提高评估的准确性和效率，节省人力资源。通过与可接受品质区间进行比较，系统能够判断产品生产数据是否符合预设的品质要求。这为后续的数据品质提升决策提供了依据，帮助用户及时采取措施来改善数据品质。数据品质提升决策模块通过对产品生产数据品质指数与生产数据可接受品质区间的比较，判断是否需要进行数据质量提升优化决策。当品质指数低于可接受品质区间下限时，系统会进行智能生产停线作业，并追溯数据异常根源，得到异常诱因数据。这样可以帮助用户及时发现和处理生产线上的数据质量异常，减少不合格产品的生产和流通。基于历史数据和异常诱因数据，系统能够生成针对性的生产数据质量增强策略。这些策略可以包括设备维护、工艺优化等方面的改进措施，帮助提升数据品质和生产线的稳定性。该模块能够更加系统地基于历史数据和异常诱因数据生成策略，提高了数据质量的改进效果和生产线的稳定性。生产过程知识图谱构建模块构建因果关系动态知识图谱，将产品生产数据品质指数、异常诱因数据和质量增强策略进行关联，形成动态数据品质因果知识图谱。将数据品质相关的信息进行结构化和可视化表示，帮助用户更好地理解数据品质的影响因素和改进策略。知识图谱的形式可以呈现数据之间的联系和关联，使得用户可以直观地掌握数据品质的整体情况。构建动态数据品质因果知识图谱可以促进知识的共享和传播。知识图谱的形式可以直观地呈现数据之间的联系和关联，帮助用户更好地理解数据品质的影响因素和改进策略，并促进知识的共享和传播。不同利益相关者可以通过知识图谱了解数据品质的相关信息，共同探讨和改进数据品质的方案。生产数据态势可视化模块通过对产品生产数据进行时空标注，并与动态数据品质因果知识图谱进行多维数据融合，生成生产质量可视化态势图。通过将多源数据进行综合分析和可视化展示，帮助用户直观地了解生产数据的状态和趋势。通过可视化态势图，用户可以一目了然地把握生产线上的数据品质情况，快速发现异常情况和改进方向。综上所述，本发明能够实现高效的自动化数据监控和分析，减少人工操作的繁琐和误差，提高数据处理的效率。智能数据质量控制方法能够减少专家规则设置的工作量，识别新型异常，并考虑数据之间的相关性，提高数据质量分析的准确性和全面性。系统性和智能化的数据质量提升方法提供全面的支持和指导，帮助改进数据质量，增强生产线的稳定性。此外，结构化和可视化的数据品质信息表示形式有助于用户更好地理解数据品质相关的信息，促进知识的共享和传播，提高数据品质管理的效率和准确性。In the present invention, the industrial data acquisition module collects industrial data through high-density sensing, and the system can timely obtain real-time data of workshop equipment, including product production data, equipment performance parameters and product quality indicators. This can provide accurate and real-time production data, eliminate the shortcomings of traditional reliance on manual operation, and improve the efficiency and accuracy of data acquisition. Such real-time monitoring and acquisition can help solve the problem of low efficiency of traditional methods in processing large-scale data, and realize real-time monitoring and rapid response to production lines. The module obtains heterogeneous multi-source industrial data sets, including data from different devices and systems. Such data integration can provide a comprehensive data perspective, so that subsequent data analysis and decision-making can comprehensively consider information from different data sources, and more comprehensively understand the relevant data on the entire production line. The data quality influencing factor mining module analyzes and models historical data, and the system can mine a set of influencing factors of data quality. These influencing factors can be used for subsequent quantitative evaluation of data quality and improvement of decision-making, providing a basis and reference for evaluating data quality. Based on historical data, the system can calculate the acceptable quality interval of production data. This interval can be used as a standard for measuring the quality of current data to help determine whether the data meets the preset quality requirements. This module can consider the correlation between data more comprehensively and systematically, and can identify new types of anomalies, which not only reduces the workload of manual rule setting, but also improves the accuracy and intelligence level of data quality analysis. The data quality assessment module realizes the automatic assessment of product production data by quantitatively evaluating the data quality of product production data and comparing it with the acceptable quality range of production data. Through automated assessment, the system can quickly and accurately quantify the quality of product production data. Compared with manual assessment, automated assessment can improve the accuracy and efficiency of assessment and save human resources. By comparing with the acceptable quality range, the system can determine whether the product production data meets the preset quality requirements. This provides a basis for subsequent data quality improvement decisions and helps users take timely measures to improve data quality. The data quality improvement decision module determines whether data quality improvement optimization decisions are needed by comparing the product production data quality index with the acceptable quality range of production data. When the quality index is lower than the lower limit of the acceptable quality range, the system will perform intelligent production line stop operation, trace the root cause of data anomalies, and obtain abnormal inducement data. This can help users to timely discover and deal with data quality anomalies on the production line and reduce the production and circulation of unqualified products. Based on historical data and abnormal inducement data, the system can generate targeted production data quality enhancement strategies. These strategies can include improvement measures in equipment maintenance, process optimization, etc., to help improve data quality and the stability of the production line. This module can generate strategies based on historical data and abnormal inducement data in a more systematic way, improving the improvement effect of data quality and the stability of the production line. The production process knowledge graph construction module constructs a causal dynamic knowledge graph, associates the product production data quality index, abnormal inducement data and quality enhancement strategy, and forms a dynamic data quality causal knowledge graph. The information related to data quality is structured and visualized to help users better understand the factors affecting data quality and improvement strategies. The form of the knowledge graph can present the connections and associations between data, so that users can intuitively grasp the overall situation of data quality. Building a dynamic data quality causal knowledge graph can promote the sharing and dissemination of knowledge. The form of the knowledge graph can intuitively present the connections and associations between data, help users better understand the factors affecting data quality and improvement strategies, and promote the sharing and dissemination of knowledge. Different stakeholders can understand the relevant information of data quality through the knowledge graph, and jointly discuss and improve the data quality plan. The production data situation visualization module generates a production quality visualization situation map by performing spatiotemporal annotation of product production data and multi-dimensional data fusion with the dynamic data quality causal knowledge map. By comprehensively analyzing and visualizing multi-source data, users can intuitively understand the status and trend of production data. Through the visualization situation map, users can grasp the data quality situation on the production line at a glance, quickly discover abnormal situations and improvement directions. In summary, the present invention can achieve efficient automated data monitoring and analysis, reduce the tediousness and errors of manual operations, and improve the efficiency of data processing. The intelligent data quality control method can reduce the workload of expert rule setting, identify new anomalies, and consider the correlation between data to improve the accuracy and comprehensiveness of data quality analysis. The systematic and intelligent data quality improvement method provides comprehensive support and guidance to help improve data quality and enhance the stability of the production line. In addition, the structured and visualized data quality information representation helps users better understand information related to data quality, promote knowledge sharing and dissemination, and improve the efficiency and accuracy of data quality management.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
通过阅读参照以下附图所作的对非限制性实施所作的详细描述，本发明的其它特征、目的和优点将会变得更明显：Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments thereof made with reference to the following drawings:
图1示出了一实施例的基于人工智能的工业数据品质监控与提升系统的模块流程示意图。FIG1 shows a module flow diagram of an industrial data quality monitoring and improvement system based on artificial intelligence according to an embodiment of the present invention.
图2示出了一实施例的S44的详细步骤流程示意图。FIG. 2 is a schematic diagram showing a detailed step flow chart of S44 in an embodiment.
具体实施方式Detailed ways
下面结合附图对本发明专利的技术方法进行清楚、完整的描述，显然，所描述的实施例是本发明的一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域所属的技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.
此外，附图仅为本发明的示意性图解，并非一定是按比例绘制。图中相同的附图标记表示相同或类似的部分，因而将省略对它们的重复描述。附图中所示的一些方框图是功能实体，不一定必须与物理或逻辑上独立的实体相对应。可以采用软件形式来实现功能实体，或在一个或多个硬件模块或集成电路中实现这些功能实体，或在不同网络和/或处理器方法和/或微控制器方法中实现这些功能实体。In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and/or processor methods and/or microcontroller methods.
应当理解的是，虽然在这里可能使用了术语“第一”、“第二”等等来描述各个单元，但是这些单元不应当受这些术语限制。使用这些术语仅仅是为了将一个单元与另一个单元进行区分。举例来说，在不背离示例性实施例的范围的情况下，第一单元可以被称为第二单元，并且类似地第二单元可以被称为第一单元。这里所使用的术语“和/或”包括其中一个或更多所列出的相关联项目的任意和所有组合。It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and/or" used herein includes any and all combinations of one or more of the listed associated items.
为实现上述目的，请参阅图1至图2，本发明提供了一种基于人工智能的工业数据品质监控与提升系统，所述系统包括以下模块：To achieve the above purpose, please refer to Figures 1 and 2. The present invention provides an industrial data quality monitoring and improvement system based on artificial intelligence, and the system includes the following modules:
S1：工业数据采集模块S1，用于获取车间设备空间布局数据；根据车间设备空间布局数据对目标生产车间进行高密度感知工业数据采集，得到异构多源工业数据集，其中异构多源工业数据集包括产品生产数据、设备性能参数与产品质量指标；S1: Industrial data collection module S1 is used to obtain the spatial layout data of workshop equipment; according to the spatial layout data of workshop equipment, high-density sensing industrial data collection is performed on the target production workshop to obtain a heterogeneous multi-source industrial data set, where the heterogeneous multi-source industrial data set includes product production data, equipment performance parameters and product quality indicators;
具体地，例如，可以在目标生产车间中，安装传感器网络或使用无线通信设备，并部署多种传感器设备，如温度传感器、湿度传感器、压力传感器和振动传感器等，以获取车间内各个设备的实时状态信息。这些传感器设备被布置在关键位置，以覆盖整个车间的设备空间布局。接下来，配置传感器设备，使其能够采集产品生产数据、设备性能参数和产品质量指标等各种工业数据。通过设置适当的采样频率和采样精度，确保高密度的数据感知和准确性。最后，采集到的数据从不同传感器和设备整合和处理，形成异构多源工业数据集。这个数据集包括了产品生产数据，如生产数量、生产速率、生产时间等；设备性能参数，如温度、压力、振动频率等；以及产品质量指标，如尺寸、重量、质量等。Specifically, for example, a sensor network can be installed or wireless communication equipment can be used in the target production workshop, and a variety of sensor devices, such as temperature sensors, humidity sensors, pressure sensors, and vibration sensors, can be deployed to obtain real-time status information of each device in the workshop. These sensor devices are arranged in key locations to cover the equipment space layout of the entire workshop. Next, the sensor devices are configured to enable them to collect various industrial data such as product production data, equipment performance parameters, and product quality indicators. By setting appropriate sampling frequency and sampling accuracy, high-density data perception and accuracy are ensured. Finally, the collected data is integrated and processed from different sensors and devices to form a heterogeneous multi-source industrial data set. This data set includes product production data, such as production quantity, production rate, production time, etc.; equipment performance parameters, such as temperature, pressure, vibration frequency, etc.; and product quality indicators, such as size, weight, quality, etc.
S2：数据品质影响因子挖掘模块S2，用于获取历史产品生产数据与历史设备性能参数；对历史产品生产数据进行低品质数据检测，得到历史低品质生产数据；基于历史产品生产数据与历史设备性能参数进行数据品质影响因子挖掘，得到产品生产数据品质影响因子集；根据产品生产数据品质影响因子集与历史低品质生产数据进行数据品质临界值范围计算，得到生产数据可接受品质区间；S2: Data quality influencing factor mining module S2 is used to obtain historical product production data and historical equipment performance parameters; perform low-quality data detection on historical product production data to obtain historical low-quality production data; perform data quality influencing factor mining based on historical product production data and historical equipment performance parameters to obtain a set of product production data quality influencing factors; calculate the data quality critical value range based on the product production data quality influencing factor set and historical low-quality production data to obtain an acceptable quality range for production data;
具体地，例如，可以通过系统集成或数据接口，获取历史产品生产数据和历史设备性能参数。对于历史产品生产数据，可以从生产数据库、生产记录或生产管理系统中提取相关数据，如生产数量、生产速率等。对于历史设备性能参数，可以通过设备监控系统、传感器数据或设备日志记录等途径获得，如设备温度、压力等。这些数据可以通过合适的数据提取方法和通信协议获取。接着，对历史产品生产数据进行低品质数据检测，以识别出历史低品质的生产数据样本。这可以通过应用统计分析方法、异常检测算法或规则引擎等技术来实现。探测方法可以根据具体的数据特征和质量标准进行选择和定制。然后，基于历史产品生产数据和历史设备性能参数，进行数据品质影响因子的挖掘。这可以运用数据挖掘技术，如机器学习算法、统计分析方法或相关性分析等，来发现对产品生产数据品质具有重要影响的因素。这些因素可以是设备状态、工艺参数、环境条件等。最后，根据产品生产数据品质影响因子集和历史低品质生产数据，进行数据品质临界值范围的计算。通过分析历史低品质生产数据与品质影响因子集之间的关联关系，可以确定生产数据可接受的品质区间。Specifically, for example, historical product production data and historical equipment performance parameters can be obtained through system integration or data interface. For historical product production data, relevant data such as production quantity, production rate, etc. can be extracted from production database, production records or production management system. For historical equipment performance parameters, such as equipment temperature, pressure, etc., can be obtained through equipment monitoring system, sensor data or equipment log records. These data can be obtained through appropriate data extraction methods and communication protocols. Next, low-quality data detection is performed on historical product production data to identify historical low-quality production data samples. This can be achieved by applying statistical analysis methods, anomaly detection algorithms or rule engines and other technologies. The detection method can be selected and customized according to specific data characteristics and quality standards. Then, based on historical product production data and historical equipment performance parameters, data quality influencing factors are mined. This can use data mining techniques, such as machine learning algorithms, statistical analysis methods or correlation analysis, to discover factors that have an important impact on the quality of product production data. These factors can be equipment status, process parameters, environmental conditions, etc. Finally, the data quality critical value range is calculated based on the product production data quality influencing factor set and historical low-quality production data. By analyzing the correlation between historical low-quality production data and the set of quality influencing factors, the acceptable quality range of production data can be determined.
S3：数据品质评估模块S3，用于根据产品生产数据品质影响因子集对产品生产数据进行数据品质量化评估，得到产品生产数据品质指数；将产品生产数据品质指数与生产数据可接受品质区间进行比较；S3: Data quality assessment module S3 is used to perform quantitative data quality assessment on product production data according to the product production data quality influencing factor set to obtain the product production data quality index; compare the product production data quality index with the acceptable quality range of production data;
具体地，例如，可以利用数据品质影响因子集，对产品生产数据进行量化评估。通过对每个数据样本的影响因子进行加权计算或建立适当的模型，可以综合考虑各个因子对数据品质的影响程度，从而得到产品生产数据的品质指数。该指数可以反映数据的准确性、完整性、一致性等方面的品质情况。接着，将产品生产数据的品质指数与生产数据可接受品质区间进行比较。Specifically, for example, a set of data quality influencing factors can be used to quantitatively evaluate product production data. By weighting the influencing factors of each data sample or establishing an appropriate model, the degree of influence of each factor on data quality can be comprehensively considered to obtain the quality index of product production data. This index can reflect the quality of data in terms of accuracy, completeness, consistency, etc. Next, the quality index of product production data is compared with the acceptable quality range of production data.
S4：数据品质提升决策模块S4，用于当产品生产数据品质指数小于生产数据可接受品质区间下限时，对目标生产车间进行智能生产停线作业，并对产品生产数据进行数据异常根源追溯，得到生产数据异常诱因数据；基于历史产品生产数据、历史低品质生产数据与生产数据异常诱因数据对目标生产车间进行数据质量提升优化决策，得到生产数据质量增强策略；根据生产数据质量增强策略对目标生产车间进行策略深度融合实施；S4: Data quality improvement decision module S4 is used to perform intelligent production line stop operation on the target production workshop when the product production data quality index is less than the lower limit of the acceptable quality range of the production data, and trace the root cause of data anomaly of the product production data to obtain the abnormal production data inducement data; based on the historical product production data, historical low-quality production data and production data abnormal inducement data, the data quality improvement optimization decision is made for the target production workshop to obtain the production data quality enhancement strategy; according to the production data quality enhancement strategy, the strategy is deeply integrated and implemented for the target production workshop;
具体地，例如，当产品生产数据的品质指数低于可接受品质区间下限时，该模块会自动触发智能生产停线作业，以避免低品质数据对后续生产和质量造成不良影响。接着，该模块对产品生产数据进行异常根源追溯，通过分析低品质数据的特征、相关因素和历史数据，可以确定导致数据异常的诱因数据。这些诱因数据可能包括设备异常、工艺变化、材料问题等。基于历史产品生产数据、历史低品质生产数据和生产数据异常诱因数据，该模块会对目标生产车间进行数据质量提升优化决策。这些决策可以基于数据挖掘、统计分析、机器学习等技术，结合专家经验和领域知识，识别出对数据品质影响较大的因素，并制定相应的策略和措施来提升生产数据的品质。最后，根据生产数据质量增强策略，该模块对目标生产车间进行策略深度融合实施。这包括实施数据质量监控与反馈机制，对生产过程中的数据进行实时监测和分析，及时发现和处理数据品质异常。同时，通过优化生产流程、改进设备维护和调整工艺参数等方式，全面提升生产数据的品质水平。Specifically, for example, when the quality index of product production data is lower than the lower limit of the acceptable quality range, the module will automatically trigger the intelligent production line stop operation to avoid the adverse effects of low-quality data on subsequent production and quality. Then, the module traces the root cause of abnormal product production data. By analyzing the characteristics, related factors and historical data of low-quality data, the inducement data that causes data abnormality can be determined. These inducement data may include equipment abnormalities, process changes, material problems, etc. Based on historical product production data, historical low-quality production data and production data abnormal inducement data, the module will make data quality improvement optimization decisions for the target production workshop. These decisions can be based on data mining, statistical analysis, machine learning and other technologies, combined with expert experience and domain knowledge, to identify factors that have a greater impact on data quality, and formulate corresponding strategies and measures to improve the quality of production data. Finally, according to the production data quality enhancement strategy, the module implements deep integration of strategies for the target production workshop. This includes implementing a data quality monitoring and feedback mechanism, real-time monitoring and analysis of data in the production process, and timely detection and processing of data quality abnormalities. At the same time, the quality level of production data can be comprehensively improved by optimizing production processes, improving equipment maintenance and adjusting process parameters.
S5：生产过程知识图谱构建模块S5，对产品生产数据品质指数、生产数据异常诱因数据与生产数据质量增强策略进行因果关系动态知识图谱构建，得到动态数据品质因果知识图谱；S5: Production process knowledge graph construction module S5, which constructs a causal relationship dynamic knowledge graph for product production data quality index, production data abnormality inducement data and production data quality enhancement strategy, and obtains a dynamic data quality causal knowledge graph;
具体地，例如，可以利用因果关系分析方法，如因果推断、因果图分析等，对产品生产数据品质指数、生产数据异常诱因数据和生产数据质量增强策略进行深入分析。这可以帮助确定不同因素之间的因果关系，并揭示它们对数据品质的影响程度。基于因果关系分析的结果，使用知识图谱构建工具或者方法，如图数据库、知识图谱建模工具等，构建动态数据品质因果知识图谱。这个知识图谱可以将不同因素、指标和策略之间的关联关系以图形化的方式进行表示，并支持动态更新和查询。Specifically, for example, causal analysis methods such as causal inference and causal graph analysis can be used to conduct in-depth analysis of product production data quality index, production data abnormality inducement data, and production data quality enhancement strategies. This can help determine the causal relationship between different factors and reveal the extent of their impact on data quality. Based on the results of causal analysis, a knowledge graph construction tool or method, such as a graph database, knowledge graph modeling tool, etc., is used to construct a dynamic data quality causal knowledge graph. This knowledge graph can graphically represent the relationship between different factors, indicators, and strategies, and supports dynamic updates and queries.
S6：生产数据态势可视化模块S6，用于当产品生产数据品质指数等于或大于生产数据可接受品质区间数据时，对生产数据设备性能参数与产品质量指标进行时空标注，得到实时生产环境状态画像数据；对动态数据品质因果知识图谱与实时生产环境状态画像数据进行多维数据融合，得到生产质量可视化态势图；将生产质量可视化态势图上传至预设的工业数据监控平台。S6: Production data situation visualization module S6 is used to perform spatiotemporal annotation on production data equipment performance parameters and product quality indicators when the product production data quality index is equal to or greater than the acceptable quality interval data of the production data, and obtain real-time production environment status portrait data; perform multi-dimensional data fusion on the dynamic data quality causal knowledge graph and the real-time production environment status portrait data to obtain a production quality visualization situation map; upload the production quality visualization situation map to the preset industrial data monitoring platform.
具体地，例如，当产品生产数据品质指数等于或大于生产数据可接受品质区间数据时，可以针对满足品质要求的数据，对生产数据设备性能参数与产品质量指标进行时空标注。这意味着对数据进行标记，以指示数据所对应的设备性能参数和产品质量指标。标注可以通过自动识别或者模型推断等方式进行。基于生产数据标注的结果，得到实时生产环境状态画像数据。这些数据反映了当前生产环境中设备性能参数和产品质量指标的状态，可以包括时间、空间以及其他相关的属性信息。将动态数据品质因果知识图谱和实时生产环境状态画像数据进行多维数据融合。这个步骤涉及将知识图谱中的因果关系和实时环境数据进行关联和整合，以建立综合的数据模型。多维数据融合可以利用数据集成工具、数据挖掘算法或机器学习方法等进行。通过对多维数据融合的结果进行可视化处理，生成生产质量可视化态势图。这个图形化展示将动态数据品质因果知识图谱和实时生产环境状态画像数据进行交互展示，以呈现生产质量的态势变化和趋势。可视化态势图可以利用数据可视化工具、图形库或者仪表盘软件等实现。最后，将生成的生产质量可视化态势图上传至预设的工业数据监控平台。这个平台可以是企业内部的数据监控系统或者云端的数据分析平台。上传后，生产质量可视化态势图可以被进一步分析、共享和应用于实时监控和决策支持。Specifically, for example, when the product production data quality index is equal to or greater than the acceptable quality interval data of the production data, the production data equipment performance parameters and product quality indicators can be spatiotemporally annotated for the data that meets the quality requirements. This means that the data is marked to indicate the equipment performance parameters and product quality indicators corresponding to the data. The annotation can be performed by automatic recognition or model inference. Based on the results of the production data annotation, real-time production environment status portrait data is obtained. These data reflect the status of equipment performance parameters and product quality indicators in the current production environment, and may include time, space and other related attribute information. The dynamic data quality causal knowledge graph and the real-time production environment status portrait data are multi-dimensionally fused. This step involves associating and integrating the causal relationship in the knowledge graph with the real-time environment data to establish a comprehensive data model. Multi-dimensional data fusion can be performed using data integration tools, data mining algorithms or machine learning methods. By visualizing the results of the multi-dimensional data fusion, a production quality visualization situation map is generated. This graphical display interactively displays the dynamic data quality causal knowledge graph and the real-time production environment status portrait data to present the situation changes and trends of production quality. The visualization situation map can be implemented using data visualization tools, graphics libraries or dashboard software. Finally, the generated production quality visualization situation map is uploaded to the preset industrial data monitoring platform. This platform can be the enterprise's internal data monitoring system or the cloud data analysis platform. After uploading, the production quality visualization situation map can be further analyzed, shared and applied to real-time monitoring and decision support.
本发明中，工业数据采集模块通过高密度感知工业数据采集，系统能够及时获取车间设备的实时数据，包括产品生产数据、设备性能参数和产品质量指标。这可以提供准确的、实时的生产数据，消除了传统依赖人工操作的缺点，提高了数据获取的效率和准确性。这样的实时监测和采集能够帮助解决传统方法在处理大规模数据时效率低下的问题，并实现对生产线的实时监控和快速响应。该模块通过获取异构多源工业数据集，包括来自不同设备和系统的数据。这样的数据集成能够提供全面的数据视角，使得后续的数据分析和决策可以综合考虑不同数据源的信息，更全面地了解整个生产线上的相关数据。数据品质影响因子挖掘模块通过对历史数据进行分析和建模，系统能够挖掘出数据品质的影响因子集。这些影响因子可以用于后续的数据品质量化评估和提升决策，为评估数据品质提供了依据和参考。基于历史数据，系统可以计算生产数据的可接受品质区间。这个区间可以用作衡量当前数据品质的标准，帮助判断数据是否符合预设的品质要求。该模块能够更全面、系统地考虑数据之间的相关性，并能够识别新型异常，不仅减少了人工规则设置的工作量，还提高了数据质量分析的准确性和智能化水平。数据品质评估模块通过对产品生产数据进行数据品质量化评估，并与生产数据可接受品质区间进行比较，实现了对产品生产数据的自动化评估。通过自动化评估，系统能够快速而准确地对产品生产数据的品质进行量化评估。相比于人工评估，自动化评估可以提高评估的准确性和效率，节省人力资源。通过与可接受品质区间进行比较，系统能够判断产品生产数据是否符合预设的品质要求。这为后续的数据品质提升决策提供了依据，帮助用户及时采取措施来改善数据品质。数据品质提升决策模块通过对产品生产数据品质指数与生产数据可接受品质区间的比较，判断是否需要进行数据质量提升优化决策。当品质指数低于可接受品质区间下限时，系统会进行智能生产停线作业，并追溯数据异常根源，得到异常诱因数据。这样可以帮助用户及时发现和处理生产线上的数据质量异常，减少不合格产品的生产和流通。基于历史数据和异常诱因数据，系统能够生成针对性的生产数据质量增强策略。这些策略可以包括设备维护、工艺优化等方面的改进措施，帮助提升数据品质和生产线的稳定性。该模块能够更加系统地基于历史数据和异常诱因数据生成策略，提高了数据质量的改进效果和生产线的稳定性。生产过程知识图谱构建模块构建因果关系动态知识图谱，将产品生产数据品质指数、异常诱因数据和质量增强策略进行关联，形成动态数据品质因果知识图谱。将数据品质相关的信息进行结构化和可视化表示，帮助用户更好地理解数据品质的影响因素和改进策略。知识图谱的形式可以呈现数据之间的联系和关联，使得用户可以直观地掌握数据品质的整体情况。构建动态数据品质因果知识图谱可以促进知识的共享和传播。知识图谱的形式可以直观地呈现数据之间的联系和关联，帮助用户更好地理解数据品质的影响因素和改进策略，并促进知识的共享和传播。不同利益相关者可以通过知识图谱了解数据品质的相关信息，共同探讨和改进数据品质的方案。生产数据态势可视化模块通过对产品生产数据进行时空标注，并与动态数据品质因果知识图谱进行多维数据融合，生成生产质量可视化态势图。通过将多源数据进行综合分析和可视化展示，帮助用户直观地了解生产数据的状态和趋势。通过可视化态势图，用户可以一目了然地把握生产线上的数据品质情况，快速发现异常情况和改进方向。综上所述，本发明能够实现高效的自动化数据监控和分析，减少人工操作的繁琐和误差，提高数据处理的效率。智能数据质量控制方法能够减少专家规则设置的工作量，识别新型异常，并考虑数据之间的相关性，提高数据质量分析的准确性和全面性。系统性和智能化的数据质量提升方法提供全面的支持和指导，帮助改进数据质量，增强生产线的稳定性。此外，结构化和可视化的数据品质信息表示形式有助于用户更好地理解数据品质相关的信息，促进知识的共享和传播，提高数据品质管理的效率和准确性。In the present invention, the industrial data acquisition module collects industrial data through high-density sensing, and the system can timely obtain real-time data of workshop equipment, including product production data, equipment performance parameters and product quality indicators. This can provide accurate and real-time production data, eliminate the shortcomings of traditional reliance on manual operation, and improve the efficiency and accuracy of data acquisition. Such real-time monitoring and acquisition can help solve the problem of low efficiency of traditional methods in processing large-scale data, and realize real-time monitoring and rapid response to production lines. The module obtains heterogeneous multi-source industrial data sets, including data from different devices and systems. Such data integration can provide a comprehensive data perspective, so that subsequent data analysis and decision-making can comprehensively consider information from different data sources and have a more comprehensive understanding of the relevant data on the entire production line. The data quality influencing factor mining module analyzes and models historical data, and the system can mine a set of influencing factors for data quality. These influencing factors can be used for subsequent quantitative evaluation of data quality and improvement of decision-making, providing a basis and reference for evaluating data quality. Based on historical data, the system can calculate the acceptable quality interval of production data. This interval can be used as a standard for measuring the quality of current data to help determine whether the data meets the preset quality requirements. This module can consider the correlation between data more comprehensively and systematically, and can identify new types of anomalies, which not only reduces the workload of manual rule setting, but also improves the accuracy and intelligence level of data quality analysis. The data quality assessment module realizes the automatic assessment of product production data by quantitatively evaluating the data quality of product production data and comparing it with the acceptable quality range of production data. Through automated assessment, the system can quickly and accurately quantify the quality of product production data. Compared with manual assessment, automated assessment can improve the accuracy and efficiency of assessment and save human resources. By comparing with the acceptable quality range, the system can determine whether the product production data meets the preset quality requirements. This provides a basis for subsequent data quality improvement decisions and helps users take timely measures to improve data quality. The data quality improvement decision module determines whether data quality improvement optimization decisions are needed by comparing the product production data quality index with the acceptable quality range of production data. When the quality index is lower than the lower limit of the acceptable quality range, the system will perform intelligent production line stop operation, trace the root cause of data anomalies, and obtain abnormal inducement data. This can help users to timely discover and deal with data quality anomalies on the production line and reduce the production and circulation of unqualified products. Based on historical data and abnormal inducement data, the system can generate targeted production data quality enhancement strategies. These strategies can include improvement measures in equipment maintenance, process optimization, etc., to help improve data quality and the stability of the production line. This module can generate strategies based on historical data and abnormal inducement data in a more systematic way, improving the improvement effect of data quality and the stability of the production line. The production process knowledge graph construction module constructs a causal relationship dynamic knowledge graph, associates the product production data quality index, abnormal inducement data and quality enhancement strategy, and forms a dynamic data quality causal knowledge graph. The information related to data quality is structured and visualized to help users better understand the factors affecting data quality and improvement strategies. The form of knowledge graph can present the connections and associations between data, so that users can intuitively grasp the overall situation of data quality. Building a dynamic data quality causal knowledge graph can promote the sharing and dissemination of knowledge. The form of knowledge graph can intuitively present the connections and associations between data, help users better understand the factors affecting data quality and improvement strategies, and promote the sharing and dissemination of knowledge. Different stakeholders can understand the relevant information of data quality through the knowledge graph, and jointly discuss and improve the data quality plan. The production data situation visualization module generates a production quality visualization situation map by performing spatiotemporal annotation of product production data and multi-dimensional data fusion with the dynamic data quality causal knowledge map. By comprehensively analyzing and visualizing multi-source data, users can intuitively understand the status and trend of production data. Through the visualization situation map, users can grasp the data quality situation on the production line at a glance, quickly discover abnormal situations and improvement directions. In summary, the present invention can achieve efficient automated data monitoring and analysis, reduce the tediousness and errors of manual operations, and improve the efficiency of data processing. The intelligent data quality control method can reduce the workload of expert rule setting, identify new anomalies, and consider the correlation between data to improve the accuracy and comprehensiveness of data quality analysis. The systematic and intelligent data quality improvement method provides comprehensive support and guidance to help improve data quality and enhance the stability of the production line. In addition, the structured and visualized data quality information representation helps users better understand information related to data quality, promote knowledge sharing and dissemination, and improve the efficiency and accuracy of data quality management.
优选地，工业数据采集模块S1具体为：Preferably, the industrial data acquisition module S1 is specifically:
S11：获取车间设备空间布局数据；S11: Obtaining workshop equipment space layout data;
具体地，例如，可以使用测量工具（如测量仪器、激光测距仪等），对车间内的设备位置、尺寸和相对位置进行测量，并进行标记和记录。根据测量数据，使用计算机辅助设计（CAD）软件或其他绘图工具，绘制车间的平面图。平面图应包括设备的位置、设备之间的距离和相对位置等详细信息。Specifically, for example, you can use measuring tools (such as measuring instruments, laser rangefinders, etc.) to measure the location, size and relative position of equipment in the workshop, and mark and record them. Based on the measurement data, use computer-aided design (CAD) software or other drawing tools to draw a floor plan of the workshop. The floor plan should include detailed information such as the location of the equipment, the distance between the equipment and the relative position.
S12：根据车间设备空间布局数据对目标生产车间进行自动采集与人工录入设备划分，得到自动采集设备标识数据与人工录入设备标识数据；S12: dividing the target production workshop into automatic collection equipment and manual entry equipment according to the workshop equipment space layout data, and obtaining automatic collection equipment identification data and manual entry equipment identification data;
具体地，例如，可以基于车间设备空间布局数据，了解设备的类型、功能和位置信息。根据设备的特征和功能，确定哪些设备可以通过自动采集系统进行数据采集。例如，具备传感器、监测装置等自动化设备可以被归类为自动采集设备。为每个自动采集设备分配一个唯一的标识符，例如设备编号、条形码或RFID标签。这些标识符可以用于识别设备和与自动采集系统进行连接。对于无法通过自动采集系统获取数据的设备，需要进行人工录入。根据设备的性质和数据获取的可行性，确定哪些设备需要通过人工手动输入数据。为每个人工录入设备分配一个唯一的标识符，例如设备编号、条形码或RFID标签。这些标识符可以用于识别设备和与数据录入过程进行关联。记录每个设备的标识数据，包括自动采集设备和人工录入设备的标识符。Specifically, for example, based on the spatial layout data of the workshop equipment, the type, function and location information of the equipment can be understood. According to the characteristics and functions of the equipment, determine which equipment can collect data through the automatic collection system. For example, automated equipment with sensors, monitoring devices, etc. can be classified as automatic collection equipment. Assign a unique identifier to each automatic collection device, such as a device number, barcode, or RFID tag. These identifiers can be used to identify the device and connect to the automatic collection system. For equipment that cannot obtain data through the automatic collection system, manual entry is required. According to the nature of the equipment and the feasibility of data acquisition, determine which equipment requires manual data entry. Assign a unique identifier to each manual entry device, such as a device number, barcode, or RFID tag. These identifiers can be used to identify the device and associate with the data entry process. Record the identification data of each device, including the identifiers of automatic collection devices and manual entry devices.
S13：根据自动采集设备标识数据对目标生产车间中相应设备部署工业物联网传感器阵列，并进行异步多源数据采集，得到产品生产数据与设备性能参数；S13: deploying an industrial Internet of Things sensor array on corresponding equipment in the target production workshop according to the automatically collected equipment identification data, and performing asynchronous multi-source data collection to obtain product production data and equipment performance parameters;
具体地，例如，可以根据自动采集设备标识数据，确定需要部署的传感器类型和数量，以覆盖目标生产车间中的设备。根据设备标识数据和传感器部署规划，工业物联网传感器阵列被部署在目标生产车间中的相应设备上。传感器的选型包括温度传感器、压力传感器、振动传感器等，用于采集设备运行状态和性能参数。同时，在产品生产环节，根据需要监测的产品生产数据，选择适当的传感器类型并在生产线上部署，例如称重传感器用于监测产品重量。配置数据采集模块与传感器连接，实现异步多源数据采集。传感器开始采集设备的数据，包括温度、压力、振动等参数，以及产品的生产数据，如产品重量。这些数据通过数据采集节点或网关设备进行接收，并存储在本地数据库或云平台中。Specifically, for example, the type and number of sensors to be deployed can be determined based on the automatic collection of device identification data to cover the equipment in the target production workshop. Based on the device identification data and the sensor deployment plan, the industrial Internet of Things sensor array is deployed on the corresponding equipment in the target production workshop. The selection of sensors includes temperature sensors, pressure sensors, vibration sensors, etc., which are used to collect equipment operation status and performance parameters. At the same time, in the product production link, according to the product production data that needs to be monitored, the appropriate sensor type is selected and deployed on the production line, such as weighing sensors for monitoring product weight. The data acquisition module is configured to connect with the sensor to realize asynchronous multi-source data acquisition. The sensor starts to collect equipment data, including parameters such as temperature, pressure, vibration, and product production data, such as product weight. These data are received through data acquisition nodes or gateway devices and stored in local databases or cloud platforms.
S14：根据人工录入设备标识数据对目标生产车间中相应设备进行人机交互数据采集，得到手工录入生产作业数据；S14: collecting human-machine interaction data of corresponding equipment in the target production workshop according to the manually entered equipment identification data to obtain manually entered production operation data;
具体地，例如，可以根据人工录入设备标识数据，确定需要进行人机交互数据采集的设备。针对每个需要采集数据的设备，设计相应的人机交互界面。界面可以是电脑应用程序、移动设备应用程序或专门的数据采集终端。操作员或相关人员通过人机交互界面，输入和记录与设备相关的生产作业数据。这些数据可以包括设备操作记录、维修记录、故障信息等。Specifically, for example, the equipment that needs to be collected through human-computer interaction data can be determined based on the manually entered equipment identification data. For each equipment that needs to collect data, a corresponding human-computer interaction interface is designed. The interface can be a computer application, a mobile device application, or a dedicated data collection terminal. The operator or relevant personnel inputs and records the production operation data related to the equipment through the human-computer interaction interface. These data may include equipment operation records, maintenance records, fault information, etc.
S15：对手工录入生产作业数据进行自适应人工误差校正，得到优化手工录入生产作业数据；S15: performing adaptive manual error correction on the manually entered production operation data to obtain optimized manually entered production operation data;
具体地，例如，可以使用数据挖掘和统计分析方法来识别手工录入生产作业数据中潜在的数据错误和异常。基于数据质量分析的结果，建立自适应人工误差校正模型。该模型可以根据数据的特征和错误模式，自动校正手工录入数据中的误差。根据建立的误差校正模型，对手工录入生产作业数据进行校正操作。可以采用自动化脚本、数据处理工具或专门的数据校正系统来实现误差校正过程。Specifically, for example, data mining and statistical analysis methods can be used to identify potential data errors and anomalies in manually entered production operation data. Based on the results of data quality analysis, an adaptive manual error correction model is established. The model can automatically correct errors in manually entered data based on the characteristics and error patterns of the data. Correction operations are performed on manually entered production operation data based on the established error correction model. The error correction process can be implemented using automated scripts, data processing tools, or specialized data correction systems.
S16：对产品生产数据与优化手工录入生产作业数据进行多模态产品质量指标评价，得到产品质量指标；S16: Perform multi-modal product quality index evaluation on product production data and optimized manually entered production operation data to obtain product quality index;
具体地，例如，可以将产品生产数据和优化手工录入生产作业数据进行整合，以建立完整的数据集。确保数据的对应关系和一致性。根据具体的应用需求和产品特性，定义多模态产品质量指标。这些指标可以包括生产效率、产品质量、故障率等方面的指标。根据产品质量指标的定义，选择适当的评价方法。例如，可以采用统计分析、数据挖掘、机器学习等方法来进行多模态产品质量指标的评价。基于整合的数据集，进行数据分析和模型构建。可以使用适当的算法和技术，建立预测模型或分类模型，以评估产品质量指标。利用建立的模型，对产品生产数据和优化手工录入生产作业数据进行评价，得到产品质量指标的结果。Specifically, for example, product production data and optimized manually entered production operation data can be integrated to establish a complete data set. Ensure the correspondence and consistency of the data. Define multimodal product quality indicators based on specific application requirements and product characteristics. These indicators can include indicators in production efficiency, product quality, failure rate, etc. According to the definition of product quality indicators, select appropriate evaluation methods. For example, statistical analysis, data mining, machine learning and other methods can be used to evaluate multimodal product quality indicators. Based on the integrated data set, perform data analysis and model building. Appropriate algorithms and techniques can be used to establish a prediction model or classification model to evaluate product quality indicators. Using the established model, evaluate product production data and optimized manually entered production operation data to obtain the results of product quality indicators.
S17：对产品生产数据、设备性能参数与产品质量指标进行智能加权异构数据聚合，得到异构多源工业数据集，其中异构多源工业数据集包括产品生产数据、设备性能参数与产品质量指标。S17: Perform intelligent weighted heterogeneous data aggregation on product production data, equipment performance parameters and product quality indicators to obtain a heterogeneous multi-source industrial data set, where the heterogeneous multi-source industrial data set includes product production data, equipment performance parameters and product quality indicators.
具体地，例如，可以针对产品生产数据、设备性能参数和产品质量指标，对其进行智能加权。加权可以基于数据质量、数据可信度、数据的重要性等因素进行。使用加权算法，如加权平均法或加权回归方法，来计算不同数据源的权重。将加权后的产品生产数据、设备性能参数和产品质量指标进行异构数据聚合。这可以通过数据连接、合并或关联操作来实现。根据数据的关联关系和特征，将数据进行聚合，形成异构多源工业数据集。Specifically, for example, product production data, equipment performance parameters, and product quality indicators can be intelligently weighted. Weighting can be based on factors such as data quality, data credibility, and data importance. Use weighting algorithms, such as weighted average or weighted regression methods, to calculate the weights of different data sources. Aggregate the weighted product production data, equipment performance parameters, and product quality indicators for heterogeneous data. This can be achieved through data connection, merging, or association operations. Aggregate the data based on the association relationship and characteristics of the data to form a heterogeneous multi-source industrial data set.
本发明通过获取车间设备空间布局数据，可以获得车间中各个设备的位置和布局信息。这为后续的自动采集和人工录入设备标识数据提供了基础，帮助确定设备的相对位置和关联性，从而更好地理解数据的来源和关系。通过根据车间设备空间布局数据，将设备划分为自动采集设备和人工录入设备，实现数据来源的区分。自动采集设备标识数据所代表的设备的数据来源可以从工业物联网传感器阵列中异步多源采集，而人工录入设备标识数据所代表的设备的数据来源可以通过人机交互进行数据采集。这样的划分提供了不同方式采集的数据来源，为后续的数据处理和质量评价提供了多样性和灵活性。基于自动采集设备标识数据，针对目标生产车间中的设备，部署工业物联网传感器阵列，实现对设备的数据采集。通过异步多源数据采集，可以获取产品生产数据和设备性能参数。这些数据对于生产过程的监控和分析具有重要价值。通过进行人机交互数据采集，记录生产作业过程中的重要数据。手工录入的生产作业数据是一种补充，可以提供与自动采集数据不同的视角和信息，丰富了数据源的多样性。对手工录入的生产作业数据进行自适应人工误差校正，通过校正过程修正可能存在的人为误差。这样可以提高手工录入数据的准确性和一致性，确保数据的可靠性和可比性。通过对产品生产数据和经过优化的手工录入生产作业数据进行多模态评价，综合考虑多个数据来源的信息，得到产品质量指标。这样的评价方式可以更全面地了解产品质量状况，将不同数据源的信息进行综合分析，提供更准确和全面的产品质量指标。通过对产品生产数据、设备性能参数和产品质量指标进行智能加权的异构数据聚合，将不同数据来源的信息融合为一个综合的工业数据集。这样的数据集包括产品生产数据、设备性能参数和产品质量指标，提供了对整个生产过程的综合视图。The present invention can obtain the location and layout information of each device in the workshop by acquiring the spatial layout data of the workshop equipment. This provides a basis for the subsequent automatic collection and manual entry of equipment identification data, helps to determine the relative position and relevance of the equipment, and thus better understands the source and relationship of the data. By dividing the equipment into automatic collection equipment and manual entry equipment according to the spatial layout data of the workshop equipment, the data source is distinguished. The data source of the equipment represented by the automatic collection equipment identification data can be asynchronously multi-source collected from the industrial Internet of Things sensor array, while the data source of the equipment represented by the manual entry equipment identification data can be collected through human-computer interaction. Such a division provides data sources collected in different ways, which provides diversity and flexibility for subsequent data processing and quality evaluation. Based on the automatic collection of equipment identification data, the industrial Internet of Things sensor array is deployed for the equipment in the target production workshop to realize data collection of the equipment. Through asynchronous multi-source data collection, product production data and equipment performance parameters can be obtained. These data are of great value for monitoring and analysis of the production process. By performing human-computer interaction data collection, important data in the production operation process is recorded. Manually entered production operation data is a supplement that can provide different perspectives and information from automatically collected data, enriching the diversity of data sources. Perform adaptive human error correction on manually entered production operation data, and correct possible human errors through the correction process. This can improve the accuracy and consistency of manually entered data and ensure the reliability and comparability of data. Through multimodal evaluation of product production data and optimized manually entered production operation data, product quality indicators are obtained by comprehensively considering information from multiple data sources. This evaluation method can provide a more comprehensive understanding of product quality status, comprehensively analyze information from different data sources, and provide more accurate and comprehensive product quality indicators. Through intelligently weighted heterogeneous data aggregation of product production data, equipment performance parameters, and product quality indicators, information from different data sources is integrated into a comprehensive industrial data set. Such a data set includes product production data, equipment performance parameters, and product quality indicators, providing a comprehensive view of the entire production process.
优选地，数据品质影响因子挖掘模块S2具体为：Preferably, the data quality influencing factor mining module S2 is specifically:
S21：获取历史产品生产数据与历史设备性能参数；S21: Obtain historical product production data and historical equipment performance parameters;
具体地，例如，可以确定历史产品生产数据和历史设备性能参数的数据源。这可以是生产数据库、设备监控系统、日志文件、传感器数据等。确保能够获取到包含所需数据的合适数据源。根据数据源的不同，可能需要使用相应的数据提取工具、API调用、数据库查询或日志分析技术来检索和提取数据。Specifically, for example, you can determine the data source of historical product production data and historical equipment performance parameters. This can be a production database, equipment monitoring system, log files, sensor data, etc. Ensure that you can obtain the appropriate data source containing the required data. Depending on the data source, you may need to use corresponding data extraction tools, API calls, database queries, or log analysis techniques to retrieve and extract data.
S22：对历史产品生产数据进行低品质数据检测，得到历史低品质生产数据；S22: Perform low-quality data detection on historical product production data to obtain historical low-quality production data;
具体地，例如，可以根据具体的数据质量要求和应用需求，定义低品质数据的评估指标。这些指标可以包括数据完整性、准确性、一致性等方面的要求。根据定义的评估指标，选择适当的低品质数据检测方法。例如，可以使用统计分析、数据挖掘、规则检测等方法来检测低品质数据。根据选择的低品质数据检测方法，对历史产品生产数据进行检测操作。可以使用相应的工具和技术，如Python的pandas库和numpy库，编写检测脚本或应用现有的数据质量工具。从检测结果中提取历史低品质生产数据。根据检测结果的标志或阈值，识别并提取低品质数据，形成一个包含历史低品质生产数据的数据集。Specifically, for example, evaluation indicators for low-quality data can be defined based on specific data quality requirements and application requirements. These indicators may include requirements for data integrity, accuracy, consistency, etc. According to the defined evaluation indicators, an appropriate low-quality data detection method is selected. For example, statistical analysis, data mining, rule detection and other methods can be used to detect low-quality data. According to the selected low-quality data detection method, the historical product production data is tested. Appropriate tools and techniques, such as Python's pandas library and numpy library, can be used to write detection scripts or apply existing data quality tools. Historical low-quality production data is extracted from the test results. According to the signs or thresholds of the test results, low-quality data is identified and extracted to form a data set containing historical low-quality production data.
S23：对历史产品生产数据与历史设备性能参数进行时间序列同步对准，得到同步历史生产-设备数据集；S23: Perform time series synchronization alignment on historical product production data and historical equipment performance parameters to obtain a synchronized historical production-equipment data set;
具体地，例如，可以确定产品生产数据和设备性能参数数据的时间戳字段，并对两个数据集进行时间序列对准操作，以确保它们在时间上对应。这可以使用Python的pandas库进行时间序列操作和对准。将对准后的产品生产数据和设备性能参数数据进行合并，形成同步的历史生产-设备数据集。可以使用Python的pandas库的合并操作（例如merge）根据时间戳字段将两个数据集进行合并。Specifically, for example, the timestamp fields of the product production data and the equipment performance parameter data can be determined, and the two data sets can be aligned in time series to ensure that they correspond in time. This can be done using Python's pandas library for time series operations and alignment. The aligned product production data and equipment performance parameter data are merged to form a synchronized historical production-equipment data set. The two data sets can be merged based on the timestamp field using the merge operation (e.g., merge) of Python's pandas library.
S24：基于同步历史生产-设备数据集对历史低品质生产数据相应时间段内的历史设备性能参数进行设备运行效率评估，得到设备运行效率数据；S24: performing equipment operation efficiency evaluation on historical equipment performance parameters within a corresponding time period of historical low-quality production data based on the synchronized historical production-equipment data set to obtain equipment operation efficiency data;
具体地，例如，可以根据历史低品质生产数据，确定需要评估的时间段。从同步的历史生产-设备数据集中，提取与评估时间段相对应的设备性能参数数据。可以使用Python的pandas库进行时间范围的筛选和提取操作。使用适当的方法或模型对提取的设备性能参数数据进行运行效率评估。具体的评估方法可以根据业务需求和领域知识来选择，例如可以使用统计分析方法、机器学习模型或专业领域的算法进行评估。根据设备运行效率评估的结果，得到相应时间段内的设备运行效率数据。Specifically, for example, the time period that needs to be evaluated can be determined based on historical low-quality production data. The equipment performance parameter data corresponding to the evaluation time period is extracted from the synchronized historical production-equipment data set. The Python pandas library can be used to filter and extract time ranges. Use appropriate methods or models to evaluate the operating efficiency of the extracted equipment performance parameter data. The specific evaluation method can be selected based on business needs and domain knowledge. For example, statistical analysis methods, machine learning models, or algorithms in professional fields can be used for evaluation. Based on the results of the equipment operating efficiency evaluation, the equipment operating efficiency data for the corresponding time period is obtained.
S25：基于历史产品生产数据与历史设备性能参数根据设备运行效率数据进行数据品质影响因子挖掘，得到产品生产数据品质影响因子集；S25: mining data quality influencing factors based on historical product production data and historical equipment performance parameters according to equipment operation efficiency data to obtain a set of product production data quality influencing factors;
具体地，例如，可以从历史产品生产数据和历史设备性能参数中提取相关特征，例如，可以计算产品生产速率、设备故障次数、设备运行时间等。这可以使用Python的pandas库进行数据处理和特征提取。根据设备运行效率数据和相关特征，使用适当的数据挖掘方法（如相关性分析、回归分析、决策树等）来挖掘数据品质影响因子。这些方法可以帮助发现哪些因素对产品生产数据的品质产生重要影响。根据数据挖掘的结果，得到产品生产数据的品质影响因子集。这些因子可以是对产品生产数据质量具有显著影响的特征或变量。可以将这些因子整理为一个集合或数据表，记录其名称和影响程度等信息。Specifically, for example, relevant features can be extracted from historical product production data and historical equipment performance parameters. For example, product production rate, number of equipment failures, equipment operation time, etc. can be calculated. This can be done using Python's pandas library for data processing and feature extraction. Based on the equipment operating efficiency data and related features, appropriate data mining methods (such as correlation analysis, regression analysis, decision trees, etc.) are used to mine data quality influencing factors. These methods can help discover which factors have a significant impact on the quality of product production data. Based on the results of data mining, a set of quality influencing factors for product production data is obtained. These factors can be features or variables that have a significant impact on the quality of product production data. These factors can be organized into a set or data table, recording information such as their names and degree of influence.
S26：根据产品生产数据品质影响因子集与历史低品质生产数据进行数据品质临界值范围计算，得到生产数据可接受品质区间。S26: Calculate the data quality critical value range based on the product production data quality influencing factor set and historical low-quality production data to obtain the acceptable quality range of the production data.
具体地，例如，可以使用产品生产数据品质影响因子集作为参考，对历史低品质生产数据进行分析，了解每个品质影响因子在低品质数据中的取值范围。可以使用统计分析方法（如箱线图、直方图）来观察数据分布和异常情况。根据历史低品质生产数据和产品生产数据品质影响因子集，计算每个品质影响因子的临界值范围。可以根据统计指标（如均值、标准差）或领域知识来确定临界值的计算方法。根据计算得到的品质影响因子的临界值范围，确定生产数据的可接受品质区间。这个区间可以定义为低品质数据和高品质数据之间的范围，小于该范围的数据被认为是低品质数据。Specifically, for example, the set of product production data quality influencing factors can be used as a reference to analyze historical low-quality production data to understand the value range of each quality influencing factor in low-quality data. Statistical analysis methods (such as box plots and histograms) can be used to observe data distribution and abnormalities. Based on the historical low-quality production data and the set of product production data quality influencing factors, calculate the critical value range of each quality influencing factor. The calculation method of the critical value can be determined based on statistical indicators (such as mean, standard deviation) or domain knowledge. Based on the calculated critical value range of the quality influencing factor, determine the acceptable quality range of the production data. This interval can be defined as the range between low-quality data and high-quality data, and data smaller than this range is considered to be low-quality data.
本发明通过获取历史产品生产数据和历史设备性能参数，可以建立一个包含过去生产和设备信息的数据集。这些数据是过去生产过程的记录，可以提供有关产品生产和设备性能的详细信息。通过对历史产品生产数据进行低品质数据检测，可以识别和标记出数据中存在的低品质数据。低品质数据可能包括异常值、缺失值、重复值等，对数据的准确性和可靠性产生负面影响。通过识别低品质数据，可以为后续的数据清洗和处理提供指导，保证数据质量的提升。通过对历史产品生产数据和历史设备性能参数进行时间序列的同步对准，确保它们具有相同的时间轴和采样间隔。这样可以实现生产数据和设备数据之间的对应关系，构建同步的历史生产-设备数据集。基于同步的历史生产-设备数据集，对历史低品质生产数据对应时间段内的设备性能参数进行评估。这样可以分析设备在低品质数据发生时的运行效率，并得到设备运行效率数据。设备运行效率数据可以帮助了解设备在不同情况下的运行状况，为后续的数据品质影响因子挖掘提供重要依据。基于历史产品生产数据、历史设备性能参数和设备运行效率数据，进行数据品质影响因子的挖掘。通过分析这些数据之间的关系，可以识别出对产品生产数据品质具有重要影响的因子。这些因子可能包括设备性能、数据采集方法、环境条件等，对数据品质产生积极或消极的影响。通过挖掘这些影响因子，可以深入了解数据品质的形成机制，为改进数据采集和生产过程提供指导和优化建议。于产品生产数据品质影响因子集和历史低品质生产数据，进行数据品质临界值范围的计算。通过分析历史低品质生产数据与品质影响因子之间的关系，可以确定数据品质的临界值范围，即数据质量在可接受的范围内。这样可以为后续的数据质量控制和异常检测提供基准，确保生产数据处于可接受的品质区间内。The present invention can establish a data set containing past production and equipment information by acquiring historical product production data and historical equipment performance parameters. These data are records of past production processes and can provide detailed information about product production and equipment performance. By performing low-quality data detection on historical product production data, low-quality data in the data can be identified and marked. Low-quality data may include outliers, missing values, duplicate values, etc., which have a negative impact on the accuracy and reliability of the data. By identifying low-quality data, guidance can be provided for subsequent data cleaning and processing to ensure the improvement of data quality. By synchronizing the time series of historical product production data and historical equipment performance parameters, it is ensured that they have the same time axis and sampling interval. In this way, the corresponding relationship between production data and equipment data can be achieved, and a synchronized historical production-equipment data set can be constructed. Based on the synchronized historical production-equipment data set, the equipment performance parameters in the corresponding time period of historical low-quality production data are evaluated. In this way, the operating efficiency of the equipment when low-quality data occurs can be analyzed, and the equipment operating efficiency data can be obtained. The equipment operating efficiency data can help understand the operating status of the equipment under different conditions, and provide an important basis for the subsequent mining of data quality influencing factors. Based on historical product production data, historical equipment performance parameters and equipment operating efficiency data, data quality influencing factors are mined. By analyzing the relationship between these data, we can identify factors that have a significant impact on the quality of product production data. These factors may include equipment performance, data collection methods, environmental conditions, etc., which have a positive or negative impact on data quality. By mining these influencing factors, we can gain a deep understanding of the formation mechanism of data quality and provide guidance and optimization suggestions for improving data collection and production processes. Based on the set of factors affecting product production data quality and historical low-quality production data, the critical value range of data quality is calculated. By analyzing the relationship between historical low-quality production data and quality influencing factors, the critical value range of data quality can be determined, that is, the data quality is within an acceptable range. This can provide a benchmark for subsequent data quality control and anomaly detection to ensure that production data is within an acceptable quality range.
优选地，S25具体为：Preferably, S25 is specifically:
S251：将设备运行效率数据与预设的设备运行效率临界值进行比较；S251: Compare the equipment operation efficiency data with a preset equipment operation efficiency critical value;
具体地，例如，可以使用逻辑判断操作（如大于、小于、等于）来判断设备运行效率数据是否满足预设的设备运行效率临界值条件。Specifically, for example, a logical judgment operation (such as greater than, less than, equal to) may be used to judge whether the equipment operation efficiency data meets a preset equipment operation efficiency critical value condition.
S252：当设备运行效率数据小于或等于预设的设备运行效率临界值时，对历史设备性能参数进行泰拉矩阵分解，得到设备性能影响因子张量；S252: When the equipment operation efficiency data is less than or equal to a preset equipment operation efficiency critical value, perform Terra matrix decomposition on the historical equipment performance parameters to obtain an equipment performance influencing factor tensor;
具体地，例如，当设备运行效率数据小于或等于预设的设备运行效率临界值时，可以对历史设备性能参数数据进行泰拉矩阵分解（Tucker decomposition）。泰拉矩阵分解是一种高阶张量分解方法，可以将高维数据分解为低维核心张量和模态矩阵的乘积。可以使用Python的TensorLy库或其他张量分解工具库来实现泰拉矩阵分解。根据泰拉矩阵分解的结果，得到设备性能影响因子张量。泰拉矩阵分解将历史设备性能参数数据分解为核心张量和模态矩阵，其中模态矩阵表示了设备性能的不同影响因子。可以将这些影响因子整理为一个张量或数据结构，记录其名称和对设备性能的影响程度等信息。Specifically, for example, when the equipment operating efficiency data is less than or equal to the preset equipment operating efficiency critical value, the historical equipment performance parameter data can be subjected to Tucker decomposition. Tucker decomposition is a high-order tensor decomposition method that can decompose high-dimensional data into the product of a low-dimensional core tensor and a modal matrix. Python's TensorLy library or other tensor decomposition tool libraries can be used to implement Tucker decomposition. According to the results of Tucker decomposition, the equipment performance influencing factor tensor is obtained. Tucker decomposition decomposes the historical equipment performance parameter data into a core tensor and a modal matrix, where the modal matrix represents the different influencing factors of the equipment performance. These influencing factors can be organized into a tensor or data structure to record information such as their names and the degree of impact on equipment performance.
S253：对设备性能影响因子张量进行超图卷积特征提取，得到设备性能影响因子集；S253: performing hypergraph convolution feature extraction on the device performance influencing factor tensor to obtain a device performance influencing factor set;
具体地，例如，可以基于设备性能影响因子张量，构建超图。超图是一种图结构，其中节点表示影响因子，边表示影响因子之间的关系。可以根据设备性能影响因子张量的特点和领域知识，确定超图中的节点和边的定义。使用超图卷积神经网络（GraphConvolutional Network，GCN）或其他超图卷积方法，对超图进行特征提取。超图卷积是一种适用于超图结构的图卷积方法，可以捕捉节点之间的高阶关系。可以使用Python的DGL库、PyTorch Geometric库或其他图神经网络库来实现超图卷积特征提取。根据超图卷积特征提取的结果，得到设备性能影响因子集。这些影响因子可以是超图卷积提取的节点特征，表示了设备性能影响因子的重要程度和相关性。可以将这些因子整理为一个集合或数据表，记录其名称和对设备性能的影响程度等信息。Specifically, for example, a hypergraph can be constructed based on the device performance impact factor tensor. A hypergraph is a graph structure in which nodes represent impact factors and edges represent relationships between impact factors. The definitions of nodes and edges in the hypergraph can be determined based on the characteristics of the device performance impact factor tensor and domain knowledge. A hypergraph convolutional neural network (GCN) or other hypergraph convolution methods are used to extract features from the hypergraph. Hypergraph convolution is a graph convolution method suitable for hypergraph structures that can capture high-order relationships between nodes. Hypergraph convolution feature extraction can be implemented using Python's DGL library, PyTorch Geometric library, or other graph neural network libraries. Based on the results of the hypergraph convolution feature extraction, a set of device performance impact factors is obtained. These impact factors can be node features extracted by hypergraph convolution, which represent the importance and relevance of the device performance impact factors. These factors can be organized into a set or data table to record information such as their names and the degree of impact on device performance.
S254：当设备运行效率数据大于预设的设备运行效率临界值时，对目标生产车间进行供应链溯源，得到供应链环节数据；S254: When the equipment operation efficiency data is greater than a preset equipment operation efficiency critical value, the supply chain is traced to the target production workshop to obtain supply chain link data;
具体地，例如，当设备运行效率数据大于预设的设备运行效率临界值时，进行供应链溯源。通过追踪原材料、零部件或组装过程等信息，获取与目标生产车间相关的供应链环节数据。这可以涉及到与供应商和其他合作伙伴进行数据交换和查询。根据供应链溯源的结果，得到目标生产车间的供应链环节数据。这些数据可以包括原材料来源、供应商信息、生产批次、质检记录等。可以将这些数据整理为一个数据集或数据表，记录相关的供应链信息。Specifically, for example, when the equipment operating efficiency data is greater than the preset equipment operating efficiency critical value, supply chain traceability is performed. By tracking information such as raw materials, parts or assembly processes, supply chain link data related to the target production workshop is obtained. This may involve data exchange and query with suppliers and other partners. Based on the results of supply chain traceability, the supply chain link data of the target production workshop is obtained. These data may include raw material sources, supplier information, production batches, quality inspection records, etc. These data can be organized into a data set or data table to record relevant supply chain information.
S255：对供应链环节数据进行供应链稳定性冲击仿真，得到供应链扰动因子集；S255: Conduct supply chain stability shock simulation on supply chain link data to obtain a supply chain disturbance factor set;
具体地，例如，可以使用供应链稳定性冲击仿真方法，对供应链环节数据进行模拟。这可以通过引入随机变量、模型参数调整或其他仿真方法来模拟供应链中的不确定性和扰动情况。可以使用Python的仿真库（如SimPy、NumPy）来实现供应链稳定性冲击仿真。根据供应链稳定性冲击仿真的结果，得到供应链扰动因子集。这些因子可以是仿真中引入的扰动变量，表示供应链中的不稳定性和风险因素。可以将这些因子整理为一个集合或数据表，记录其名称和对供应链稳定性的影响程度等信息。Specifically, for example, the supply chain stability shock simulation method can be used to simulate the supply chain link data. This can be done by introducing random variables, adjusting model parameters, or other simulation methods to simulate the uncertainty and disturbance in the supply chain. Python simulation libraries (such as SimPy, NumPy) can be used to implement supply chain stability shock simulation. According to the results of the supply chain stability shock simulation, a set of supply chain disturbance factors is obtained. These factors can be disturbance variables introduced in the simulation, representing the instability and risk factors in the supply chain. These factors can be organized into a set or data table, recording their names and the degree of impact on the stability of the supply chain.
S256：对目标生产车间进行历史人机交互数据采集，得到历史人机交互数据；对历史人机交互数据进行行为模式挖掘与偏差检测，得到人工录入操作失误影响因子集；S256: Collect historical human-computer interaction data of the target production workshop to obtain historical human-computer interaction data; perform behavior pattern mining and deviation detection on the historical human-computer interaction data to obtain a set of influencing factors of manual input operation errors;
具体地，例如，可以使用传感器、记录设备或其他数据采集方法来获取与人机交互相关的数据，包括操作记录、输入数据、时间戳等。使用行为模式挖掘方法，对历史人机交互数据进行分析和挖掘。可以使用机器学习算法（如聚类、关联规则挖掘）或序列模式挖掘方法（如序列模式、序列频繁项挖掘）来发现人机交互的行为模式。可以使用Python的机器学习库（如scikit-learn、TensorFlow）或序列分析库（如PrefixSpan、GSPY）来实现行为模式挖掘。对行为模式进行偏差检测，识别潜在的人工录入操作失误。可以使用异常检测算法（如离群点检测、统计方法）或专门设计的规则和模型来检测偏差。根据异常检测的结果，确定可能存在的人工录入操作失误。根据偏差检测的结果，得到人工录入操作失误影响因子集。这些因子可以是与人工录入操作失误相关的特征、模式或其他指标，表示其对生产车间操作的影响程度和频率。可以将这些因子整理为一个集合或数据表，记录其名称和对生产车间操作的影响程度等信息。Specifically, for example, sensors, recording devices or other data collection methods can be used to obtain data related to human-computer interaction, including operation records, input data, timestamps, etc. Use behavioral pattern mining methods to analyze and mine historical human-computer interaction data. Machine learning algorithms (such as clustering, association rule mining) or sequence pattern mining methods (such as sequence patterns, sequence frequent item mining) can be used to discover behavioral patterns of human-computer interaction. Python machine learning libraries (such as scikit-learn, TensorFlow) or sequence analysis libraries (such as PrefixSpan, GSPY) can be used to implement behavioral pattern mining. Deviation detection is performed on behavioral patterns to identify potential manual input operation errors. Anomaly detection algorithms (such as outlier detection, statistical methods) or specially designed rules and models can be used to detect deviations. Based on the results of anomaly detection, possible manual input operation errors are determined. Based on the results of deviation detection, a set of influencing factors for manual input operation errors is obtained. These factors can be features, patterns or other indicators related to manual input operation errors, indicating the degree and frequency of their impact on production workshop operations. These factors can be organized into a set or data table to record information such as their names and the degree of impact on production workshop operations.
S257：将设备性能影响因子集、供应链扰动因子集与人工录入操作失误影响因子集作为产品生产数据品质影响因子集。S257: The equipment performance influencing factor set, the supply chain disturbance factor set and the manual input operation error influencing factor set are used as the product production data quality influencing factor set.
具体地，例如，可以将设备性能影响因子集、供应链扰动因子集和人工录入操作失误影响因子集进行整合。根据数据的结构和需求，可以使用数据处理工具（如Python的pandas库）将这些因子集合并为一个数据集或数据表。确保每个因子都有对应的标识符或关联信息，以便后续分析和处理。将整合后的因子集作为产品生产数据品质影响因子集。这个影响因子集包含了设备性能、供应链稳定性和人工录入操作等多个方面对产品生产数据品质的影响因素。可以将这些因子整理为一个集合或数据表，记录其名称和对产品生产数据品质的影响程度等信息。Specifically, for example, the set of factors affecting equipment performance, the set of factors affecting supply chain disturbances, and the set of factors affecting manual entry errors can be integrated. Depending on the structure and requirements of the data, data processing tools (such as Python's pandas library) can be used to merge these factor sets into a data set or data table. Make sure that each factor has a corresponding identifier or associated information for subsequent analysis and processing. The integrated factor set is used as the set of factors affecting the quality of product production data. This set of factors includes factors affecting the quality of product production data from multiple aspects such as equipment performance, supply chain stability, and manual entry operations. These factors can be organized into a set or data table, recording their names and information such as the degree of impact on the quality of product production data.
本发明通过将设备运行效率数据与预设的设备运行效率临界值进行比较，可以评估设备的运行情况。如果设备运行效率数据低于或等于预设的临界值，可以判定设备的运行状态存在问题，需要进一步分析设备性能的影响因素。当设备运行效率数据小于或等于预设的设备运行效率临界值时，采用泰拉矩阵分解的方法对历史设备性能参数进行分析。泰拉矩阵分解可以从历史数据中提取出设备性能的影响因子，即设备性能影响因子张量。这些因子可以用于后续的数据分析和挖掘，帮助理解设备性能的变化和影响因素。通过对设备性能影响因子张量进行超图卷积特征提取，可以从中提取出设备性能的关键特征。超图卷积是一种用于处理图数据的方法，可以捕捉图中节点之间的复杂关系。通过超图卷积特征提取，可以得到设备性能影响因子集，这些因子具有更高的表征能力，可以更好地描述设备性能的影响因素。当设备运行效率数据大于预设的设备运行效率临界值时，说明设备的运行状态正常。在这种情况下，可以通过供应链溯源的方法追踪目标生产车间的供应链环节数据，包括原材料存量、供应商、生产过程等信息。对供应链环节数据进行供应链稳定性冲击仿真，模拟不同情况下的供应链稳定性，并从中提取供应链中的扰动因子。这些扰动因子可以帮助理解供应链的脆弱性和稳定性，为改进供应链管理和应对潜在风险提供依据。通过对历史人机交互数据进行行为模式挖掘与偏差检测，可以识别出人工录入操作中的偏差和错误。提取出的人工录入操作失误影响因子集可以帮助分析和改进人机交互过程，减少人为因素对产品生产数据品质的影响。将之前提取的设备性能影响因子集、供应链扰动因子集和人工录入操作失误影响因子集整合在一起，形成产品生产数据品质影响因子集。这个集合包含了影响产品生产数据品质的关键因素，可以用于评估和优化产品生产过程，提高数据品质和生产效率。The present invention can evaluate the operation of the equipment by comparing the equipment operation efficiency data with the preset equipment operation efficiency critical value. If the equipment operation efficiency data is lower than or equal to the preset critical value, it can be determined that there is a problem with the operation state of the equipment, and it is necessary to further analyze the influencing factors of the equipment performance. When the equipment operation efficiency data is less than or equal to the preset equipment operation efficiency critical value, the historical equipment performance parameters are analyzed by the Terra matrix decomposition method. Terra matrix decomposition can extract the influencing factors of equipment performance from the historical data, that is, the equipment performance influencing factor tensor. These factors can be used for subsequent data analysis and mining to help understand the changes and influencing factors of equipment performance. By performing hypergraph convolution feature extraction on the equipment performance influencing factor tensor, the key features of the equipment performance can be extracted from it. Hypergraph convolution is a method for processing graph data, which can capture the complex relationship between nodes in the graph. Through hypergraph convolution feature extraction, a set of equipment performance influencing factors can be obtained, which have higher characterization capabilities and can better describe the influencing factors of equipment performance. When the equipment operation efficiency data is greater than the preset equipment operation efficiency critical value, it means that the operation state of the equipment is normal. In this case, the supply chain link data of the target production workshop can be tracked by the supply chain traceability method, including information such as raw material inventory, suppliers, and production processes. Supply chain stability impact simulation is performed on the supply chain link data to simulate the supply chain stability under different circumstances, and the disturbance factors in the supply chain are extracted from it. These disturbance factors can help understand the fragility and stability of the supply chain, and provide a basis for improving supply chain management and responding to potential risks. By mining behavioral patterns and detecting deviations in historical human-computer interaction data, deviations and errors in manual input operations can be identified. The extracted set of factors affecting manual input errors can help analyze and improve the human-computer interaction process and reduce the impact of human factors on the quality of product production data. The previously extracted set of equipment performance influencing factors, supply chain disturbance factors, and manual input error influencing factors are integrated together to form a set of factors affecting product production data quality. This set contains the key factors that affect the quality of product production data, which can be used to evaluate and optimize the product production process and improve data quality and production efficiency.
优选地，S26具体为：Preferably, S26 is specifically:
S261：对历史低品质生产数据进行缺陷成本估算，得到缺陷成本数据；S261: Estimating defect cost for historical low-quality production data to obtain defect cost data;
具体地，例如，可以使用缺陷成本估算方法，对历史低品质生产数据进行分析和计算。缺陷成本包括直接成本（如废品处理、返工成本）、间接成本（如停产损失、客户投诉处理成本）等。具体的成本计算方法可以根据企业的需求和实际情况进行选择和定制。可以使用Excel或专门的成本估算工具来进行计算和分析。根据缺陷成本估算的结果，得到缺陷成本数据。这些数据可以是每个缺陷类型或生产批次对应的成本值，用于衡量低品质生产数据对企业的经济影响。可以将这些数据整理为一个集合或数据表，记录其名称、对应的缺陷类型或生产批次，以及对企业经济的影响程度等信息。Specifically, for example, the defect cost estimation method can be used to analyze and calculate historical low-quality production data. Defect costs include direct costs (such as scrap disposal, rework costs), indirect costs (such as production stoppage losses, customer complaint handling costs), etc. The specific cost calculation method can be selected and customized according to the needs and actual situation of the enterprise. Excel or a special cost estimation tool can be used for calculation and analysis. According to the results of the defect cost estimation, the defect cost data is obtained. These data can be the cost values corresponding to each defect type or production batch, which are used to measure the economic impact of low-quality production data on the enterprise. These data can be organized into a collection or data table, recording its name, the corresponding defect type or production batch, and the degree of impact on the enterprise's economy.
S262：对缺陷成本数据进行可接受损失阈值设定，得到可接受损失阈值；S262: Setting an acceptable loss threshold for the defect cost data to obtain an acceptable loss threshold;
具体地，例如，可以使用适当的方法或模型来设定可接受损失阈值。常用的方法包括风险评估、成本效益分析、质量目标设定等。可以根据企业的需求选择合适的方法。这个阈值应该能够平衡经济投入和质量风险，确保企业在可接受的损失范围内运营。根据阈值设定的结果，得到可接受损失阈值。这个阈值可以是一个具体的数值，表示企业能够接受的最大缺陷成本或损失金额。也可以是一个范围或区间，以考虑不同的情况和灵活性。确保记录阈值的具体数值和相关的定义或说明。Specifically, for example, an appropriate method or model can be used to set an acceptable loss threshold. Common methods include risk assessment, cost-benefit analysis, quality goal setting, etc. Appropriate methods can be selected according to the needs of the enterprise. This threshold should be able to balance economic input and quality risk to ensure that the enterprise operates within an acceptable loss range. Based on the results of the threshold setting, an acceptable loss threshold is obtained. This threshold can be a specific value, indicating the maximum defect cost or loss amount that the enterprise can accept. It can also be a range or interval to take into account different situations and flexibility. Make sure to record the specific value of the threshold and the relevant definition or description.
S263：根据可接受损失阈值对历史低品质生产数据进行可接受历史低品质生产数据反求，得到可接受历史低品质生产数据；S263: reverse the historical low-quality production data according to the acceptable loss threshold to obtain the acceptable historical low-quality production data;
具体地，例如，可以根据可接受的缺陷成本范围，筛选历史低品质生产数据，找出符合阈值要求的数据。可以根据缺陷成本数据和阈值进行筛选和过滤，找到缺陷成本在可接受范围内的生产数据。得到符合可接受缺陷成本范围的可接受历史低品质生产数据。Specifically, for example, historical low-quality production data can be screened according to an acceptable defect cost range to find data that meets the threshold requirements. Production data with defect costs within an acceptable range can be screened and filtered according to defect cost data and thresholds to find acceptable historical low-quality production data that meets the acceptable defect cost range.
S264：获取可接受历史低品质生产数据相应设备性能参数；S264: Obtaining equipment performance parameters corresponding to acceptable historical low-quality production data;
具体地，例如，可以根据可接受历史低品质生产数据，获取相应的设备性能参数。确定与设备性能相关的参数和指标。这些参数可以是设备的运行状态、工作效率、故障记录等。根据需求和实际情况，选择适当的性能参数来评估设备的运行质量。将可接受历史低品质生产数据与设备性能参数进行整合和关联。根据生产批次或产品缺陷信息，找到对应的设备性能参数数据。这可以通过批次号、时间戳或其他唯一标识符来进行匹配和关联。根据关联的设备性能参数定义和数据整合结果，获取相应的设备性能参数。这可以涉及从设备监控系统、传感器数据、维修记录或其他相关数据源中提取所需的性能参数。Specifically, for example, the corresponding equipment performance parameters can be obtained based on the acceptable historical low-quality production data. Determine the parameters and indicators related to equipment performance. These parameters can be the operating status, work efficiency, fault records, etc. of the equipment. According to the needs and actual conditions, select appropriate performance parameters to evaluate the operating quality of the equipment. Integrate and associate the acceptable historical low-quality production data with the equipment performance parameters. Find the corresponding equipment performance parameter data based on the production batch or product defect information. This can be matched and associated by batch number, timestamp or other unique identifier. Obtain the corresponding equipment performance parameters based on the associated equipment performance parameter definition and data integration results. This can involve extracting the required performance parameters from the equipment monitoring system, sensor data, maintenance records or other relevant data sources.
S265：基于产品生产数据品质影响因子集对可接受历史低品质生产数据与可接受历史低品质生产数据相应设备性能参数进行影响因子赋权，得到加权影响因子数据；S265: weighting the acceptable historical low-quality production data and the equipment performance parameters corresponding to the acceptable historical low-quality production data based on the product production data quality impact factor set to obtain weighted impact factor data;
具体地，例如，可以使用合适的方法或模型，对影响因子进行赋权。可以使用的方法包括专家评估、层次分析法（AHP）、加权平均法等。根据企业的需求和可用的数据，选择最适合的赋权方法。确保赋权过程可重复且具有合理的逻辑。根据所选的赋权方法，对可接受历史低品质生产数据和相应设备性能参数的影响因子进行赋权。这可以涉及根据定义的影响因子集，使用赋权表或公式来计算每个影响因子的权重。使用Excel或类似的工具进行计算和分析。根据所选的赋权方法，对可接受历史低品质生产数据和相应设备性能参数的影响因子进行赋权。Specifically, for example, appropriate methods or models can be used to weight influencing factors. Available methods include expert evaluation, analytic hierarchy process (AHP), weighted average method, etc. Select the most suitable weighting method based on the needs of the enterprise and the available data. Ensure that the weighting process is repeatable and has reasonable logic. According to the selected weighting method, weight the influencing factors of acceptable historical low-quality production data and corresponding equipment performance parameters. This may involve using a weighting table or formula to calculate the weight of each influencing factor based on the defined set of influencing factors. Use Excel or similar tools for calculation and analysis. According to the selected weighting method, weight the influencing factors of acceptable historical low-quality production data and corresponding equipment performance parameters.
S266：根据产品生产数据品质影响因子集与加权影响因子数据对可接受历史低品质生产数据与可接受历史低品质生产数据相应设备性能参数进行数据品质临界值范围计算，得到生产数据可接受品质区间；S266: Calculate the data quality critical value range of acceptable historical low-quality production data and corresponding equipment performance parameters of acceptable historical low-quality production data according to the product production data quality influencing factor set and weighted influencing factor data, and obtain the acceptable quality range of production data;
具体地，例如，可以使用适当的方法或模型来计算数据品质的临界值范围。可以使用的方法包括基于统计分析、质量控制图、专家判断等。根据企业的需求选择合适的方法。确保计算方法合理、可靠且与目标相符。根据影响因子的加权结果，对可接受历史低品质生产数据和设备性能参数进行加权求和，得到综合的品质评估值。这可以通过使用加权逐项相加的方法，将每个数据点的加权影响因子与对应的数据值相乘，并将结果求和得到一个综合评估值。根据数据品质临界值计算方法，将综合的品质评估值转化为数据品质临界值范围。这可以通过可接受历史低品质生产数据对应的数据品质临界值进行范围确定，或者通过根据企业的需求和标准，设定上限和下限的阈值来确定数据品质的可接受范围。使用适当的工具或方法，对综合的品质评估值进行数据品质临界值范围的计算。这可以涉及使用统计分析软件（如Excel、Python的numpy库）进行计算和处理。根据计算结果，得到生产数据的可接受品质区间。这可以是一个范围，例如上限和下限的数值，表示数据品质的可接受范围。Specifically, for example, an appropriate method or model can be used to calculate the critical value range of data quality. The methods that can be used include those based on statistical analysis, quality control charts, expert judgment, etc. Select an appropriate method according to the needs of the enterprise. Ensure that the calculation method is reasonable, reliable and consistent with the target. According to the weighted results of the influencing factors, the acceptable historical low-quality production data and equipment performance parameters are weighted and summed to obtain a comprehensive quality evaluation value. This can be done by using the weighted item-by-item addition method to multiply the weighted influencing factor of each data point by the corresponding data value, and sum the results to obtain a comprehensive evaluation value. According to the data quality critical value calculation method, the comprehensive quality evaluation value is converted into a data quality critical value range. This can be determined by the data quality critical value corresponding to the acceptable historical low-quality production data, or by setting upper and lower thresholds according to the needs and standards of the enterprise to determine the acceptable range of data quality. Use appropriate tools or methods to calculate the data quality critical value range for the comprehensive quality evaluation value. This may involve using statistical analysis software (such as Excel, Python's numpy library) for calculation and processing. According to the calculation results, the acceptable quality interval of the production data is obtained. This can be a range, such as upper and lower numerical values, indicating an acceptable range of data quality.
本发明通过对历史低品质生产数据进行缺陷成本估算，可以计算出与缺陷相关的成本，如返工成本、废品成本等。这些缺陷成本数据有助于评估低品质数据对生产过程和企业经济的影响，为后续的决策提供定量依据。通过对缺陷成本数据进行分析和考量，可以设定一个可接受的损失阈值。这个阈值表示企业在生产过程中愿意承担的最大损失程度，超过该阈值的低品质数据将被视为不可接受的质量问题。可接受损失阈值的设定有助于明确生产数据品质的目标和限制。根据设定的可接受损失阈值，对历史低品质生产数据进行反求，筛选出符合可接受标准的数据。这些可接受历史低品质生产数据是在企业可接受损失范围内的数据。根据可接受历史低品质生产数据，获取与这些数据相关联的设备性能参数。这些参数反映了设备在生产过程中的关键性能指标，与低品质数据的产生可能存在关联关系。通过获取这些设备性能参数，可以进一步分析和理解设备对低品质数据的影响。据产品生产数据品质影响因子集，对可接受历史低品质生产数据与相应设备性能参数进行影响因子赋权。这些影响因子可以根据其对产品生产数据品质的重要程度进行赋权，以反映它们对品质的相对重要性。通过加权影响因子数据，可以更准确地评估设备性能对可接受历史低品质生产数据的影响程度。基于产品生产数据品质影响因子集和加权影响因子数据，计算可接受历史低品质生产数据与相应设备性能参数的数据品质临界值范围。这个品质临界值范围定义了可接受的数据品质边界，超过这个范围的数据将被视为不可接受的低品质数据。通过确定生产数据的可接受品质区间，可以为数据的质量控制和改进提供指导The present invention can calculate the costs related to defects, such as rework costs, scrap costs, etc., by estimating the defect costs of historical low-quality production data. These defect cost data are helpful to evaluate the impact of low-quality data on the production process and enterprise economy, and provide quantitative basis for subsequent decision-making. By analyzing and considering the defect cost data, an acceptable loss threshold can be set. This threshold represents the maximum loss degree that the enterprise is willing to bear in the production process, and low-quality data exceeding this threshold will be regarded as unacceptable quality problems. The setting of an acceptable loss threshold helps to clarify the goals and limitations of production data quality. According to the set acceptable loss threshold, the historical low-quality production data is reversed to screen out data that meet the acceptable standards. These acceptable historical low-quality production data are data within the acceptable loss range of the enterprise. According to the acceptable historical low-quality production data, the equipment performance parameters associated with these data are obtained. These parameters reflect the key performance indicators of the equipment in the production process, and may be associated with the generation of low-quality data. By obtaining these equipment performance parameters, the impact of the equipment on low-quality data can be further analyzed and understood. According to the product production data quality influencing factor set, the acceptable historical low-quality production data and the corresponding equipment performance parameters are weighted by influencing factors. These influencing factors can be weighted according to their importance to the quality of product production data to reflect their relative importance to quality. By weighting the influencing factor data, the degree of influence of equipment performance on acceptable historical low-quality production data can be more accurately assessed. Based on the product production data quality influencing factor set and the weighted influencing factor data, the data quality critical value range of acceptable historical low-quality production data and corresponding equipment performance parameters is calculated. This quality critical value range defines the acceptable data quality boundary, and data exceeding this range will be considered unacceptable low-quality data. By determining the acceptable quality range of production data, guidance can be provided for data quality control and improvement.
优选地，数据品质评估模块S3具体为：Preferably, the data quality assessment module S3 is specifically:
S31：根据产品生产数据品质影响因子集对产品生产数据进行非影响因子属性数据剔除，得到精简产品生产数据；S31: removing attribute data of non-influencing factors from the product production data according to the product production data quality influencing factor set to obtain streamlined product production data;
具体地，例如，可以针对每个数据属性，根据产品生产数据品质影响因子集确定其是否为影响因子属性。影响因子属性是那些对产品生产数据品质有重要影响的属性。基于影响因子属性的定义和规则，对产品生产数据进行筛选和剔除。剔除那些不属于影响因子属性集的数据。确保数据剔除过程是可追溯和可重复的。记录剔除的数据属性和对应的剔除规则。在非影响因子属性数据剔除后，得到精简的产品生产数据集。Specifically, for example, for each data attribute, it can be determined whether it is an influencing factor attribute based on the product production data quality influencing factor set. Influencing factor attributes are those attributes that have a significant impact on the quality of product production data. Based on the definition and rules of influencing factor attributes, the product production data is screened and eliminated. Data that does not belong to the influencing factor attribute set is eliminated. Ensure that the data elimination process is traceable and repeatable. Record the eliminated data attributes and the corresponding elimination rules. After the non-influencing factor attribute data is eliminated, a streamlined product production data set is obtained.
S32：基于产品生产数据品质影响因子集构建产品生产数据品质评价模型；S32: Construct a product production data quality evaluation model based on the product production data quality influencing factor set;
具体地，例如，可以选择适当的评价模型来构建产品生产数据品质评价模型。可以使用的评价模型包括逻辑回归、支持向量机、决策树等。根据数据的特点和评价目标，选择最适合的模型。根据产品生产数据品质影响因子集，准备用于模型构建的数据集。这包括从产品生产数据中提取和整理与影响因子相关的数据属性，以及为每个数据样本标记品质评价结果（例如好/坏、合格/不合格）。将准备好的数据集分为训练集和测试集。使用训练集数据进行模型的训练。根据选择的评价模型算法，通过学习和调整模型参数，使其能够对产品生产数据的品质进行准确评价。使用测试集数据对训练好的模型进行验证。评估模型在未见过的数据上的性能表现，检查其对产品生产数据品质的评价准确性和可靠性，从而得到产品生产数据品质评价模型。Specifically, for example, an appropriate evaluation model can be selected to construct a product production data quality evaluation model. The evaluation models that can be used include logistic regression, support vector machine, decision tree, etc. Select the most suitable model according to the characteristics of the data and the evaluation objectives. Prepare a data set for model construction based on the set of factors affecting the quality of product production data. This includes extracting and organizing data attributes related to the influencing factors from the product production data, and marking the quality evaluation results (such as good/bad, qualified/unqualified) for each data sample. Divide the prepared data set into a training set and a test set. Use the training set data to train the model. According to the selected evaluation model algorithm, by learning and adjusting the model parameters, it can accurately evaluate the quality of product production data. Use the test set data to verify the trained model. Evaluate the performance of the model on unseen data, check the accuracy and reliability of its evaluation of the quality of product production data, and thus obtain a product production data quality evaluation model.
S33：将精简产品生产数据输入至产品生产数据品质评价模型进行数据品质量化评估，得到产品生产数据品质指数；将产品生产数据品质指数与生产数据可接受品质区间进行比较。S33: Input the streamlined product production data into the product production data quality evaluation model to perform quantitative evaluation of data quality to obtain the product production data quality index; compare the product production data quality index with the acceptable quality range of the production data.
具体地，例如，可以将精简产品生产数据输入到产品生产数据品质评价模型中进行评估。使用评价模型对输入的产品生产数据进行预测或分类。模型会根据已学习的规律，对数据的品质进行评估并生成相应的品质指数。对每个数据样本，得到相应的产品生产数据品质指数。该指数表示该样本的数据品质程度，可以是一个连续值或离散的评价结果。根据评价模型的输出结果，得到每个数据样本的产品生产数据品质指数。将产品生产数据品质指数与生产数据可接受品质区间进行比较，以确定数据的品质是否符合要求。Specifically, for example, the streamlined product production data can be input into the product production data quality evaluation model for evaluation. The evaluation model is used to predict or classify the input product production data. The model will evaluate the quality of the data based on the learned rules and generate a corresponding quality index. For each data sample, a corresponding product production data quality index is obtained. The index represents the data quality degree of the sample and can be a continuous value or a discrete evaluation result. Based on the output results of the evaluation model, the product production data quality index of each data sample is obtained. The product production data quality index is compared with the acceptable quality interval of the production data to determine whether the quality of the data meets the requirements.
本发明通过剔除产品生产数据中的非影响因子属性，得到精简的产品生产数据，提高数据的质量和相关性。根据产品生产数据品质影响因子集，识别并剔除那些对产品生产数据品质评估没有直接影响或无关的属性数据。这样做的价值在于减少数据集的复杂性，去除冗余信息，使得后续的数据分析和评估更加准确和高效。通过利用产品生产数据品质影响因子集，建立一个评价模型，该模型可以根据各个影响因子的权重和相关数据进行计算，从而得出产品生产数据的品质评分。构建产品生产数据品质评价模型的价值在于提供了一种客观、可量化的评估手段，使得企业能够更加准确地了解其产品生产数据的品质水平。通过将经过精简处理的产品生产数据输入到产品生产数据品质评价模型中，通过计算和分析得出产品生产数据的品质指数。品质指数可以反映数据的质量水平，以便进一步判断数据是否符合预期的品质要求。通过将产品生产数据品质指数与生产数据可接受品质区间进行比较，可以确定数据的合格性，并及时采取措施来改进和优化生产过程，保证数据的品质达到可接受水平。The present invention obtains streamlined product production data by eliminating non-influencing factor attributes in product production data, thereby improving the quality and relevance of the data. According to the product production data quality influencing factor set, attribute data that has no direct impact or is irrelevant to the product production data quality assessment is identified and eliminated. The value of doing so is to reduce the complexity of the data set, remove redundant information, and make subsequent data analysis and evaluation more accurate and efficient. By using the product production data quality influencing factor set, an evaluation model is established, which can be calculated according to the weights of each influencing factor and related data to obtain the quality score of the product production data. The value of constructing a product production data quality evaluation model is to provide an objective and quantifiable evaluation method, so that enterprises can more accurately understand the quality level of their product production data. By inputting the streamlined product production data into the product production data quality evaluation model, the quality index of the product production data is obtained by calculation and analysis. The quality index can reflect the quality level of the data, so as to further determine whether the data meets the expected quality requirements. By comparing the product production data quality index with the acceptable quality interval of the production data, the eligibility of the data can be determined, and measures can be taken in time to improve and optimize the production process to ensure that the quality of the data reaches an acceptable level.
优选地，数据品质提升决策模块S4具体为：Preferably, the data quality improvement decision module S4 is specifically:
S41：当产品生产数据品质指数小于生产数据可接受品质区间下限时，对目标生产车间进行智能生产停线作业；S41: When the product production data quality index is less than the lower limit of the acceptable quality range of the production data, the target production workshop is intelligently shut down;
具体地，例如，当产品生产数据品质指数小于生产数据可接受品质区间下限时，根据预先设定的生产停线策略，通过自动化控制系统停止目标生产车间的生产线。可以使用基于规则的逻辑控制或基于机器学习的智能控制方法来实现智能停线作业。Specifically, for example, when the product production data quality index is less than the lower limit of the acceptable quality range of the production data, the production line of the target production workshop is stopped by the automated control system according to the pre-set production stop strategy. Intelligent line stop operation can be achieved using rule-based logical control or machine learning-based intelligent control methods.
S42：根据产品生产数据品质影响因子集对产品生产数据进行因果追溯规则构建，得到因果追溯规则数据；S42: constructing causal tracing rules for product production data according to a set of product production data quality influencing factors to obtain causal tracing rule data;
具体地，例如，可以使用关联规则挖掘算法（如Apriori算法）来发现产品生产数据品质影响因子集对产品生产数据之间的关联性和因果关系。使用Python的数据挖掘和关联规则挖掘库（如mlxtend）来实现规则构建过程。根据构建的因果追溯规则，将规则转化为可读性高的形式，生成因果追溯规则数据。这些规则数据可以包括规则的条件、结果和支持度等信息。Specifically, for example, association rule mining algorithms (such as Apriori algorithm) can be used to discover the correlation and causal relationship between the product production data quality influencing factor set and the product production data. Use Python's data mining and association rule mining libraries (such as mlxtend) to implement the rule construction process. According to the constructed causal tracing rules, the rules are converted into a highly readable form to generate causal tracing rule data. These rule data can include information such as the conditions, results, and support of the rules.
S43：获取当前供应链数据；基于因果追溯规则数据、设备性能参数与当前供应链数据对产品生产数据进行数据异常根源追溯，得到生产数据异常诱因数据；S43: Acquire current supply chain data; trace the root cause of data anomaly of product production data based on causal tracing rule data, equipment performance parameters and current supply chain data to obtain production data anomaly inducement data;
具体地，例如，可以与供应链相关的各个环节，包括原材料供应商、物流信息、生产计划等，提供数据接口或采集方式，以获取当前的供应链数据。可以使用企业资源计划（ERP）系统、供应链管理（SCM）系统或其他相关系统来获取数据。基于因果追溯规则数据对当前供应链数据与设备性能参数进行关联和分析，以追溯产品生产数据的异常根源。可以使用数据挖掘和关联规则挖掘的方法，根据规则数据中的条件和结果进行匹配和推理，找出导致生产数据异常的诱因。使用Python的数据分析和关联规则挖掘库（如mlxtend）进行数据处理和分析。根据数据异常根源追溯的结果，提取出导致生产数据异常的诱因数据。这些诱因数据可以包括供应链环节、设备状态、工艺参数等方面的信息。Specifically, for example, data interfaces or collection methods can be provided for various links related to the supply chain, including raw material suppliers, logistics information, production plans, etc., to obtain current supply chain data. Enterprise resource planning (ERP) systems, supply chain management (SCM) systems, or other related systems can be used to obtain data. Based on the causal traceability rule data, the current supply chain data and equipment performance parameters are associated and analyzed to trace the abnormal root causes of product production data. Data mining and association rule mining methods can be used to match and infer the conditions and results in the rule data to find out the causes of abnormal production data. Use Python's data analysis and association rule mining library (such as mlxtend) for data processing and analysis. According to the results of the data anomaly root cause tracing, extract the cause data that causes abnormal production data. These cause data can include information on supply chain links, equipment status, process parameters, etc.
S44：基于历史产品生产数据、历史低品质生产数据与生产数据异常诱因数据对目标生产车间进行数据质量提升优化决策，得到生产数据质量增强策略；S44: Based on historical product production data, historical low-quality production data and production data abnormality inducement data, data quality improvement optimization decision is made for the target production workshop to obtain production data quality enhancement strategy;
具体地，例如，可以结合历史产品生产数据、历史低品质生产数据和生产数据异常诱因数据，制定数据质量提升的优化决策。基于数据分析的结果，识别和分析导致低品质生产的因素和异常诱因数据之间的关联。根据分析结果，制定一系列数据质量提升的优化决策，包括调整生产工艺参数、改进供应链管理、优化设备维护计划等。Specifically, for example, historical product production data, historical low-quality production data, and production data abnormality inducement data can be combined to make optimization decisions to improve data quality. Based on the results of data analysis, the association between factors leading to low-quality production and abnormal inducement data is identified and analyzed. Based on the analysis results, a series of optimization decisions to improve data quality are made, including adjusting production process parameters, improving supply chain management, optimizing equipment maintenance plans, etc.
S45：根据生产数据质量增强策略对目标生产车间进行策略深度融合实施。S45: Implement deep integration of strategies in the target production workshop based on the production data quality enhancement strategy.
具体地，例如，可以根据生产数据质量增强策略中的应对手段对目标生产车间进行及时应用。Specifically, for example, the response measures in the production data quality enhancement strategy can be promptly applied to the target production workshop.
本发明通过监测产品生产数据品质指数，如果发现指数低于生产数据可接受品质区间下限，就触发智能生产停线作业。低品质的工业数据可能对供应链产生负面影响。通过及时停线作业，可以避免不良数据传递给下游环节，确保供应链中的数据质量。通过构建因果追溯规则，可以识别出对产品生产数据品质影响较大的因素，并提供因果关系的参考依据，帮助理解和解决数据品质问题。这些规则可以帮助企业了解各个因素对产品生产数据品质的影响程度，从而更好地进行数据品质管理和改进。因果追溯规则数据的生成为企业提供了深入了解数据异常的能力，有助于确定数据质量问题的根本原因。通过获取当前供应链数据，并结合之前构建的因果追溯规则数据和设备性能参数，对产品生产数据进行异常根源追溯。通过分析数据异常的原因和来源，可以得到导致生产数据异常的具体诱因，包括供应链中的问题、设备性能等。这样的追溯有助于准确定位问题，提供解决方案的依据，为改进生产数据品质提供指导。通过利用历史产品生产数据、历史低品质生产数据以及之前得到的生产数据异常诱因数据，对目标生产车间进行数据质量提升优化决策。通过分析历史数据和异常诱因数据，可以识别出生产数据质量问题的主要原因，并制定相应的改进策略和措施。这些决策可能涉及供应链管理、设备维护、工艺调整等方面，旨在消除或减少导致数据异常的因素，提高生产数据的品质和稳定性。根据制定的生产数据质量增强策略，将这些策略深度融合到目标生产车间的实际生产过程中。这可能包括更新和调整供应链管理流程、设备维护计划、数据采集和监控系统等，以确保策略的有效实施和持续改进。通过策略的深度融合实施，可以提高数据质量的稳定性和可持续性。The present invention monitors the quality index of product production data. If the index is found to be lower than the lower limit of the acceptable quality range of production data, the intelligent production line stop operation is triggered. Low-quality industrial data may have a negative impact on the supply chain. By stopping the line operation in time, bad data can be prevented from being passed to downstream links, ensuring the data quality in the supply chain. By constructing causal tracing rules, factors that have a greater impact on the quality of product production data can be identified, and a reference basis for causal relationships can be provided to help understand and solve data quality problems. These rules can help companies understand the degree of influence of various factors on the quality of product production data, so as to better manage and improve data quality. The generation of causal tracing rule data provides companies with the ability to deeply understand data anomalies and helps to determine the root cause of data quality problems. By obtaining the current supply chain data and combining the previously constructed causal tracing rule data and equipment performance parameters, the abnormal root cause of product production data is traced. By analyzing the causes and sources of data anomalies, the specific causes of production data anomalies can be obtained, including problems in the supply chain, equipment performance, etc. Such tracing helps to accurately locate problems, provide a basis for solutions, and provide guidance for improving production data quality. By using historical product production data, historical low-quality production data, and previously obtained production data abnormality inducement data, data quality improvement optimization decisions are made for the target production workshop. By analyzing historical data and abnormality inducement data, the main causes of production data quality problems can be identified, and corresponding improvement strategies and measures can be formulated. These decisions may involve supply chain management, equipment maintenance, process adjustment, etc., aiming to eliminate or reduce factors that cause data abnormalities and improve the quality and stability of production data. According to the formulated production data quality enhancement strategy, these strategies are deeply integrated into the actual production process of the target production workshop. This may include updating and adjusting supply chain management processes, equipment maintenance plans, data collection and monitoring systems, etc., to ensure the effective implementation and continuous improvement of the strategy. Through the deep integration and implementation of strategies, the stability and sustainability of data quality can be improved.
优选地，S44具体为：Preferably, S44 is specifically:
S441：获取历史生产数据异常诱因数据集；S441: Obtaining a dataset of abnormal causes of historical production data;
具体地，例如，可以根据企业数据库的类型和结构，编写查询语句以获取历史生产数据异常诱因数据集。Specifically, for example, a query statement may be written according to the type and structure of an enterprise database to obtain a dataset of abnormality causes of historical production data.
S442：对历史生产数据异常诱因数据集进行异质诱因识别与分类，得到异质诱因数据集，其中异质诱因数据集包括设备故障异质诱因数据、人工操作失误异质诱因数据、供应链中断异质诱因数据以及环境影响异质诱因数据；S442: performing heterogeneous inducement identification and classification on the abnormal inducement data set of historical production data to obtain a heterogeneous inducement data set, wherein the heterogeneous inducement data set includes equipment failure heterogeneous inducement data, manual operation error heterogeneous inducement data, supply chain interruption heterogeneous inducement data, and environmental impact heterogeneous inducement data;
具体地，例如，可以从历史生产数据异常诱因数据集中提取相关特征，以便进行异质诱因的识别和分类。特征可以包括但不限于设备标识、时间戳、操作员信息、供应链信息和环境参数等。根据特征提取的结果，构建异质诱因识别与分类的模型。可以使用各种机器学习算法（如决策树、支持向量机、随机森林等）或深度学习方法（如神经网络）来构建模型。准备标注好的训练数据集，其中包含已经人工标注的异质诱因类别标签或由机器自动学习并标注的异质诱因类别标签。使用训练好的异质诱因识别与分类模型，对历史生产数据异常诱因数据集进行预测和分类。将预测结果与原始数据集进行关联，生成包括设备故障异质诱因数据、人工操作失误异质诱因数据、供应链中断异质诱因数据以及环境影响异质诱因数据的异质诱因数据集。Specifically, for example, relevant features can be extracted from the historical production data abnormal inducement data set to identify and classify heterogeneous inducements. Features may include but are not limited to equipment identification, timestamp, operator information, supply chain information, and environmental parameters. Based on the results of feature extraction, a model for heterogeneous inducement identification and classification is constructed. Various machine learning algorithms (such as decision trees, support vector machines, random forests, etc.) or deep learning methods (such as neural networks) can be used to build models. Prepare a labeled training data set, which contains manually labeled heterogeneous inducement category labels or heterogeneous inducement category labels automatically learned and labeled by the machine. Use the trained heterogeneous inducement identification and classification model to predict and classify the historical production data abnormal inducement data set. Associate the prediction results with the original data set to generate a heterogeneous inducement data set including equipment failure heterogeneous inducement data, manual operation error heterogeneous inducement data, supply chain interruption heterogeneous inducement data, and environmental impact heterogeneous inducement data.
S443： 基于历史产品生产数据、历史低品质生产数据与历史生产数据异常诱因数据集对异质诱因数据集中每一类异质诱因数据进行异常发生概率统计分析，得到各类异常诱因发生概率数据；S443: Based on the historical product production data, the historical low-quality production data and the historical production data abnormal inducement data set, a statistical analysis of the abnormal occurrence probability of each type of heterogeneous inducement data in the heterogeneous inducement data set is performed to obtain the occurrence probability data of each type of abnormal inducement;
具体地，例如，可以对异质诱因数据集中的每一类异质诱因数据进行异常发生概率的统计分析。针对每一类异质诱因数据，将历史产品生产数据中的对应低品质数据与历史生产数据异常诱因数据集进行关联。统计分析可以基于关联数据，计算每一类异质诱因数据的异常发生概率，如使用频率统计或概率分布拟合等方法。Specifically, for example, a statistical analysis of the probability of abnormal occurrence can be performed on each type of heterogeneous inducement data in the heterogeneous inducement data set. For each type of heterogeneous inducement data, the corresponding low-quality data in the historical product production data is associated with the historical production data abnormal inducement data set. The statistical analysis can calculate the probability of abnormal occurrence of each type of heterogeneous inducement data based on the associated data, such as using frequency statistics or probability distribution fitting methods.
S444：对异质诱因数据集中每一类异质诱因数据进行经济损失风险分布分析，得到经济损失风险分布数据；S444: performing economic loss risk distribution analysis on each type of heterogeneous inducement data in the heterogeneous inducement data set to obtain economic loss risk distribution data;
具体地，例如，可以收集与每一类异质诱因相关的经济损失数据，确保数据准确、完整且可靠。将经济损失数据与异质诱因数据集中的每一类数据进行关联，确保对应关系正确。针对异质诱因数据集中的每一类异质诱因数据，进行经济损失风险分布分析。使用适当的统计方法和风险评估模型，如概率分布分析、风险价值（VaR）计算等，来计算每一类异质诱因数据的经济损失风险分布。这些方法可以根据具体需求选择，如正态分布、蒙特卡洛模拟等。Specifically, for example, economic loss data related to each type of heterogeneous inducement can be collected to ensure that the data is accurate, complete and reliable. The economic loss data is associated with each type of data in the heterogeneous inducement data set to ensure that the corresponding relationship is correct. For each type of heterogeneous inducement data in the heterogeneous inducement data set, an economic loss risk distribution analysis is performed. Appropriate statistical methods and risk assessment models, such as probability distribution analysis, value at risk (VaR) calculation, etc., are used to calculate the economic loss risk distribution of each type of heterogeneous inducement data. These methods can be selected according to specific needs, such as normal distribution, Monte Carlo simulation, etc.
S445：对异质诱因数据集、各类异常诱因发生概率数据与经济损失风险分布数据进行数据异常诱因知识图谱构建，得到数据异常诱因知识图谱；S445: constructing a data anomaly inducement knowledge graph for the heterogeneous inducement data set, the probability data of various abnormal inducements, and the economic loss risk distribution data to obtain a data anomaly inducement knowledge graph;
具体地，例如，可以将异质诱因数据集中的每一类异质诱因数据作为节点，利用概率统计分析得到的异常发生概率数据作为节点属性。将经济损失风险分布数据与各类异质诱因数据进行关联，形成边，表示异常诱因与经济损失之间的关系。可以使用图数据库或知识图谱构建工具，如Neo4j、Apache Jena等，来构建数据异常诱因知识图谱。Specifically, for example, each type of heterogeneous inducement data in the heterogeneous inducement data set can be used as a node, and the abnormal occurrence probability data obtained by probability statistical analysis can be used as the node attribute. The economic loss risk distribution data is associated with each type of heterogeneous inducement data to form an edge, which represents the relationship between abnormal inducement and economic loss. A graph database or knowledge graph construction tool, such as Neo4j, Apache Jena, etc., can be used to construct a data abnormal inducement knowledge graph.
S446：将生产数据异常诱因数据与数据异常诱因知识图谱进行异常根源匹配，得到生产数据异常诱因匹配结果数据；S446: Match the production data anomaly cause data with the data anomaly cause knowledge graph to obtain the production data anomaly cause matching result data;
具体地，例如，可以将生产数据异常诱因数据与数据异常诱因知识图谱进行匹配。根据生产数据异常诱因数据的属性，与数据异常诱因知识图谱中的节点属性进行匹配，寻找与之相似或相关的节点。可以利用图数据库或知识图谱查询语言（如Cypher查询语言）进行查询和匹配操作。Specifically, for example, the production data anomaly inducement data can be matched with the data anomaly inducement knowledge graph. According to the attributes of the production data anomaly inducement data, the node attributes in the data anomaly inducement knowledge graph are matched to find similar or related nodes. The query and matching operations can be performed using a graph database or a knowledge graph query language (such as Cypher query language).
S447：基于数据异常诱因知识图谱与生产数据异常诱因匹配结果数据对目标生产车间进行数据质量提升优化决策，得到生产数据质量增强策略。S447: Based on the data anomaly cause knowledge graph and the production data anomaly cause matching result data, the data quality improvement optimization decision is made for the target production workshop to obtain the production data quality enhancement strategy.
具体地，例如，可以基于生产数据异常诱因匹配结果数据，分析目标生产车间的数据质量问题和异常情况。考虑异常发生概率、经济损失风险分布以及与异常诱因相关的节点属性等信息，对数据质量问题进行评估和分析。基于数据异常诱因知识图谱和数据质量分析结果，制定生产数据质量增强策略。根据知识图谱中异常诱因节点的属性和边的关系，确定数据质量改进的目标和方向。可以制定针对不同异常诱因的数据质量改进措施，如数据采集优化、传感器校准、工艺参数调整等。Specifically, for example, based on the matching result data of production data anomaly inducements, the data quality problems and abnormal situations of the target production workshop can be analyzed. Considering information such as the probability of anomaly occurrence, the distribution of economic loss risk, and the node attributes related to the anomaly inducements, the data quality problems are evaluated and analyzed. Based on the knowledge graph of data anomaly inducements and the results of data quality analysis, a production data quality enhancement strategy is formulated. According to the attributes of the anomaly inducement nodes in the knowledge graph and the relationship between the edges, the goals and directions of data quality improvement are determined. Data quality improvement measures can be formulated for different anomaly inducements, such as data acquisition optimization, sensor calibration, process parameter adjustment, etc.
本发明通过获取历史生产数据异常诱因数据集，提供了历史生产数据异常诱因的数据集，包括导致数据异常的各种因素。这为后续的异常诱因分析和优化决策提供了基础数据。通过对历史生产数据异常诱因进行识别与分类，将不同类型的异常诱因进行归类。这有助于企业清晰地了解不同类型的异常来源，并有针对性地进行异常根因分析和处理。生成了异质诱因数据集，其中包括设备故障、人工操作失误、供应链中断和环境影响等不同类型的异常诱因数据。这为后续的统计分析和风险评估提供了基础数据。通过统计分析，计算了每一类异质诱因数据的异常发生概率。这提供了企业评估不同异常诱因的风险程度的依据。帮助企业了解不同类型的异常诱因数据发生的频率，从而有助于制定相应的风险预防和数据质量改进策略。通过对异质诱因数据集中的每一类异常诱因数据进行经济损失风险分析，帮助企业评估不同异常诱因的经济损失风险程度。提供了经济损失风险分布数据，这有助于企业了解并优先处理那些可能导致较大经济损失的异常诱因。通过构建数据异常诱因知识图谱，将异质诱因数据、异常发生概率数据和经济损失风险分布数据有机地结合起来。数据异常诱因知识图谱提供了全面的异常诱因信息，帮助企业更好地理解不同异常诱因之间的关联和影响，为后续的根因匹配和数据质量提升决策提供了依据。将生产数据异常诱因数据与数据异常诱因知识图谱进行匹配，可以确定生产数据异常的根本原因。通过匹配结果数据，企业可以准确地了解导致数据异常的具体异常诱因，为解决数据质量问题提供指导和依据。基于数据异常诱因知识图谱和生产数据异常诱因匹配结果数据，制定针对目标生产车间的数据质量提升优化决策。通过制定生产数据质量增强策略，企业可以有效地改善数据质量，减少异常发生的可能性，提高产品质量和生产效率。本发明能够帮助企业深入了解异常诱因、评估风险、解决数据质量问题，并最终提升生产数据的质量和效益。The present invention provides a data set of abnormal inducements of historical production data, including various factors that cause data abnormality, by acquiring a data set of abnormal inducements of historical production data. This provides basic data for subsequent abnormal inducement analysis and optimization decision-making. By identifying and classifying the abnormal inducements of historical production data, different types of abnormal inducements are classified. This helps enterprises to clearly understand different types of abnormal sources and conduct abnormal root cause analysis and processing in a targeted manner. A heterogeneous inducement data set is generated, including different types of abnormal inducement data such as equipment failure, manual operation errors, supply chain interruptions and environmental impacts. This provides basic data for subsequent statistical analysis and risk assessment. Through statistical analysis, the abnormal occurrence probability of each type of heterogeneous inducement data is calculated. This provides a basis for enterprises to evaluate the risk level of different abnormal inducements. It helps enterprises understand the frequency of occurrence of different types of abnormal inducement data, thereby helping to formulate corresponding risk prevention and data quality improvement strategies. By performing economic loss risk analysis on each type of abnormal inducement data in the heterogeneous inducement data set, it helps enterprises to evaluate the economic loss risk level of different abnormal inducements. Economic loss risk distribution data is provided, which helps enterprises understand and prioritize those abnormal inducements that may cause greater economic losses. By constructing a knowledge graph of data anomaly inducements, heterogeneous inducement data, anomaly probability data and economic loss risk distribution data are organically combined. The knowledge graph of data anomaly inducements provides comprehensive anomaly inducement information, helping enterprises to better understand the association and influence between different anomaly inducements, and providing a basis for subsequent root cause matching and data quality improvement decisions. By matching the production data anomaly inducement data with the data anomaly inducement knowledge graph, the root cause of the production data anomaly can be determined. Through the matching result data, enterprises can accurately understand the specific anomaly inducements that cause data anomalies, providing guidance and basis for solving data quality problems. Based on the data anomaly inducement knowledge graph and the production data anomaly inducement matching result data, data quality improvement optimization decisions for the target production workshop are formulated. By formulating a production data quality enhancement strategy, enterprises can effectively improve data quality, reduce the possibility of anomalies, and improve product quality and production efficiency. The present invention can help enterprises gain an in-depth understanding of anomaly inducements, assess risks, solve data quality problems, and ultimately improve the quality and benefits of production data.
优选地，S447具体为：Preferably, S447 is specifically:
S4471：当生产数据异常诱因匹配结果数据为设备故障异质诱因数据时，对目标生产车间进行设备健康状态扫描，得到故障设备标识数据；对故障设备标识数据相应设备进行设备维修决策支持，得到车间设备优化维护策略；S4471: When the production data abnormal cause matching result data is the equipment failure heterogeneous cause data, the target production workshop is scanned for equipment health status to obtain faulty equipment identification data; equipment maintenance decision support is provided for the equipment corresponding to the faulty equipment identification data to obtain the workshop equipment optimization maintenance strategy;
具体地，例如，可以当生产数据异常诱因匹配结果数据为设备故障异质诱因数据时，目标生产车间中的设备进行健康状态扫描。可以使用传感器、监测设备或远程监控系统等技术手段，实时或定期获取设备的运行状态、故障信息等。根据设备健康状态扫描结果，识别出存在故障的设备，并生成故障设备标识数据。故障设备标识数据可以包括设备编号、设备类型、故障类型、故障程度等信息，用于后续的设备维修决策支持。基于故障设备标识数据，进行设备维修决策支持。可以利用故障设备的标识数据，结合设备维修记录、维修历史数据等，进行故障原因分析和维修优先级排序。根据维修优先级和资源可用性等因素，制定车间设备优化维护策略，如设备更换、维修计划调整等。Specifically, for example, when the production data abnormal inducement matching result data is the equipment failure heterogeneous inducement data, the equipment in the target production workshop can be scanned for health status. Technical means such as sensors, monitoring equipment or remote monitoring systems can be used to obtain the operating status, fault information, etc. of the equipment in real time or regularly. According to the equipment health status scan results, the faulty equipment is identified and the faulty equipment identification data is generated. The faulty equipment identification data may include information such as equipment number, equipment type, fault type, and fault degree, which is used for subsequent equipment maintenance decision support. Equipment maintenance decision support is performed based on the faulty equipment identification data. The identification data of the faulty equipment can be used in combination with equipment maintenance records, maintenance history data, etc. to analyze the cause of the failure and sort the maintenance priorities. According to factors such as maintenance priority and resource availability, an optimized maintenance strategy for workshop equipment is formulated, such as equipment replacement and maintenance plan adjustment.
S4472：当生产数据异常诱因匹配结果数据为人工操作失误异质诱因数据时，对目标生产车间中的关键工序进行人机交互分析，得到人机优化需求数据；基于人机优化需求数据实施虚拟现实交互式培训，得到车间人员技能提升策略；S4472: When the production data abnormal cause matching result data is heterogeneous cause data of manual operation errors, human-machine interaction analysis is performed on the key processes in the target production workshop to obtain human-machine optimization demand data; virtual reality interactive training is implemented based on the human-machine optimization demand data to obtain workshop personnel skill improvement strategies;
具体地，例如，可以当生产数据异常诱因匹配结果数据为人工操作失误异质诱因数据时，根据生产数据异常诱因匹配结果数据中的人工操作失误异质诱因数据，重点关注目标生产车间中的关键工序。进行人机交互分析，评估人机界面设计、工序操作流程、工具设备等方面可能导致的人工操作失误问题。基于人机交互分析结果，得到人机优化需求数据。人机优化需求数据可以包括对界面设计的改进要求、对工序流程的调整建议、对培训和技能提升的需求等。基于人机优化需求数据，实施虚拟现实交互式培训。利用虚拟现实技术，构建模拟的生产工作环境，让车间人员通过交互式培训进行实践操作和技能提升。培训内容可以包括正确的操作流程、应对异常情况的处理方法等，以提升车间人员的操作技能和减少人工操作失误。根据虚拟现实交互式培训的结果和反馈，制定车间人员技能提升策略。可以根据培训的效果评估，确定进一步的培训计划和措施，如定期培训、技能认证等，以确保车间人员的技能水平持续提升。Specifically, for example, when the production data abnormal inducement matching result data is manual operation error heterogeneous inducement data, the key processes in the target production workshop can be focused on according to the manual operation error heterogeneous inducement data in the production data abnormal inducement matching result data. Human-computer interaction analysis is performed to evaluate the human-computer interface design, process operation flow, tools and equipment and other aspects that may cause manual operation errors. Based on the results of the human-computer interaction analysis, human-computer optimization demand data is obtained. The human-computer optimization demand data may include requirements for improvement of interface design, suggestions for adjustment of process flow, requirements for training and skill improvement, etc. Based on the human-computer optimization demand data, virtual reality interactive training is implemented. Using virtual reality technology, a simulated production work environment is constructed to allow workshop personnel to perform practical operations and improve their skills through interactive training. The training content may include correct operating procedures, methods for dealing with abnormal situations, etc., to improve the operating skills of workshop personnel and reduce manual operation errors. According to the results and feedback of virtual reality interactive training, a skill improvement strategy for workshop personnel is formulated. Further training plans and measures, such as regular training and skill certification, can be determined based on the evaluation of the effectiveness of the training to ensure that the skill level of workshop personnel continues to improve.
S4473：当生产数据异常诱因匹配结果数据为供应链中断异质诱因数据时，构建供应链数字孪生模型，并获取车间原料物流数据；利用供应链数字孪生模型结合车间原料物流数据对目标生产车间进行智能物流调度优化实施，得到车间供应链稳定策略；S4473: When the matching result data of the abnormal inducement of production data is the heterogeneous inducement data of supply chain interruption, a supply chain digital twin model is constructed and the raw material logistics data of the workshop is obtained; the supply chain digital twin model is used in combination with the raw material logistics data of the workshop to optimize the intelligent logistics scheduling of the target production workshop and obtain the workshop supply chain stabilization strategy;
具体地，例如，当生产数据异常诱因匹配结果数据为供应链中断异质诱因数据时，基于供应链的结构和运作流程，建立供应链数字孪生模型。数字孪生模型是一个虚拟的供应链镜像，能够模拟供应链中的各个环节和相互关系。收集目标生产车间的原料物流数据。这些数据可以包括原料的供应商信息、运输时间、库存水平等。将车间原料物流数据与供应链数字孪生模型结合起来。使用智能算法和优化技术，对供应链进行调度和优化，以实现更稳定的物流运作和减少中断的风险。优化的目标可以包括减少供应链中断次数、降低物流成本、提高物流效率等。根据智能物流调度优化的结果，制定车间供应链稳定策略。这些策略可以包括与供应商的合作协议调整、库存管理策略优化、物流运输方式调整等，旨在确保供应链的稳定性和可靠性。Specifically, for example, when the matching result data of the abnormal inducement of production data is the heterogeneous inducement data of the supply chain interruption, a supply chain digital twin model is established based on the structure and operation process of the supply chain. The digital twin model is a virtual supply chain mirror that can simulate the various links and relationships in the supply chain. Collect the raw material logistics data of the target production workshop. These data may include supplier information, transportation time, inventory level, etc. of the raw materials. Combine the workshop raw material logistics data with the supply chain digital twin model. Use intelligent algorithms and optimization techniques to schedule and optimize the supply chain to achieve more stable logistics operations and reduce the risk of interruptions. The optimization goals may include reducing the number of supply chain interruptions, reducing logistics costs, and improving logistics efficiency. According to the results of intelligent logistics scheduling optimization, formulate workshop supply chain stabilization strategies. These strategies may include adjustments to cooperation agreements with suppliers, optimization of inventory management strategies, adjustments to logistics transportation methods, etc., aiming to ensure the stability and reliability of the supply chain.
S4474：当生产数据异常诱因匹配结果数据为环境影响异质诱因数据时，对目标生产车间进行车间环境参数感知网络部署，得到车间实时环境参数；根据车间实时环境参数对目标生产车间进行智能环境控制优化实施，得到车间环境优化策略；S4474: When the production data abnormal inducement matching result data is environmental impact heterogeneous inducement data, a workshop environment parameter perception network is deployed for the target production workshop to obtain the real-time environment parameters of the workshop; intelligent environment control optimization is implemented for the target production workshop according to the real-time environment parameters of the workshop to obtain the workshop environment optimization strategy;
具体地，例如，当生产数据异常诱因匹配结果数据为环境影响异质诱因数据时，在目标生产车间中部署环境参数感知网络。这可以包括传感器、监测设备和数据采集系统等，用于实时感知车间的环境参数，如温度、湿度、空气质量等。通过部署的环境参数感知网络，获取目标生产车间的实时环境参数数据。结合车间实时环境参数数据，使用智能算法和控制技术，对车间环境进行优化和调控。例如，根据温度和湿度数据进行空调和通风系统的自动调节，以保持适宜的工作环境。根据智能环境控制优化的结果，制定车间环境优化策略。这些策略可以包括设定环境参数的目标范围、优化设备调度和控制策略等，旨在提供舒适、安全和高效的工作环境。Specifically, for example, when the production data abnormal inducement matching result data is environmental impact heterogeneous inducement data, an environmental parameter perception network is deployed in the target production workshop. This may include sensors, monitoring equipment, and data acquisition systems, etc., for real-time perception of the environmental parameters of the workshop, such as temperature, humidity, air quality, etc. The real-time environmental parameter data of the target production workshop is obtained through the deployed environmental parameter perception network. Combined with the real-time environmental parameter data of the workshop, the workshop environment is optimized and regulated using intelligent algorithms and control technologies. For example, the air conditioning and ventilation systems are automatically adjusted according to the temperature and humidity data to maintain a suitable working environment. According to the results of the intelligent environmental control optimization, workshop environment optimization strategies are formulated. These strategies may include setting the target range of environmental parameters, optimizing equipment scheduling and control strategies, etc., aiming to provide a comfortable, safe and efficient working environment.
S4475：将车间设备优化维护策略或车间人员技能提升策略或车间供应链稳定策略或车间环境优化策略作为生产数据质量增强策略；S4475: Use workshop equipment optimization and maintenance strategy, workshop personnel skill improvement strategy, workshop supply chain stabilization strategy, or workshop environment optimization strategy as production data quality enhancement strategy;
具体地，例如，当生产数据异常诱因匹配结果数据为设备故障异质诱因数据、人工操作失误异质诱因数据、供应链中断异质诱因数据以及环境影响异质诱因数据中某一中异质诱因数据时，将异质诱因数据对应的质量增强策略作为生产数据质量增强策略。Specifically, for example, when the production data abnormal cause matching result data is one of the heterogeneous cause data including equipment failure heterogeneous cause data, manual operation error heterogeneous cause data, supply chain interruption heterogeneous cause data and environmental impact heterogeneous cause data, the quality enhancement strategy corresponding to the heterogeneous cause data is used as the production data quality enhancement strategy.
S4476：当生产数据异常诱因匹配结果数据为设备故障异质诱因数据、人工操作失误异质诱因数据、供应链中断异质诱因数据以及环境影响异质诱因数据中的任一组合时，将车间设备优化维护策略、车间人员技能提升策略、车间供应链稳定策略与车间环境优化策略中对应的组合作为生产数据质量增强策略。S4476: When the matching result data of the abnormal inducement of production data is any combination of equipment failure heterogeneous inducement data, manual operation error heterogeneous inducement data, supply chain interruption heterogeneous inducement data and environmental impact heterogeneous inducement data, the corresponding combination of workshop equipment optimization and maintenance strategy, workshop personnel skill improvement strategy, workshop supply chain stabilization strategy and workshop environment optimization strategy shall be used as the production data quality enhancement strategy.
具体地，例如，当生产数据异常诱因匹配结果数据为设备故障异质诱因数据、人工操作失误异质诱因数据、供应链中断异质诱因数据以及环境影响异质诱因数据中的任一组合时，将车间设备优化维护策略、车间人员技能提升策略、车间供应链稳定策略与车间环境优化策略中对应的组合作为生产数据质量增强策略。Specifically, for example, when the production data abnormal cause matching result data is any combination of equipment failure heterogeneous cause data, manual operation error heterogeneous cause data, supply chain interruption heterogeneous cause data, and environmental impact heterogeneous cause data, the corresponding combination of workshop equipment optimization and maintenance strategy, workshop personnel skills improvement strategy, workshop supply chain stabilization strategy and workshop environment optimization strategy is used as the production data quality enhancement strategy.
本发明中，当生产数据异常诱因匹配结果数据为设备故障异质诱因数据时，通过设备健康状态扫描，及时发现目标生产车间中存在的故障设备，准确标识出故障设备。这有助于企业及时采取维修措施，防止设备故障对生产造成更大的影响。基于故障设备标识数据，为设备维修决策提供支持。这使企业能够根据设备故障的严重程度、影响范围等因素，制定相应的设备优化维护策略。有效的设备维护可以减少生产中断和工业生产数据连贯性质量问题，提高生产效率和稳定性。当生产数据异常诱因匹配结果数据为人工操作失误异质诱因数据时，通过人机交互分析，深入了解目标生产车间关键工序中可能存在的人工操作失误问题。这有助于识别潜在的问题点，改进工序设计和操作流程，减少人工操作失误的发生。基于人机优化需求数据，实施虚拟现实交互式培训。这种培训方式能够模拟真实工作环境，提供身临其境的体验，帮助车间人员提升技能和操作水平，从而确保生产过程中人工手动输入数据的质量。当生产数据异常诱因匹配结果数据为供应链中断异质诱因数据时，构建供应链数字孪生模型并获取车间原料物流数据，可以实现对供应链中断异质诱因的全面分析和建模。这使企业能够更好地理解供应链中的问题和瓶颈，并提前预测可能的中断情况，为供应链管理提供更准确的数据支持。利用供应链数字孪生模型结合车间原料物流数据，对目标生产车间进行智能物流调度优化实施。通过优化物流调度，可以实现原料的准时供应、减少库存积压、优化运输路线等。这有助于降低供应链中断的风险，提高生产车间的供应链稳定性和运作效率。当生产数据异常诱因匹配结果数据为环境影响异质诱因数据时，通过部署车间环境参数感知网络，能够实时感知目标生产车间的环境参数，如温度、湿度、气体浓度等。这有助于对环境影响因素进行准确监测和数据收集。基于实时环境参数，通过智能环境控制优化实施，对目标生产车间进行环境调节和优化。通过调整环境参数，如温度、湿度等，以满足生产过程中的要求，提高产品质量、稳定生产流程，从而提高产品质量指标。将车间设备优化维护策略作为生产数据质量增强策略，可以确保设备的正常运行和提高设备可靠性，减少设备故障对生产数据质量的影响。将车间人员技能提升策略作为生产数据质量增强策略，可以提高员工的操作水平和技能，减少人为因素导致的生产数据异常和质量问题。将车间供应链稳定策略作为生产数据质量增强策略，可以保证供应链的稳定性，减少供应链中断对生产数据的影响。将车间环境优化策略作为生产数据质量增强策略，可以优化生产环境，减少环境因素对生产数据和产品质量的影响。In the present invention, when the production data abnormal inducement matching result data is the equipment failure heterogeneous inducement data, the equipment health status scan is used to timely discover the faulty equipment in the target production workshop and accurately identify the faulty equipment. This helps enterprises to take maintenance measures in time to prevent equipment failure from having a greater impact on production. Based on the faulty equipment identification data, support is provided for equipment maintenance decisions. This enables enterprises to formulate corresponding equipment optimization and maintenance strategies based on factors such as the severity of equipment failure and the scope of impact. Effective equipment maintenance can reduce production interruptions and industrial production data consistency quality problems, and improve production efficiency and stability. When the production data abnormal inducement matching result data is the manual operation error heterogeneous inducement data, through human-computer interaction analysis, an in-depth understanding of the possible manual operation error problems in the key processes of the target production workshop is obtained. This helps to identify potential problem points, improve process design and operation procedures, and reduce the occurrence of manual operation errors. Based on the human-computer optimization demand data, virtual reality interactive training is implemented. This training method can simulate a real working environment, provide an immersive experience, and help workshop personnel improve their skills and operation levels, thereby ensuring the quality of manually input data during the production process. When the result data of the abnormal inducement matching of production data is heterogeneous inducement data of supply chain disruption, building a supply chain digital twin model and obtaining workshop raw material logistics data can realize comprehensive analysis and modeling of heterogeneous inducements of supply chain disruption. This enables enterprises to better understand the problems and bottlenecks in the supply chain and predict possible disruptions in advance, providing more accurate data support for supply chain management. The supply chain digital twin model is combined with the workshop raw material logistics data to optimize the intelligent logistics scheduling of the target production workshop. By optimizing logistics scheduling, it is possible to achieve on-time supply of raw materials, reduce inventory backlogs, optimize transportation routes, etc. This helps to reduce the risk of supply chain disruption and improve the supply chain stability and operation efficiency of the production workshop. When the result data of the abnormal inducement matching of production data is heterogeneous inducement data of environmental impact, by deploying the workshop environmental parameter perception network, the environmental parameters of the target production workshop, such as temperature, humidity, gas concentration, etc., can be perceived in real time. This helps to accurately monitor and collect data on environmental influencing factors. Based on real-time environmental parameters, the target production workshop is adjusted and optimized through intelligent environmental control optimization implementation. By adjusting environmental parameters such as temperature and humidity to meet the requirements of the production process, improve product quality, stabilize the production process, and thus improve product quality indicators. Using workshop equipment optimization and maintenance strategies as production data quality enhancement strategies can ensure the normal operation of equipment and improve equipment reliability, reducing the impact of equipment failures on production data quality. Using workshop personnel skill improvement strategies as production data quality enhancement strategies can improve employees' operating levels and skills, and reduce production data anomalies and quality problems caused by human factors. Using workshop supply chain stabilization strategies as production data quality enhancement strategies can ensure the stability of the supply chain and reduce the impact of supply chain disruptions on production data. Using workshop environment optimization strategies as production data quality enhancement strategies can optimize the production environment and reduce the impact of environmental factors on production data and product quality.
优选地，生产过程知识图谱构建模块S5具体为：Preferably, the production process knowledge graph construction module S5 is specifically:
S51：对产品生产数据品质指数与生产数据异常诱因数据进行分形维度分析，得到数据质量异常分形关联数据；S51: Perform fractal dimension analysis on the product production data quality index and the production data abnormality inducement data to obtain data quality abnormality fractal correlation data;
具体地，例如，可以使用分形维度分析算法如Hurst指数、分形维度计算等。将产品生产数据品质指数和生产数据异常诱因数据输入选择的分形维度分析工具。运行分析工具，计算数据的分形维度。这将揭示数据的自相似特征和分形结构。综合产品生产数据品质指数和生产数据异常诱因数据的分形维度分析结果。提取数据质量异常分形关联数据，即异常数据在分形维度上的相关性信息。Specifically, for example, fractal dimension analysis algorithms such as Hurst index, fractal dimension calculation, etc. can be used. Input the product production data quality index and the production data abnormality inducement data into the selected fractal dimension analysis tool. Run the analysis tool to calculate the fractal dimension of the data. This will reveal the self-similar characteristics and fractal structure of the data. Combine the fractal dimension analysis results of the product production data quality index and the production data abnormality inducement data. Extract the data quality abnormal fractal correlation data, that is, the correlation information of the abnormal data on the fractal dimension.
S52：对生产数据异常诱因数据与生产数据质量增强策略进行过程语义分析，得到异常诱因策略语义数据；S52: Perform process semantic analysis on the production data abnormality inducement data and the production data quality enhancement strategy to obtain abnormality inducement strategy semantic data;
具体地，例如，可以使用自然语言处理（NLP）技术、语义分析算法等。将生产数据异常诱因数据和生产数据质量增强策略数据输入选择的过程语义分析工具。运行分析工具，提取数据中的语义信息和关联关系。这将揭示异常诱因数据和策略数据之间的语义关联。综合异常诱因与策略数据的语义分析结果。提取异常诱因策略语义数据，即异常诱因数据与策略数据之间的语义关系和语义表示。对异常诱因策略语义数据进行解释和分析。根据分析结果，可以理解异常诱因与策略之间的语义关系和逻辑。Specifically, for example, natural language processing (NLP) technology, semantic analysis algorithms, etc. can be used. Input the production data anomaly inducement data and the production data quality enhancement strategy data into the selected process semantic analysis tool. Run the analysis tool to extract the semantic information and association relationship in the data. This will reveal the semantic association between the anomaly inducement data and the strategy data. Combine the semantic analysis results of the anomaly inducement and strategy data. Extract the anomaly inducement strategy semantic data, that is, the semantic relationship and semantic representation between the anomaly inducement data and the strategy data. Interpret and analyze the anomaly inducement strategy semantic data. Based on the analysis results, the semantic relationship and logic between the anomaly inducement and the strategy can be understood.
S53：基于数据质量异常分形关联数据与异常诱因策略语义数据对产品生产数据品质指数、生产数据异常诱因数据与生产数据质量增强策略进行异构数据图嵌入融合，得到初始数据品质因果知识图谱；S53: Based on the data quality anomaly fractal association data and anomaly inducement strategy semantic data, the product production data quality index, production data anomaly inducement data and production data quality enhancement strategy are embedded and fused into heterogeneous data graphs to obtain the initial data quality causal knowledge graph;
具体地，例如，可以选择适合进行异构数据图嵌入的工具或方法。异构数据图嵌入工具可以将不同类型的数据和关联关系嵌入到统一的图结构中。将数据质量异常分形关联数据、异常诱因策略语义数据、产品生产数据品质指数、生产数据异常诱因数据和生产数据质量增强策略以异构数据图的方式表示。在图中，节点表示数据和策略元素，边表示它们之间的关联关系。将异构数据图输入选择的异构数据图嵌入工具。运行嵌入工具，例如，GraphSAGE、DeepWalk等，将不同类型的数据嵌入到统一的图结构中，保留数据之间的关联关系。这将生成一个嵌入后的初始数据品质因果知识图谱。Specifically, for example, a tool or method suitable for heterogeneous data graph embedding can be selected. The heterogeneous data graph embedding tool can embed different types of data and association relationships into a unified graph structure. Data quality anomaly fractal association data, anomaly inducement strategy semantic data, product production data quality index, production data anomaly inducement data, and production data quality enhancement strategy are represented in the form of a heterogeneous data graph. In the graph, nodes represent data and policy elements, and edges represent the association relationships between them. Input the heterogeneous data graph into the selected heterogeneous data graph embedding tool. Run the embedding tool, such as GraphSAGE, DeepWalk, etc., to embed different types of data into a unified graph structure and retain the association relationships between the data. This will generate an embedded initial data quality causal knowledge graph.
S54：对目标生产车间进行物联网实时监测，得到实时生产状态监控数据集；基于实时生产状态监控数据集对初始数据品质因果知识图谱进行时序知识图谱演化，得到动态数据品质因果知识图谱。S54: Conduct real-time IoT monitoring of the target production workshop to obtain a real-time production status monitoring data set; perform temporal knowledge graph evolution on the initial data quality causal knowledge graph based on the real-time production status monitoring data set to obtain a dynamic data quality causal knowledge graph.
具体地，例如，可以在目标生产车间中部署物联网监测系统。系统可以收集各种传感器和设备的实时数据，例如温度、湿度、压力、设备状态等。使用物联网监测系统收集目标生产车间中的实时生产状态监控数据。这些数据可以是实时传感器读数、设备运行状态、工作流程信息等。选择适合进行时序知识图谱演化的算法或方法。例如，Temporal GraphConvolutional Networks(TGCN)：TGCN是一种基于图卷积网络的算法，用于处理时序图数据。它可以捕捉节点和边在不同时间步之间的关系演化。时序知识图谱演化算法可以将时间序列数据嵌入到知识图谱中，反映数据的时序变化和演化过程。将预处理的实时生产状态监控数据集与初始数据品质因果知识图谱输入选择的时序知识图谱演化算法。运行演化算法，将实时数据的时序变化和演化过程嵌入到初始数据品质因果知识图谱中。这将生成一个动态数据品质因果知识图谱，其中节点和边表示数据和策略元素在时间上的变化和演化。Specifically, for example, an IoT monitoring system can be deployed in the target production workshop. The system can collect real-time data from various sensors and equipment, such as temperature, humidity, pressure, equipment status, etc. Use the IoT monitoring system to collect real-time production status monitoring data in the target production workshop. These data can be real-time sensor readings, equipment operating status, workflow information, etc. Select an algorithm or method suitable for the evolution of the temporal knowledge graph. For example, Temporal Graph Convolutional Networks (TGCN): TGCN is an algorithm based on graph convolutional networks for processing temporal graph data. It can capture the evolution of the relationship between nodes and edges at different time steps. The temporal knowledge graph evolution algorithm can embed time series data into the knowledge graph to reflect the temporal changes and evolution process of the data. The preprocessed real-time production status monitoring data set and the initial data quality causal knowledge graph are input into the selected temporal knowledge graph evolution algorithm. Run the evolution algorithm to embed the temporal changes and evolution process of the real-time data into the initial data quality causal knowledge graph. This will generate a dynamic data quality causal knowledge graph in which nodes and edges represent the changes and evolution of data and policy elements over time.
本发明通过分形维度分析，可以揭示产品生产数据品质指数和生产数据异常诱因数据之间的关联关系。分形维度分析是一种用于研究数据的自相似性和复杂性的方法，能够提取数据的隐藏模式和特征。数据质量异常分形关联数据提供了关于数据质量异常的定量度量和可视化表示，有助于企业进行数据质量监控、异常检测和问题排查，提高数据质量管理效果。通过过程语义分析，可以提取出生产数据异常诱因数据和生产数据质量增强策略之间的关键语义信息，帮助企业理解和解释数据异常的原因和改进策略。异常诱因策略语义数据可以作为知识图谱中的重要元素，为知识表示和推理提供语义上的丰富性和准确性，提高知识图谱的应用效果和决策支持能力。通过初始数据品质因果知识图谱将不同类型的数据和信息整合在一起，提供了一个全面的视角来理解和分析产品生产数据的品质因果关系。它可以帮助企业发现数据质量异常的根本原因，识别潜在的改进策略，并为数据质量管理和决策提供指导。通过异构数据图嵌入融合，知识图谱中的数据和信息相互关联，形成更加丰富和综合的知识表示。这可以提高企业对数据质量问题的感知和理解能力，加强数据驱动的决策支持和问题解决能力。实时生产状态监控数据集提供了对目标生产车间的实时了解，帮助企业及时掌握生产过程中的变化和异常情况，支持实时监控和决策调整。动态数据品质因果知识图谱反映了数据品质因果关系的时序演化，能够帮助企业发现数据品质变化的规律和趋势，预测潜在的数据质量问题，并及时采取相应的措施进行调整和改进。The present invention can reveal the correlation between the product production data quality index and the production data abnormality inducement data through fractal dimension analysis. Fractal dimension analysis is a method for studying the self-similarity and complexity of data, which can extract the hidden patterns and features of data. The data quality abnormal fractal correlation data provides quantitative measurement and visual representation of data quality abnormality, which helps enterprises to monitor data quality, detect abnormalities and troubleshoot problems, and improve the data quality management effect. Through process semantic analysis, the key semantic information between the production data abnormality inducement data and the production data quality enhancement strategy can be extracted to help enterprises understand and explain the causes and improvement strategies of data abnormalities. The abnormal inducement strategy semantic data can be used as an important element in the knowledge graph to provide semantic richness and accuracy for knowledge representation and reasoning, and improve the application effect and decision support ability of the knowledge graph. Different types of data and information are integrated together through the initial data quality causal knowledge graph, providing a comprehensive perspective to understand and analyze the quality causal relationship of product production data. It can help enterprises discover the root causes of data quality abnormalities, identify potential improvement strategies, and provide guidance for data quality management and decision-making. Through the embedding and fusion of heterogeneous data graphs, the data and information in the knowledge graph are interconnected to form a richer and more comprehensive knowledge representation. This can improve the company's perception and understanding of data quality issues, and strengthen data-driven decision support and problem-solving capabilities. The real-time production status monitoring data set provides real-time understanding of the target production workshop, helping companies to promptly grasp changes and abnormalities in the production process, and support real-time monitoring and decision adjustments. The dynamic data quality causal knowledge graph reflects the temporal evolution of data quality causal relationships, which can help companies discover the laws and trends of data quality changes, predict potential data quality issues, and take corresponding measures to adjust and improve them in a timely manner.
因此，无论从哪一点来看，均应将实施例看作是示范性的，而且是非限制性的，本发明的范围由所附权利要求而不是上述说明限定，因此旨在将落在申请文件的等同要件的含义和范围内的所有变化涵括在本发明内。Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.
以上所述仅是本发明的具体实施方式，使本领域技术人员能够理解或实现本发明。对这些实施例的多种修改对本领域的技术人员来说将是显而易见的，本文中所定义的一般原理可以在不脱离本发明的精神或范围的情况下，在其它实施例中实现。因此，本发明将不会被限制于本文所示的这些实施例，而是要符合与本文所发明的原理和新颖特点相一致的最宽的范围。The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410424860.1ACN118011990B (en) | 2024-04-10 | 2024-04-10 | Industrial data quality monitoring and improving system based on artificial intelligence |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410424860.1ACN118011990B (en) | 2024-04-10 | 2024-04-10 | Industrial data quality monitoring and improving system based on artificial intelligence |
| Publication Number | Publication Date |
|---|---|
| CN118011990A CN118011990A (en) | 2024-05-10 |
| CN118011990Btrue CN118011990B (en) | 2024-06-14 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410424860.1AActiveCN118011990B (en) | 2024-04-10 | 2024-04-10 | Industrial data quality monitoring and improving system based on artificial intelligence |
| Country | Link |
|---|---|
| CN (1) | CN118011990B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118331212B (en)* | 2024-06-13 | 2024-09-06 | 江苏天弘颢宇信息技术有限公司 | Industrial electrical machine room environment monitoring system and method with multi-source acquisition and data modeling |
| CN118551230B (en)* | 2024-07-30 | 2024-10-18 | 厦门辉腾盛节能材料有限公司 | Operation monitoring method of high-strength protective coating automatic production line |
| CN118656762B (en)* | 2024-08-19 | 2024-11-15 | 广东鑫光智能系统有限公司 | Intelligent plate processing workshop fault monitoring method and system |
| CN119130372A (en)* | 2024-09-07 | 2024-12-13 | 莆田市顺莲电子商务有限公司 | Sales chain management method based on Internet identity resolution |
| CN119228397A (en)* | 2024-09-11 | 2024-12-31 | 四川省农业科学院农业资源与环境研究所 | Organic fertilizer quality and safety traceability management method based on multi-source data |
| CN119670866A (en)* | 2024-12-11 | 2025-03-21 | 中国人民解放军92942部队 | A quality problem information analysis method based on knowledge graph |
| CN119479952A (en)* | 2025-01-17 | 2025-02-18 | 上海交通大学 | A quality rating and recommendation method for magnesium-based alloys |
| CN119556665B (en)* | 2025-02-05 | 2025-07-04 | 晶芯成(北京)科技有限公司 | Product quality monitoring method and system for semiconductor manufacturing factory |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110275898A (en)* | 2018-03-16 | 2019-09-24 | 埃森哲环球解决方案有限公司 | Use the integrated monitoring and communication system of the explanatory equipment management of knowledge based figure |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP7556239B2 (en)* | 2020-09-04 | 2024-09-26 | 株式会社プロテリアル | Computer program, defect cause analysis method, and defect cause analysis device |
| CN112488489B (en)* | 2020-11-26 | 2024-02-27 | 江苏科技大学 | Digital twin-driven processing quality tracing and dynamic control method |
| CN115186054A (en)* | 2022-06-29 | 2022-10-14 | 国网江苏省电力有限公司灌云县供电分公司 | A kind of fault visualization analysis method and system |
| CN115293507A (en)* | 2022-07-04 | 2022-11-04 | 中国电力科学研究院有限公司 | Situation knowledge graph construction method and system for power grid dispatching operation management |
| KR102647438B1 (en)* | 2022-08-10 | 2024-03-12 | 에스케이가스 주식회사 | Method for analysis of root cause of chemical processes and method for modeling of yield prediction |
| KR20240025792A (en)* | 2022-08-19 | 2024-02-27 | (주)더블유앤이케이 | Method and system for detecting the quality abnormality based on the artificial intelligence, and controlling facility |
| KR102622569B1 (en)* | 2022-12-20 | 2024-01-09 | (주)위세아이텍 | Intelligent apparatus for controlling manufacturing facility and method for controlling the same |
| CN117406689A (en)* | 2023-10-13 | 2024-01-16 | 中国电子科技集团公司第十四研究所 | Data driving and knowledge guiding fault diagnosis method and system |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110275898A (en)* | 2018-03-16 | 2019-09-24 | 埃森哲环球解决方案有限公司 | Use the integrated monitoring and communication system of the explanatory equipment management of knowledge based figure |
| Title |
|---|
| 基于数据分析的业务风险防控研究;邢巍 等;;《现代商业》;20200328(第09期);第17-19页* |
| Publication number | Publication date |
|---|---|
| CN118011990A (en) | 2024-05-10 |
| Publication | Publication Date | Title |
|---|---|---|
| CN118011990B (en) | Industrial data quality monitoring and improving system based on artificial intelligence | |
| US10902368B2 (en) | Intelligent decision synchronization in real time for both discrete and continuous process industries | |
| CN115310952B (en) | A method and system for enterprise asset management based on Internet of Things | |
| CN117055502A (en) | Intelligent control system based on Internet of things and big data analysis | |
| Gorski et al. | Towards a smart workflow in CMMS/EAM systems: An approach based on ML and MCDM | |
| CN118014165B (en) | Traceability management method and traceability management system for lithium ion battery production | |
| CN118378196B (en) | Industrial control host abnormal behavior identification method based on multi-mode data fusion | |
| CN113468022B (en) | Automatic operation and maintenance method for centralized monitoring of products | |
| Fernandes et al. | Machine learning and process mining applied to process optimization: Bibliometric and systemic analysis | |
| CN118114023A (en) | Pharmaceutical production line state monitoring and alarming method and system based on Internet of things | |
| CN119228397A (en) | Organic fertilizer quality and safety traceability management method based on multi-source data | |
| CN119676055A (en) | An operation and maintenance management system and method based on artificial intelligence | |
| CN119322991A (en) | Data interaction method based on intelligent control platform | |
| CN119724530B (en) | A hospital equipment budget management method, system and medium based on holographic view | |
| CN120338430A (en) | A MES digital collaborative management method and system based on deep learning | |
| CN117828148A (en) | Land sample data management method and system for land management | |
| CN118569737B (en) | A method and system for tracing the quality and safety of lens assembly production | |
| Donhauser et al. | Rolling-reactive optimization of production processes in a calcium silicate masonry unit plant using online simulation | |
| CN119273296A (en) | Safety inspection method and system applied to project management | |
| CN119048113A (en) | Food supply chain tracing method based on multi-source data | |
| CN118396606A (en) | Manufacturing industry product production optimization method based on multi-mode large model | |
| Li et al. | Challenges in developing a computational platform to integrate data analytics with simulation-based optimization | |
| Cai et al. | Active monitoring of production status in discrete manufacturing workshops driven by digital twins | |
| CN118400396B (en) | Industrial Internet security monitoring method and system based on big data | |
| CN119250792A (en) | A full life cycle management method and system for an emergency command system information support system |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |