技术领域Technical Field
本发明涉及运输预警技术领域，尤其涉及一种基于行驶轨迹的大件运输监管及预警方法、系统及装置。The present invention relates to the field of transportation early warning technology, and in particular to a large-scale transportation supervision and early warning method, system and device based on driving trajectory.
背景技术Background technique
大件货物的运输主要依赖于人工经验和简单的物流管理系统，缺乏科学化、智能化的监管手段。随着信息技术的不断发展，大数据分析和物联网技术逐渐应用于物流领域，为大件运输监管提供了新的思路和技术支持。在大数据分析方面，各种传感器技术的发展为运输车辆提供了丰富的数据采集手段，如GPS定位、车辆行驶状态监测等。这些数据通过云计算和大数据分析技术进行处理和挖掘，可以实现对大件运输车辆行驶轨迹、运输速度、路况等信息的实时监测和分析，为监管部门提供了更加全面、准确的数据支持。基于人工智能的预警系统也逐渐成熟。通过对历史数据的分析和机器学习算法的训练，可以构建出针对大件运输安全的预警模型，及时发现运输过程中存在的安全隐患和异常情况，并采取相应的预防和应对措施，提高了大件运输的安全性和效率。然而，目前传统的技术通常缺乏对异常轨迹的有效检测方法，导致在运输过程中无法及时发现异常情况，同时大件运输过程中的异常行为受到环境因素的影响，如交通拥堵、天气变化等，但现有技术往往缺乏对环境感知数据的充分利用，导致异常行驶的监管预警的准确性和效率较差。The transportation of large-scale goods mainly relies on manual experience and simple logistics management systems, lacking scientific and intelligent supervision methods. With the continuous development of information technology, big data analysis and Internet of Things technology are gradually applied to the field of logistics, providing new ideas and technical support for the supervision of large-scale transportation. In terms of big data analysis, the development of various sensor technologies has provided a wealth of data collection methods for transportation vehicles, such as GPS positioning and vehicle driving status monitoring. These data are processed and mined through cloud computing and big data analysis technology, which can realize real-time monitoring and analysis of the driving trajectory, transportation speed, road conditions and other information of large-scale transportation vehicles, providing more comprehensive and accurate data support for regulatory authorities. Early warning systems based on artificial intelligence are also gradually maturing. Through the analysis of historical data and the training of machine learning algorithms, an early warning model for large-scale transportation safety can be constructed to timely discover safety hazards and abnormal situations in the transportation process, and take corresponding preventive and response measures, thereby improving the safety and efficiency of large-scale transportation. However, current traditional technologies usually lack effective detection methods for abnormal trajectories, resulting in the inability to detect abnormal situations in a timely manner during transportation. At the same time, abnormal behaviors during large-scale transportation are affected by environmental factors, such as traffic congestion, weather changes, etc., but existing technologies often lack full utilization of environmental perception data, resulting in poor accuracy and efficiency of regulatory warnings for abnormal driving.
发明内容Summary of the invention
基于此，有必要提供一种基于行驶轨迹的大件运输监管及预警方法、系统及装置，以解决至少一个上述技术问题。Based on this, it is necessary to provide a large-scale transportation supervision and early warning method, system and device based on driving trajectory to solve at least one of the above technical problems.
为实现上述目的，一种基于行驶轨迹的大件运输监管及预警方法，所述方法包括以下步骤：To achieve the above purpose, a large-scale transportation supervision and early warning method based on driving trajectory is provided, and the method comprises the following steps:
步骤S1：获取大件运输车辆实时轨迹数据；对大件运输车辆实时轨迹数据进行数据预处理，生成标准大件运输车辆轨迹数据；Step S1: Acquire the real-time trajectory data of the large-scale transport vehicle; perform data preprocessing on the real-time trajectory data of the large-scale transport vehicle to generate standard large-scale transport vehicle trajectory data;
步骤S2：对标准大件运输车辆轨迹数据进行运输行驶轨迹分段，得到大件运输轨迹行为片段；对大件运输轨迹行为片段进行轨迹偏差计算，得到轨迹偏差数据；将行驶轨迹偏差数据和预设的行驶轨迹偏差程度阈值进行对比，生成异常轨迹片段和正常轨迹片段；Step S2: segmenting the standard large-cargo transport vehicle trajectory data into large-cargo transport trajectory behavior segments; calculating the trajectory deviation of the large-cargo transport trajectory behavior segments to obtain trajectory deviation data; comparing the driving trajectory deviation data with a preset driving trajectory deviation degree threshold to generate abnormal trajectory segments and normal trajectory segments;
步骤S3：获取运输环境感知数据；基于运输环境感知数据对异常轨迹片段进行异常运输行为识别，生成异常运输行为数据；对异常运输行为数据进行异常标签处理，生成异常运输轨迹分类标签；Step S3: Acquire transportation environment perception data; identify abnormal transportation behaviors of abnormal trajectory segments based on the transportation environment perception data to generate abnormal transportation behavior data; perform abnormal label processing on the abnormal transportation behavior data to generate abnormal transportation trajectory classification labels;
步骤S4：基于异常运输行为数据进行时空关联分析，生成异常运输行为时空关联数据；对异常运输行为时空关联数据进行异常行驶轨迹预测，生成异常行驶轨迹预测数据；Step S4: performing spatiotemporal correlation analysis based on the abnormal transportation behavior data to generate spatiotemporal correlation data of abnormal transportation behavior; performing abnormal driving trajectory prediction on the spatiotemporal correlation data of abnormal transportation behavior to generate abnormal driving trajectory prediction data;
步骤S5：对正常轨迹片段进行车辆信息监管，生成正常轨迹车辆行驶监管数据；基于异常轨迹片段进行车辆预警，生成异常轨迹车辆行驶预警数据；将正常轨迹车辆行驶监管数据和异常轨迹车辆行驶预警数据进行数据集成，生成行驶轨迹监管预警架构；利用异常行驶轨迹预测数据对行驶轨迹监管预警架构进行监管及预警推送，以执行行驶轨迹的大件运输监管及预警作业。Step S5: Carry out vehicle information supervision on normal trajectory segments to generate normal trajectory vehicle driving supervision data; carry out vehicle warning based on abnormal trajectory segments to generate abnormal trajectory vehicle driving warning data; integrate normal trajectory vehicle driving supervision data and abnormal trajectory vehicle driving warning data to generate a driving trajectory supervision and warning framework; use abnormal driving trajectory prediction data to supervise and push warnings on the driving trajectory supervision and warning framework to perform large-scale transportation supervision and warning operations on the driving trajectory.
本发明通过对实时轨迹数据进行预处理，可以确保数据的准确性和一致性，为后续的分析和处理提供可靠的数据基础。通过对轨迹行为片段和环境感知数据的分析，可以识别出异常的运输行为，并对异常轨迹进行分类和标记，从而及时发现潜在的问题和安全风险。通过时空关联分析和异常行驶轨迹预测，可以深入了解异常行为的时空特征，并预测出现的异常行驶轨迹，为制定相应的监管和预警策略提供依据。对正常轨迹进行监管，并针对异常轨迹进行预警，可以及时采取措施避免潜在的安全事故，保障运输过程的安全性和顺利进行。将正常轨迹监管数据和异常轨迹预警数据进行集成，构建行驶轨迹监管预警架构，提高了对行驶轨迹的全面监管和预警能力，增强了运输安全管理的效果。因此，本发明通过数据处理、异常检测、环境感知和时空关联分析，提高了监管预警的准确性和效率。The present invention can ensure the accuracy and consistency of data by preprocessing real-time trajectory data, and provide a reliable data basis for subsequent analysis and processing. By analyzing the trajectory behavior fragments and environmental perception data, abnormal transportation behavior can be identified, and abnormal trajectories can be classified and marked, so as to timely discover potential problems and safety risks. Through spatiotemporal correlation analysis and abnormal driving trajectory prediction, the spatiotemporal characteristics of abnormal behavior can be deeply understood, and the abnormal driving trajectory that occurs can be predicted, providing a basis for formulating corresponding supervision and early warning strategies. By supervising the normal trajectory and giving early warning for the abnormal trajectory, measures can be taken in time to avoid potential safety accidents and ensure the safety and smooth progress of the transportation process. The normal trajectory supervision data and the abnormal trajectory early warning data are integrated to construct a driving trajectory supervision and early warning architecture, which improves the comprehensive supervision and early warning capabilities of the driving trajectory and enhances the effect of transportation safety management. Therefore, the present invention improves the accuracy and efficiency of supervision and early warning through data processing, anomaly detection, environmental perception and spatiotemporal correlation analysis.
本发明提供了一种基于行驶轨迹的大件运输监管及预警装置，所述基于行驶轨迹的大件运输监管及预警装置包括：The present invention provides a large-scale transport supervision and early warning device based on a driving trajectory, and the large-scale transport supervision and early warning device based on a driving trajectory comprises:
GPS定位装置、北斗定位装置、车载传感器、位置感知设备、车载摄像头、激光雷达和超声波传感器；GPS positioning devices, Beidou positioning devices, vehicle-mounted sensors, location awareness devices, vehicle-mounted cameras, lidar and ultrasonic sensors;
以及至少一个处理器；and at least one processor;
与所述至少一个处理器通信连接的存储器及多种传感器模块，传感器模块包括北斗定位模块、车载传感器模块、环境感知模块以及数据采集模块；A memory and a plurality of sensor modules communicatively connected to the at least one processor, wherein the sensor modules include a Beidou positioning module, a vehicle-mounted sensor module, an environment perception module, and a data acquisition module;
其中，所述存储器存储有可被所述至少一个处理器执行的计算机程序，所述计算机程序被所述至少一个处理器执行，以使所述至少一个处理器能够执行如上所述的任一种基于行驶轨迹的大件运输监管及预警方法。Among them, the memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute any of the large-scale transportation supervision and early warning methods based on driving trajectories as described above.
本发明还提供了一种基于行驶轨迹的大件运输监管及预警系统，用于执行如上所述的基于行驶轨迹的大件运输监管及预警方法，该基于行驶轨迹的大件运输监管及预警系统包括：The present invention also provides a large-scale transport supervision and early warning system based on driving trajectory, which is used to execute the large-scale transport supervision and early warning method based on driving trajectory as described above. The large-scale transport supervision and early warning system based on driving trajectory includes:
数据处理模块，用于获取大件运输车辆实时轨迹数据；对大件运输车辆实时轨迹数据进行数据预处理，生成标准大件运输车辆轨迹数据；The data processing module is used to obtain the real-time trajectory data of large-scale transport vehicles; perform data preprocessing on the real-time trajectory data of large-scale transport vehicles to generate standard large-scale transport vehicle trajectory data;
轨迹片段划分模块，用于对标准大件运输车辆轨迹数据进行运输行驶轨迹分段，得到大件运输轨迹行为片段；对大件运输轨迹行为片段进行轨迹偏差计算，得到轨迹偏差数据；将行驶轨迹偏差数据和预设的行驶轨迹偏差程度阈值进行对比，生成异常轨迹片段和正常轨迹片段；The trajectory segmentation module is used to segment the standard large-scale transport vehicle trajectory data into large-scale transport trajectory behavior segments; calculate the trajectory deviation of the large-scale transport trajectory behavior segments to obtain trajectory deviation data; compare the driving trajectory deviation data with a preset driving trajectory deviation degree threshold to generate abnormal trajectory segments and normal trajectory segments;
异常行为识别模块，用于获取运输环境感知数据；基于运输环境感知数据对异常轨迹片段进行异常运输行为识别，生成异常运输行为数据；对异常运输行为数据进行异常标签处理，生成异常运输轨迹分类标签；The abnormal behavior recognition module is used to obtain the transport environment perception data; based on the transport environment perception data, the abnormal transport behavior of the abnormal trajectory segment is recognized to generate the abnormal transport behavior data; the abnormal transport behavior data is processed with abnormal labels to generate abnormal transport trajectory classification labels;
异常轨迹预测模块，用于基于异常运输行为数据进行时空关联分析，生成异常运输行为时空关联数据；对异常运输行为时空关联数据进行异常行驶轨迹预测，生成异常行驶轨迹预测数据；The abnormal trajectory prediction module is used to perform spatiotemporal correlation analysis based on abnormal transportation behavior data to generate spatiotemporal correlation data of abnormal transportation behavior; perform abnormal driving trajectory prediction on the spatiotemporal correlation data of abnormal transportation behavior to generate abnormal driving trajectory prediction data;
监管及预警模块，用于对正常轨迹片段进行车辆信息监管，生成正常轨迹车辆行驶监管数据；基于异常轨迹片段进行车辆预警，生成异常轨迹车辆行驶预警数据；将正常轨迹车辆行驶监管数据和异常轨迹车辆行驶预警数据进行数据集成，生成行驶轨迹监管预警架构；利用异常行驶轨迹预测数据对行驶轨迹监管预警架构进行监管及预警推送，以执行行驶轨迹的大件运输监管及预警作业。The supervision and warning module is used to supervise the vehicle information of normal trajectory segments and generate normal trajectory vehicle driving supervision data; to carry out vehicle warning based on abnormal trajectory segments and generate abnormal trajectory vehicle driving warning data; to integrate the normal trajectory vehicle driving supervision data and the abnormal trajectory vehicle driving warning data to generate a driving trajectory supervision and warning framework; to use the abnormal driving trajectory prediction data to supervise and push warnings to the driving trajectory supervision and warning framework, so as to perform large-scale transportation supervision and warning operations on the driving trajectory.
本发明的有益效果在于通过获取大件运输车辆的实时轨迹数据并进行处理分析，能够实现对大件运输过程的实时监管和异常预警，及时发现问题并采取措施。利用运输环境感知数据对异常轨迹片段进行异常运输行为识别，能够准确识别出异常的运输行为，帮助预防事故和损失。通过时空关联分析，可以深入理解异常运输行为的发生机制和规律，有助于提高对异常情况的理解和应对能力。基于异常运输行为数据进行时空关联分析，并生成异常行驶轨迹预测数据，可以提前预警潜在的异常情况，有效减少事故发生的可能性。将正常轨迹车辆行驶监管数据与异常轨迹车辆行驶预警数据进行数据集成，形成综合的行驶轨迹监管预警架构，使监管工作更加全面和有效。通过监管和预警，能够提升大件运输的安全性和效率，保障货物和人员的安全，同时优化运输流程，降低成本。因此，本发明通过数据处理、异常检测、环境感知和时空关联分析，提高了监管预警的准确性和效率。The beneficial effect of the present invention is that by acquiring the real-time trajectory data of large-scale transport vehicles and processing and analyzing them, it is possible to realize real-time supervision and abnormal warning of the large-scale transportation process, discover problems in time and take measures. By using the transportation environment perception data to identify abnormal transportation behaviors of abnormal trajectory segments, it is possible to accurately identify abnormal transportation behaviors and help prevent accidents and losses. Through spatiotemporal correlation analysis, the occurrence mechanism and law of abnormal transportation behaviors can be deeply understood, which helps to improve the understanding and response capabilities of abnormal situations. Based on the abnormal transportation behavior data, spatiotemporal correlation analysis is performed, and abnormal driving trajectory prediction data is generated, which can warn potential abnormal situations in advance and effectively reduce the possibility of accidents. The normal trajectory vehicle driving supervision data and the abnormal trajectory vehicle driving warning data are integrated to form a comprehensive driving trajectory supervision and warning framework, making the supervision work more comprehensive and effective. Through supervision and warning, the safety and efficiency of large-scale transportation can be improved, the safety of goods and personnel can be guaranteed, and the transportation process can be optimized and the cost can be reduced. Therefore, the present invention improves the accuracy and efficiency of supervision and warning through data processing, anomaly detection, environmental perception and spatiotemporal correlation analysis.
附图说明BRIEF DESCRIPTION OF THE DRAWINGS
图1为一种基于行驶轨迹的大件运输监管及预警方法的步骤流程示意图；FIG1 is a schematic diagram of the steps of a large-scale transportation supervision and early warning method based on driving trajectory;
图2为图1中步骤S2的详细实施步骤流程示意图；FIG2 is a schematic diagram of a detailed implementation process of step S2 in FIG1 ;
图3为图1中步骤S3的详细实施步骤流程示意图；FIG3 is a schematic flow chart of detailed implementation steps of step S3 in FIG1 ;
图4为图3中步骤S32的详细实施步骤流程示意图；FIG4 is a schematic diagram of a detailed implementation process of step S32 in FIG3 ;
本发明目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings.
具体实施方式Detailed ways
下面结合附图对本发明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 technical method of the present invention is described clearly and completely below 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4，一种基于行驶轨迹的大件运输监管及预警方法，所述方法包括以下步骤：To achieve the above purpose, please refer to Figures 1 to 4, a large-scale transportation supervision and early warning method based on driving trajectory, the method includes the following steps:
步骤S1：获取大件运输车辆实时轨迹数据；对大件运输车辆实时轨迹数据进行数据预处理，生成标准大件运输车辆轨迹数据；Step S1: Acquire the real-time trajectory data of the large-scale transport vehicle; perform data preprocessing on the real-time trajectory data of the large-scale transport vehicle to generate standard large-scale transport vehicle trajectory data;
步骤S2：对标准大件运输车辆轨迹数据进行运输行驶轨迹分段，得到大件运输轨迹行为片段；对大件运输轨迹行为片段进行轨迹偏差计算，得到轨迹偏差数据；将行驶轨迹偏差数据和预设的行驶轨迹偏差程度阈值进行对比，生成异常轨迹片段和正常轨迹片段；Step S2: segmenting the standard large-cargo transport vehicle trajectory data into large-cargo transport trajectory behavior segments; calculating the trajectory deviation of the large-cargo transport trajectory behavior segments to obtain trajectory deviation data; comparing the driving trajectory deviation data with a preset driving trajectory deviation degree threshold to generate abnormal trajectory segments and normal trajectory segments;
步骤S3：获取运输环境感知数据；基于运输环境感知数据对异常轨迹片段进行异常运输行为识别，生成异常运输行为数据；对异常运输行为数据进行异常标签处理，生成异常运输轨迹分类标签；Step S3: Acquire transportation environment perception data; identify abnormal transportation behaviors of abnormal trajectory segments based on the transportation environment perception data to generate abnormal transportation behavior data; perform abnormal label processing on the abnormal transportation behavior data to generate abnormal transportation trajectory classification labels;
步骤S4：基于异常运输行为数据进行时空关联分析，生成异常运输行为时空关联数据；对异常运输行为时空关联数据进行异常行驶轨迹预测，生成异常行驶轨迹预测数据；Step S4: performing spatiotemporal correlation analysis based on the abnormal transportation behavior data to generate spatiotemporal correlation data of abnormal transportation behavior; performing abnormal driving trajectory prediction on the spatiotemporal correlation data of abnormal transportation behavior to generate abnormal driving trajectory prediction data;
步骤S5：对正常轨迹片段进行车辆信息监管，生成正常轨迹车辆行驶监管数据；基于异常轨迹片段进行车辆预警，生成异常轨迹车辆行驶预警数据；将正常轨迹车辆行驶监管数据和异常轨迹车辆行驶预警数据进行数据集成，生成行驶轨迹监管预警架构；利用异常行驶轨迹预测数据对行驶轨迹监管预警架构进行监管及预警推送，以执行行驶轨迹的大件运输监管及预警作业。Step S5: Carry out vehicle information supervision on normal trajectory segments to generate normal trajectory vehicle driving supervision data; carry out vehicle warning based on abnormal trajectory segments to generate abnormal trajectory vehicle driving warning data; integrate normal trajectory vehicle driving supervision data and abnormal trajectory vehicle driving warning data to generate a driving trajectory supervision and warning framework; use abnormal driving trajectory prediction data to supervise and push warnings on the driving trajectory supervision and warning framework to perform large-scale transportation supervision and warning operations on the driving trajectory.
本发明通过对实时轨迹数据进行预处理，可以确保数据的准确性和一致性，为后续的分析和处理提供可靠的数据基础。通过对轨迹行为片段和环境感知数据的分析，可以识别出异常的运输行为，并对异常轨迹进行分类和标记，从而及时发现潜在的问题和安全风险。通过时空关联分析和异常行驶轨迹预测，可以深入了解异常行为的时空特征，并预测出现的异常行驶轨迹，为制定相应的监管和预警策略提供依据。对正常轨迹进行监管，并针对异常轨迹进行预警，可以及时采取措施避免潜在的安全事故，保障运输过程的安全性和顺利进行。将正常轨迹监管数据和异常轨迹预警数据进行集成，构建行驶轨迹监管预警架构，提高了对行驶轨迹的全面监管和预警能力，增强了运输安全管理的效果。因此，本发明通过数据处理、异常检测、环境感知和时空关联分析，提高了监管预警的准确性和效率。The present invention can ensure the accuracy and consistency of data by preprocessing real-time trajectory data, and provide a reliable data basis for subsequent analysis and processing. By analyzing the trajectory behavior fragments and environmental perception data, abnormal transportation behavior can be identified, and abnormal trajectories can be classified and marked, so as to timely discover potential problems and safety risks. Through spatiotemporal correlation analysis and abnormal driving trajectory prediction, the spatiotemporal characteristics of abnormal behavior can be deeply understood, and the abnormal driving trajectory that occurs can be predicted, providing a basis for formulating corresponding supervision and early warning strategies. By supervising the normal trajectory and giving early warning for the abnormal trajectory, measures can be taken in time to avoid potential safety accidents and ensure the safety and smooth progress of the transportation process. The normal trajectory supervision data and the abnormal trajectory early warning data are integrated to construct a driving trajectory supervision and early warning architecture, which improves the comprehensive supervision and early warning capabilities of the driving trajectory and enhances the effect of transportation safety management. Therefore, the present invention improves the accuracy and efficiency of supervision and early warning through data processing, anomaly detection, environmental perception and spatiotemporal correlation analysis.
本发明实施例中，参考图1所述，为本发明一种基于行驶轨迹的大件运输监管及预警方法的步骤流程示意图，在本实例中，所述一种基于行驶轨迹的大件运输监管及预警方法包括以下步骤：In the embodiment of the present invention, referring to FIG. 1, a schematic diagram of the steps of a large-scale transportation supervision and early warning method based on a driving trajectory of the present invention is shown. In this example, the large-scale transportation supervision and early warning method based on a driving trajectory includes the following steps:
步骤S1：获取大件运输车辆实时轨迹数据；对大件运输车辆实时轨迹数据进行数据预处理，生成标准大件运输车辆轨迹数据；Step S1: Acquire the real-time trajectory data of the large-scale transport vehicle; perform data preprocessing on the real-time trajectory data of the large-scale transport vehicle to generate standard large-scale transport vehicle trajectory data;
本发明实施例中，通过使用GPS设备或车载传感器等装置，实时采集大件运输车辆的位置信息和轨迹数据。运输公司或相关机构建立数据采集系统，确保及时、准确地获取车辆轨迹数据。清洗和过滤原始数据，去除存在的错误、异常或不完整的数据。根据需要对数据进行时间戳校准，以确保数据的一致性和准确性。将原始数据转换为统一的数据格式，以便后续处理和分析。确保数据格式符合标准，便于后续的数据预处理和分析。对轨迹数据进行去噪处理，滤除噪声或异常点。对轨迹数据进行平滑处理，以消除数据中的抖动或不连续性，提高数据的可读性和稳定性。标准化轨迹数据的格式和单位，确保数据的一致性和可比性。根据需要对轨迹数据进行空间校正，以消除偏差或误差。将预处理后的标准轨迹数据存储到数据库或数据仓库中，便于后续的分析和应用。确保数据存储和管理的安全性和可靠性，防止数据丢失或泄露。In the embodiment of the present invention, the location information and trajectory data of large-scale transport vehicles are collected in real time by using devices such as GPS devices or vehicle-mounted sensors. The transport company or relevant agency establishes a data collection system to ensure timely and accurate acquisition of vehicle trajectory data. The raw data is cleaned and filtered to remove existing errors, abnormalities or incomplete data. The data is time-stamped and calibrated as needed to ensure the consistency and accuracy of the data. The raw data is converted into a unified data format for subsequent processing and analysis. Ensure that the data format meets the standard to facilitate subsequent data preprocessing and analysis. De-noise the trajectory data to filter out noise or abnormal points. Smooth the trajectory data to eliminate jitter or discontinuity in the data and improve the readability and stability of the data. Standardize the format and unit of the trajectory data to ensure the consistency and comparability of the data. Perform spatial correction on the trajectory data as needed to eliminate deviations or errors. The preprocessed standard trajectory data is stored in a database or data warehouse for subsequent analysis and application. Ensure the security and reliability of data storage and management to prevent data loss or leakage.
步骤S2：对标准大件运输车辆轨迹数据进行运输行驶轨迹分段，得到大件运输轨迹行为片段；对大件运输轨迹行为片段进行轨迹偏差计算，得到轨迹偏差数据；将行驶轨迹偏差数据和预设的行驶轨迹偏差程度阈值进行对比，生成异常轨迹片段和正常轨迹片段；Step S2: segmenting the standard large-cargo transport vehicle trajectory data into large-cargo transport trajectory behavior segments; calculating the trajectory deviation of the large-cargo transport trajectory behavior segments to obtain trajectory deviation data; comparing the driving trajectory deviation data with a preset driving trajectory deviation degree threshold to generate abnormal trajectory segments and normal trajectory segments;
本发明实施例中，通过将标准大件运输车辆轨迹数据按照一定的时间或距离间隔进行分段，形成连续的运输行驶轨迹片段。利用轨迹分段技术，如分段聚类、分段切割等方法，将轨迹数据分割成具有一定长度的子段。对每个运输行驶轨迹片段进行轨迹偏差计算，评估车辆行驶轨迹与预期行驶轨迹之间的差异。常用的偏差计算方法包括距离偏差、角度偏差、速度偏差等，具体根据具体情况选择合适的计算方法。将轨迹偏差数据与预设的行驶轨迹偏差程度阈值进行对比，判断每个轨迹片段是否异常。设定适当的阈值，超过阈值的轨迹片段被视为异常，否则为正常。将被识别为异常的轨迹片段提取出来，形成异常轨迹片段集合。异常轨迹片段的提取具体采用基于阈值的方法或基于机器学习的方法。将未被识别为异常的轨迹片段提取出来，形成正常轨迹片段集合。正常轨迹片段的提取是异常检测的补充，有助于全面理解车辆的行驶状况。In an embodiment of the present invention, the trajectory data of a standard large-scale transport vehicle is segmented according to a certain time or distance interval to form a continuous transport driving trajectory segment. The trajectory segmentation technology, such as segment clustering, segment cutting and other methods, is used to segment the trajectory data into sub-segments with a certain length. The trajectory deviation is calculated for each transport driving trajectory segment to evaluate the difference between the vehicle driving trajectory and the expected driving trajectory. Common deviation calculation methods include distance deviation, angle deviation, speed deviation, etc., and a suitable calculation method is selected according to the specific situation. The trajectory deviation data is compared with a preset driving trajectory deviation degree threshold to determine whether each trajectory segment is abnormal. An appropriate threshold is set, and the trajectory segment exceeding the threshold is regarded as abnormal, otherwise it is normal. The trajectory segment identified as abnormal is extracted to form an abnormal trajectory segment set. The extraction of abnormal trajectory segments specifically adopts a threshold-based method or a machine learning-based method. The trajectory segment that is not identified as abnormal is extracted to form a normal trajectory segment set. The extraction of normal trajectory segments is a supplement to abnormality detection and helps to fully understand the driving condition of the vehicle.
步骤S3：获取运输环境感知数据；基于运输环境感知数据对异常轨迹片段进行异常运输行为识别，生成异常运输行为数据；对异常运输行为数据进行异常标签处理，生成异常运输轨迹分类标签；Step S3: Acquire transportation environment perception data; identify abnormal transportation behaviors of abnormal trajectory segments based on the transportation environment perception data to generate abnormal transportation behavior data; perform abnormal label processing on the abnormal transportation behavior data to generate abnormal transportation trajectory classification labels;
本发明实施例中，通过部署传感器网络或安装环境感知设备，实时监测运输过程中的环境因素，如路面情况、天气状况、交通情况等。利用现有的交通监控系统、气象站、路况信息服务等渠道获取相关的环境感知数据。将异常轨迹片段与运输环境感知数据进行关联分析，找出异常行为与环境因素之间的关联性。基于机器学习算法，如分类算法、聚类算法等，对异常轨迹片段进行异常运输行为的识别和分类。对识别出的异常运输行为数据进行标签化处理，区分不同类型的异常行为，如急加速、急刹车、超速行驶等。根据异常行为的性质和程度，为每个异常轨迹片段生成相应的异常标签，以便后续的分析和应用。对异常运输行为数据进行统计分析和可视化展示，深入理解异常行为的分布和趋势，发现潜在的安全隐患和问题点。利用数据挖掘和可视化技术，挖掘异常行为背后的规律和原因，为制定针对性的监管和预防措施提供参考。In the embodiment of the present invention, by deploying a sensor network or installing an environmental perception device, the environmental factors in the transportation process, such as road conditions, weather conditions, traffic conditions, etc., are monitored in real time. Relevant environmental perception data is obtained by using existing traffic monitoring systems, weather stations, road condition information services and other channels. The abnormal trajectory fragments are associated with the transportation environment perception data to find out the correlation between abnormal behavior and environmental factors. Based on machine learning algorithms, such as classification algorithms and clustering algorithms, abnormal transportation behaviors are identified and classified for abnormal trajectory fragments. The identified abnormal transportation behavior data is labeled to distinguish different types of abnormal behaviors, such as sudden acceleration, sudden braking, speeding, etc. According to the nature and degree of abnormal behavior, corresponding abnormal labels are generated for each abnormal trajectory fragment for subsequent analysis and application. Statistical analysis and visualization of abnormal transportation behavior data are performed to deeply understand the distribution and trend of abnormal behavior and discover potential safety hazards and problem points. Data mining and visualization technology are used to explore the laws and causes behind abnormal behavior, providing reference for the formulation of targeted supervision and preventive measures.
步骤S4：基于异常运输行为数据进行时空关联分析，生成异常运输行为时空关联数据；对异常运输行为时空关联数据进行异常行驶轨迹预测，生成异常行驶轨迹预测数据；Step S4: performing spatiotemporal correlation analysis based on the abnormal transportation behavior data to generate spatiotemporal correlation data of abnormal transportation behavior; performing abnormal driving trajectory prediction on the spatiotemporal correlation data of abnormal transportation behavior to generate abnormal driving trajectory prediction data;
本发明实施例中，通过将异常运输行为数据与时空因素进行关联分析，例如与时间、地点、运输路线等相关联。利用时空数据挖掘技术，如时空聚类、时空关联规则挖掘等，发现异常运输行为在时空维度上的分布和关联关系。基于已识别的异常运输行为时空关联数据，利用机器学习或统计模型进行异常行驶轨迹的预测。具体采用时间序列分析、回归分析、神经网络等方法进行轨迹预测，考虑时空因素对轨迹的影响。利用历史数据进行模型训练，包括异常运输行为数据和相关的时空因素数据。使用交叉验证等技术对模型进行评估，确保模型的预测性能和泛化能力。根据训练好的预测模型，对当前的异常运输行为数据进行预测，生成异常行驶轨迹预测数据。预测数据具体包括轨迹路径、发生异常的时间点、异常程度等信息。将生成的异常行驶轨迹预测数据存储到数据库或数据文件中，便于后续的分析和应用。In an embodiment of the present invention, the abnormal transportation behavior data is analyzed by associating with spatiotemporal factors, such as time, location, transportation route, etc. Spatiotemporal data mining techniques, such as spatiotemporal clustering, spatiotemporal association rule mining, etc., are used to discover the distribution and association relationship of abnormal transportation behaviors in the spatiotemporal dimension. Based on the identified spatiotemporal association data of abnormal transportation behaviors, machine learning or statistical models are used to predict abnormal driving trajectories. Specifically, time series analysis, regression analysis, neural network and other methods are used to predict trajectories, considering the influence of spatiotemporal factors on trajectories. Historical data are used to train the model, including abnormal transportation behavior data and related spatiotemporal factor data. The model is evaluated using cross-validation and other techniques to ensure the prediction performance and generalization ability of the model. According to the trained prediction model, the current abnormal transportation behavior data is predicted to generate abnormal driving trajectory prediction data. The prediction data specifically includes information such as trajectory path, time point of abnormal occurrence, degree of abnormality, etc. The generated abnormal driving trajectory prediction data is stored in a database or data file for subsequent analysis and application.
步骤S5：对正常轨迹片段进行车辆信息监管，生成正常轨迹车辆行驶监管数据；基于异常轨迹片段进行车辆预警，生成异常轨迹车辆行驶预警数据；将正常轨迹车辆行驶监管数据和异常轨迹车辆行驶预警数据进行数据集成，生成行驶轨迹监管预警架构；利用异常行驶轨迹预测数据对行驶轨迹监管预警架构进行监管及预警推送，以执行行驶轨迹的大件运输监管及预警作业。Step S5: Carry out vehicle information supervision on normal trajectory segments to generate normal trajectory vehicle driving supervision data; carry out vehicle warning based on abnormal trajectory segments to generate abnormal trajectory vehicle driving warning data; integrate normal trajectory vehicle driving supervision data and abnormal trajectory vehicle driving warning data to generate a driving trajectory supervision and warning framework; use abnormal driving trajectory prediction data to supervise and push warnings on the driving trajectory supervision and warning framework to perform large-scale transportation supervision and warning operations on the driving trajectory.
本发明实施例中，通过对正常轨迹片段进行车辆信息监管，包括车辆位置、速度、行驶时间等信息的监测和记录。根据监管要求和标准，生成正常轨迹车辆行驶监管数据，以便监管部门进行监督和管理。基于异常轨迹片段进行车辆预警，识别存在安全隐患的车辆和行驶情况。根据异常轨迹的特征和严重程度，生成相应的异常轨迹车辆行驶预警数据，包括预警信息和建议措施。将正常轨迹车辆行驶监管数据和异常轨迹车辆行驶预警数据进行数据集成，形成综合的行驶轨迹监管预警架构。利用数据集成技术，将不同来源和类型的数据整合在一起，实现数据的统一管理和综合分析。利用异常行驶轨迹预测数据对行驶轨迹监管预警架构进行监管和预警推送。基于预测结果和监管标准，及时向相关部门或人员发送监管通知、预警警报或建议措施，以执行大件运输的监管和预警作业。建立监管预警平台，集成数据处理、分析和推送功能，实现自动化的监管预警流程。利用先进的技术手段，如人工智能、大数据分析等，提升监管预警的准确性和效率。In the embodiment of the present invention, vehicle information supervision is performed on normal trajectory segments, including monitoring and recording of vehicle location, speed, driving time and other information. According to regulatory requirements and standards, normal trajectory vehicle driving supervision data is generated so that the regulatory department can supervise and manage. Vehicle warning is performed based on abnormal trajectory segments to identify vehicles and driving conditions with safety hazards. According to the characteristics and severity of the abnormal trajectory, corresponding abnormal trajectory vehicle driving warning data is generated, including warning information and recommended measures. The normal trajectory vehicle driving supervision data and the abnormal trajectory vehicle driving warning data are integrated to form a comprehensive driving trajectory supervision and warning framework. Using data integration technology, data from different sources and types are integrated to achieve unified data management and comprehensive analysis. The driving trajectory supervision and warning framework is supervised and warned using abnormal driving trajectory prediction data. Based on the prediction results and regulatory standards, supervision notices, warning alarms or recommended measures are sent to relevant departments or personnel in a timely manner to perform supervision and warning operations for large-scale transportation. A supervision and warning platform is established to integrate data processing, analysis and push functions to realize an automated supervision and warning process. Advanced technical means, such as artificial intelligence and big data analysis, are used to improve the accuracy and efficiency of supervision and warning.
优选的，步骤S1包括以下步骤：Preferably, step S1 comprises the following steps:
步骤S11：利用传感器获取大件运输车辆实时轨迹数据；Step S11: using sensors to obtain real-time trajectory data of large-scale transport vehicles;
步骤S12：对大件运输车辆实时轨迹数据进行数据噪声过滤，生成大件运输车辆过滤数据；Step S12: filtering the data noise of the real-time trajectory data of the large-scale transport vehicle to generate the large-scale transport vehicle filtering data;
步骤S13：对大件运输车辆过滤数据进行数据平滑，生成大件运输车辆平滑数据；Step S13: smoothing the filtered data of large-scale transport vehicles to generate smoothed data of large-scale transport vehicles;
步骤S14：利用Z-score标准化方法对大件运输车辆平滑数据进行数据标准化，生成标准大件运输车辆轨迹数据。Step S14: using the Z-score standardization method to standardize the smoothed data of the large-scale transport vehicle to generate standard large-scale transport vehicle trajectory data.
本发明通过实时获取大件运输车辆的轨迹数据，为后续的监管和预警提供了实时数据支持，有助于及时发现异常情况和安全隐患。去除轨迹数据中的噪声和异常值，提高数据的质量和可靠性，减少误判和偏差，确保后续分析的准确性。平滑轨迹数据，消除数据中的抖动和不连续性，使轨迹信息更加清晰和连贯，便于后续的分析和理解。标准化轨迹数据的格式和单位，使数据具有统一的尺度和量级，便于不同来源的数据进行比较和整合，提高数据的可比性和可用性。The present invention provides real-time data support for subsequent supervision and early warning by acquiring the trajectory data of large-scale transport vehicles in real time, which helps to timely discover abnormal situations and safety hazards. Remove noise and outliers in trajectory data, improve data quality and reliability, reduce misjudgments and deviations, and ensure the accuracy of subsequent analysis. Smooth trajectory data, eliminate jitter and discontinuity in the data, make trajectory information clearer and more coherent, and facilitate subsequent analysis and understanding. Standardize the format and unit of trajectory data so that the data has a unified scale and magnitude, facilitate comparison and integration of data from different sources, and improve data comparability and availability.
本发明实施例中，通过部署全球定位系统（GPS）传感器或其他位置感知设备在大件运输车辆上，实时采集车辆的位置信息。利用车载传感器、无线通信技术等设备，将实时轨迹数据传输至数据中心或监控中心。利用滤波算法，如均值滤波、中值滤波等，对实时轨迹数据进行噪声过滤，去除异常值和误差。设定阈值或规则，排除异常位置点或速度点，确保数据的准确性和可靠性。应用平滑算法，如移动平均法、指数平滑法等，对过滤后的轨迹数据进行平滑处理，消除数据中的抖动和波动。确保平滑处理后的数据保持原有特征，同时具有一定的连续性和稳定性。使用Z-score标准化方法，对平滑后的轨迹数据进行标准化处理，将数据转换为具有零均值和单位标准差的标准分布。确保标准化后的数据具有统一的尺度和量级，便于后续的分析和比较。In an embodiment of the present invention, a global positioning system (GPS) sensor or other location sensing device is deployed on a large transport vehicle to collect the location information of the vehicle in real time. The real-time trajectory data is transmitted to a data center or a monitoring center using on-board sensors, wireless communication technology and other equipment. The real-time trajectory data is filtered for noise using filtering algorithms, such as mean filtering, median filtering, etc., to remove outliers and errors. Thresholds or rules are set to exclude abnormal position points or speed points to ensure the accuracy and reliability of the data. Smoothing algorithms, such as moving average method, exponential smoothing method, etc., are applied to smooth the filtered trajectory data to eliminate jitter and fluctuations in the data. It is ensured that the smoothed data maintains the original characteristics and has a certain degree of continuity and stability. The smoothed trajectory data is standardized using the Z-score standardization method to convert the data into a standard distribution with zero mean and unit standard deviation. It is ensured that the standardized data has a uniform scale and magnitude to facilitate subsequent analysis and comparison.
优选的，步骤S2包括以下步骤：Preferably, step S2 comprises the following steps:
步骤S21：对标准大件运输车辆轨迹数据进行运输行驶轨迹分段，得到大件运输轨迹行为片段；Step S21: segmenting the standard large-scale transport vehicle trajectory data into large-scale transport trajectory behavior segments;
步骤S22：对大件运输轨迹行为片段进行行驶特征提取，得到轨迹片段特征向量，其中形式特征提取包括行驶速度提取、加速度变化提取和转弯角度提取；Step S22: extracting driving features from the large-scale transport trajectory behavior segment to obtain a trajectory segment feature vector, wherein the form feature extraction includes driving speed extraction, acceleration change extraction and turning angle extraction;
步骤S23：利用轨迹偏差计算公式对轨迹片段特征向量进行轨迹偏差计算，得到轨迹偏差数据；Step S23: using a trajectory deviation calculation formula to calculate the trajectory deviation of the trajectory segment feature vector to obtain trajectory deviation data;
步骤S24：将行驶轨迹偏差数据和预设的行驶轨迹偏差程度阈值进行对比，当行驶轨迹偏差程度大于或等于预设的行驶轨迹偏差程度阈值时，则将对应的轨迹片段特征向量标记为异常轨迹片段；当行驶轨迹偏差程度小于预设的行驶轨迹偏差程度阈值时，则将对应的轨迹片段特征向量标记为正常轨迹片段。Step S24: Compare the driving trajectory deviation data with the preset driving trajectory deviation degree threshold. When the driving trajectory deviation degree is greater than or equal to the preset driving trajectory deviation degree threshold, the corresponding trajectory segment feature vector is marked as an abnormal trajectory segment; when the driving trajectory deviation degree is less than the preset driving trajectory deviation degree threshold, the corresponding trajectory segment feature vector is marked as a normal trajectory segment.
本发明通过将整体的轨迹数据划分为多个片段，有助于对不同行驶阶段的行为特征进行分析和处理，提高对异常行为的检测和识别能力。通过提取轨迹片段的特征向量，可以更全面地描述轨迹行为，为后续的轨迹偏差计算和异常检测提供更丰富的数据基础。通过轨迹偏差数据的计算，可以量化轨迹片段的偏差程度，为后续的异常识别和预警提供依据。比较轨迹偏差数据与预设的偏差程度阈值，确定轨迹片段的异常与否，能够及时识别出异常的轨迹片段，提高对异常行为的检测和预警效率，保障运输安全。The present invention divides the overall trajectory data into multiple segments, which helps to analyze and process the behavioral characteristics of different driving stages and improve the detection and identification capabilities of abnormal behaviors. By extracting the feature vectors of the trajectory segments, the trajectory behavior can be described more comprehensively, providing a richer data basis for subsequent trajectory deviation calculations and anomaly detection. By calculating the trajectory deviation data, the degree of deviation of the trajectory segments can be quantified, providing a basis for subsequent anomaly identification and early warning. By comparing the trajectory deviation data with the preset deviation degree threshold to determine whether the trajectory segment is abnormal or not, abnormal trajectory segments can be identified in a timely manner, improving the efficiency of abnormal behavior detection and early warning, and ensuring transportation safety.
作为本发明的一个实例，参考图2所示，在本实例中所述步骤S2包括：As an example of the present invention, referring to FIG. 2 , in this example, step S2 includes:
步骤S21：对标准大件运输车辆轨迹数据进行运输行驶轨迹分段，得到大件运输轨迹行为片段；Step S21: segmenting the standard large-scale transport vehicle trajectory data into large-scale transport trajectory behavior segments;
本发明实施例中，通过确定数据分段的策略，具体基于时间间隔或者基于距离间隔进行分段。时间间隔是固定的，比如每隔一分钟或者每隔五分钟分段一次；距离间隔具体根据实际情况设定，比如每隔一公里或者每隔十公里分段一次。对标准大件运输车辆轨迹数据进行处理，确保数据格式的一致性和准确性。需要对数据进行清洗和预处理，包括去除重复数据、修复缺失数据、处理异常数据等。选择合适的分段算法，根据车辆的行驶特性和轨迹数据的分布情况，确定如何将轨迹数据进行划分。常见的分段算法包括基于时间的分段、基于距离的分段以及基于速度变化的分段等。根据选择的分段策略和分段算法，对标准大件运输车辆轨迹数据进行分段操作，得到大件运输轨迹行为片段。每个片段代表车辆在一段时间或一段距离内的行驶情况。将分段后的轨迹数据存储到数据库或数据文件中，便于后续的分析和应用。确保数据存储和管理的安全性和可靠性，以防止数据丢失或泄露。In an embodiment of the present invention, by determining a data segmentation strategy, segmentation is specifically performed based on a time interval or a distance interval. The time interval is fixed, such as segmenting once every one minute or every five minutes; the distance interval is specifically set according to the actual situation, such as segmenting once every one kilometer or every ten kilometers. The trajectory data of standard large-scale transport vehicles is processed to ensure the consistency and accuracy of the data format. The data needs to be cleaned and preprocessed, including removing duplicate data, repairing missing data, and processing abnormal data. Select a suitable segmentation algorithm to determine how to divide the trajectory data according to the driving characteristics of the vehicle and the distribution of the trajectory data. Common segmentation algorithms include time-based segmentation, distance-based segmentation, and speed-change-based segmentation. According to the selected segmentation strategy and segmentation algorithm, the trajectory data of standard large-scale transport vehicles is segmented to obtain large-scale transport trajectory behavior segments. Each segment represents the driving condition of the vehicle within a period of time or a distance. The segmented trajectory data is stored in a database or a data file for subsequent analysis and application. Ensure the security and reliability of data storage and management to prevent data loss or leakage.
步骤S22：对大件运输轨迹行为片段进行行驶特征提取，得到轨迹片段特征向量，其中形式特征提取包括行驶速度提取、加速度变化提取和转弯角度提取；Step S22: extracting driving features from the large-scale transport trajectory behavior segment to obtain a trajectory segment feature vector, wherein the form feature extraction includes driving speed extraction, acceleration change extraction and turning angle extraction;
本发明实施例中，通过使用轨迹数据中的位置信息和时间戳，计算车辆在每个轨迹片段中的平均行驶速度。具体采用简单的方法，如计算两个连续位置点之间的距离，并除以时间间隔来得到平均速度。根据轨迹数据中的速度信息和时间戳，计算车辆在每个轨迹片段中的加速度变化情况。具体通过计算速度的变化率来近似估计加速度，或者使用数值微分等方法来精确计算加速度。利用轨迹数据中的位置信息，计算车辆在每个轨迹片段中的转弯角度。具体采用向量运算或三角函数等方法，计算车辆运动方向的变化角度，以此作为转弯角度的估计。将提取的行驶特征（如速度、加速度变化、转弯角度等）组合成特征向量。特征向量的每个维度对应一个行驶特征，可以将不同特征的数值归一化处理，确保它们具有相同的量纲和重要性。In an embodiment of the present invention, the average driving speed of the vehicle in each trajectory segment is calculated by using the position information and timestamp in the trajectory data. Specifically, a simple method is used, such as calculating the distance between two consecutive position points and dividing it by the time interval to obtain the average speed. According to the speed information and timestamp in the trajectory data, the acceleration change of the vehicle in each trajectory segment is calculated. Specifically, the acceleration is approximately estimated by calculating the rate of change of the speed, or the acceleration is accurately calculated using methods such as numerical differentiation. Using the position information in the trajectory data, the turning angle of the vehicle in each trajectory segment is calculated. Specifically, vector operations or trigonometric functions are used to calculate the angle of change in the direction of movement of the vehicle, which is used as an estimate of the turning angle. The extracted driving features (such as speed, acceleration change, turning angle, etc.) are combined into a feature vector. Each dimension of the feature vector corresponds to a driving feature, and the values of different features can be normalized to ensure that they have the same dimension and importance.
步骤S23：利用轨迹偏差计算公式对轨迹片段特征向量进行轨迹偏差计算，得到轨迹偏差数据；Step S23: using a trajectory deviation calculation formula to calculate the trajectory deviation of the trajectory segment feature vector to obtain trajectory deviation data;
本发明实施例中，通过根据具体需求和应用场景，选择合适的轨迹偏差计算公式，公式具体基于轨迹片段的特征向量进行计算，以衡量实际轨迹与理想轨迹之间的偏差程度。根据所选的计算公式，将轨迹片段的特征向量代入，计算轨迹片段的偏差值，偏差值具体表示轨迹片段在速度、加速度变化、转弯角度等方面与预期行驶轨迹的差异程度。将计算得到的轨迹偏差值组成轨迹偏差数据。每个轨迹片段都对应着一个轨迹偏差值，用来描述该片段与预期轨迹的偏差情况。In an embodiment of the present invention, a suitable trajectory deviation calculation formula is selected according to specific needs and application scenarios. The formula is specifically based on the characteristic vector of the trajectory segment for calculation to measure the degree of deviation between the actual trajectory and the ideal trajectory. According to the selected calculation formula, the characteristic vector of the trajectory segment is substituted to calculate the deviation value of the trajectory segment. The deviation value specifically indicates the degree of difference between the trajectory segment and the expected driving trajectory in terms of speed, acceleration change, turning angle, etc. The calculated trajectory deviation value is composed of trajectory deviation data. Each trajectory segment corresponds to a trajectory deviation value, which is used to describe the deviation of the segment from the expected trajectory.
步骤S24：将行驶轨迹偏差数据和预设的行驶轨迹偏差程度阈值进行对比，当行驶轨迹偏差程度大于或等于预设的行驶轨迹偏差程度阈值时，则将对应的轨迹片段特征向量标记为异常轨迹片段；当行驶轨迹偏差程度小于预设的行驶轨迹偏差程度阈值时，则将对应的轨迹片段特征向量标记为正常轨迹片段。Step S24: Compare the driving trajectory deviation data with the preset driving trajectory deviation degree threshold. When the driving trajectory deviation degree is greater than or equal to the preset driving trajectory deviation degree threshold, the corresponding trajectory segment feature vector is marked as an abnormal trajectory segment; when the driving trajectory deviation degree is less than the preset driving trajectory deviation degree threshold, the corresponding trajectory segment feature vector is marked as a normal trajectory segment.
本发明实施例中，通过设定行驶轨迹偏差程度的阈值，阈值具体是预先设定的固定值，也可以根据历史数据或者模型预测得到。将步骤S23中得到的轨迹偏差数据与设定的行驶轨迹偏差程度阈值进行对比。如果轨迹偏差数据大于或等于预设的阈值，说明该轨迹片段存在异常，应将其标记为异常轨迹片段。如果轨迹偏差数据小于预设的阈值，则说明该轨迹片段正常，应将其标记为正常轨迹片段。将根据对比结果标记的异常轨迹片段和正常轨迹片段记录下来，形成异常轨迹片段和正常轨迹片段的列表或数据集。In an embodiment of the present invention, by setting a threshold value of the driving trajectory deviation degree, the threshold value is specifically a pre-set fixed value, and can also be obtained based on historical data or model prediction. The trajectory deviation data obtained in step S23 is compared with the set driving trajectory deviation degree threshold value. If the trajectory deviation data is greater than or equal to the preset threshold value, it means that the trajectory segment is abnormal and should be marked as an abnormal trajectory segment. If the trajectory deviation data is less than the preset threshold value, it means that the trajectory segment is normal and should be marked as a normal trajectory segment. The abnormal trajectory segments and normal trajectory segments marked according to the comparison results are recorded to form a list or data set of abnormal trajectory segments and normal trajectory segments.
优选的，步骤S23中的轨迹偏差计算公式如下所示：Preferably, the trajectory deviation calculation formula in step S23 is as follows:
； ;
式中，表示为第一个轨迹片段和第二个轨迹片段之间的偏差，表示为第一个轨迹片段在第一个时间点的位置，表示为第二个轨迹片段在第一个时间点的位置，表示为轨迹片段的起始时间，表示为轨迹片段的结束时间，表示为轨迹特征向量中的特征数量，表示为用于调节各个特征对总偏差的贡献参数，表示为用于调节各个特征对总偏差的影响程度参数，表示为第一个轨迹片段特征向量中的第个特征在第一个时间点的值，表示为第二个轨迹片段特征向量中的第个特征在第一个时间点的值，表示为用于调节积分项对总偏差的贡献参数，表示为用于调节积分项对总偏差的影响程度参数，表示为第一个轨迹片段在第二个时间点的速度特性参数，表示为第二个轨迹片段在第二个时间点的速度特性参数。In the formula, Represented as the first trajectory segment and the second trajectory segment The deviation between Represented as the first trajectory segment at the first time point s position, Represented as the second trajectory segment at the first time point s position, is the start time of the trajectory segment, Represents the end time of the trajectory segment, Expressed as the number of features in the trajectory feature vector, It is expressed as a parameter used to adjust the contribution of each feature to the total deviation. It is expressed as a parameter used to adjust the influence of each feature on the total deviation. Represented as the first trajectory segment feature vector Features at the first time point The value of Represented as the first Features at the first time point The value of Expressed as a parameter used to adjust the contribution of the integral term to the total deviation, It is expressed as a parameter used to adjust the influence of the integral term on the total deviation. Represented as the first trajectory segment at the second time point The speed characteristic parameters, Represented as the second trajectory segment at the second time point Speed characteristic parameters.
本发明通过分析并整合了一种轨迹偏差计算公式，公式描述了两个轨迹片段之间的偏差计算方法，其主要目的是衡量两个轨迹片段之间的相似程度或差异程度。公式中的总体偏差计算（整个公式的外部部分）：这部分包括了轨迹片段间位置偏差和速度特性偏差的综合计算。偏差的计算采用了两个轨迹片段间位置和速度的差异，通过加权求和来综合衡量两者的差异程度。公式中的计算了两个轨迹片段在每个时间点上位置的差异。每个特征的差异被权衡和调节，以便于不同特征对总体偏差的影响可以进行调节。参数用于调节各个特征对总偏差的贡献，参数用于调节各个特征对总偏差的影响程度。公式中的速度特性偏差计算计算了两个轨迹片段速度特性的差异。这部分通过对速度的积分来量化速度特性的整体差异。参数用于调节积分项对总偏差的贡献，参数用于调节积分项对总偏差的影响程度。公式的整体参数、控制了位置偏差的计算方式，允许不同特征对总体偏差的贡献和影响程度不同。、控制了速度特性偏差的计算方式，允许速度特性对总体偏差的贡献和影响程度不同。在使用本领域常规的轨迹偏差计算公式时，可以得到第一个轨迹片段和第二个轨迹片段之间的偏差值，通过应用本发明提供的轨迹偏差计算公式，可以更加精确的计算出第一个轨迹片段和第二个轨迹片段之间的偏差值。公式综合考虑了轨迹片段在位置和速度特性上的差异，综合考量了轨迹的整体偏差。这种综合性能够更全面地描述轨迹之间的相似性或差异性，提供了更全面的比较依据。公式可以适用于不同类型的轨迹数据，包括但不限于运动轨迹、路径规划、车辆轨迹等，广泛的适用性使得该公式可以应用于多个领域，包括交通管理、地理信息系统、移动应用等。通过将轨迹片段的差异量化为数值，公式提供了一种可量化的方式来比较和评估轨迹之间的差异程度，有助于进行更精确的轨迹匹配、分类或异常检测。The present invention analyzes and integrates a trajectory deviation calculation formula, which describes the deviation calculation method between two trajectory segments. Its main purpose is to measure the similarity or difference between the two trajectory segments. The overall deviation calculation in the formula (the external part of the entire formula): This part includes the comprehensive calculation of the position deviation and speed characteristic deviation between the trajectory segments. The deviation calculation uses the difference in position and speed between the two trajectory segments, and comprehensively measures the difference between the two through weighted summation. The difference between the positions of the two trajectory segments at each time point is calculated. The difference of each feature is weighed and adjusted so that the impact of different features on the overall deviation can be adjusted. The parameter is used to adjust the contribution of each feature to the total deviation. The parameters are used to adjust the influence of each characteristic on the total deviation. The speed characteristic deviation calculation formula The difference in the velocity characteristics of the two trajectory segments is calculated. This part quantifies the overall difference in the velocity characteristics by integrating the velocity. The parameter is used to adjust the contribution of the integral term to the total deviation. The parameter is used to adjust the degree of influence of the integral term on the total deviation. The overall parameter of the formula , The calculation method of position deviation is controlled, allowing different features to contribute and influence the overall deviation to different degrees. , The calculation method of speed characteristic deviation is controlled to allow different contributions and influences of speed characteristics on the overall deviation. When using the conventional trajectory deviation calculation formula in the field, the first trajectory segment can be obtained and the second trajectory segment By applying the trajectory deviation calculation formula provided by the present invention, the first trajectory segment can be calculated more accurately. and the second trajectory segment The formula comprehensively considers the differences in position and speed characteristics of trajectory segments, and comprehensively considers the overall deviation of the trajectory. This comprehensiveness can more comprehensively describe the similarities or differences between trajectories, providing a more comprehensive basis for comparison. The formula can be applied to different types of trajectory data, including but not limited to motion trajectories, path planning, vehicle trajectories, etc. Its wide applicability allows the formula to be applied to multiple fields, including traffic management, geographic information systems, mobile applications, etc. By quantifying the differences in trajectory segments into numerical values, the formula provides a quantifiable way to compare and evaluate the degree of difference between trajectories, which helps to perform more accurate trajectory matching, classification, or anomaly detection.
优选的，步骤S3包括以下步骤：Preferably, step S3 comprises the following steps:
步骤S31：利用车载摄像头获取运输环境感知数据；Step S31: using the vehicle-mounted camera to obtain transportation environment perception data;
步骤S32：基于运输环境感知数据对异常轨迹片段进行异常运输行为识别，生成异常运输行为数据；Step S32: identifying abnormal transportation behaviors of abnormal trajectory segments based on the transportation environment perception data, and generating abnormal transportation behavior data;
步骤S33：对异常运输行为数据进行异常模式分类，生成异常运输模式分类数据；Step S33: classifying the abnormal transport behavior data into abnormal patterns to generate abnormal transport pattern classification data;
步骤S34：对异常运输模式分类数据进行异常标签处理，生成异常运输轨迹分类标签。Step S34: performing abnormal label processing on the abnormal transportation mode classification data to generate abnormal transportation trajectory classification labels.
本发明通过车载摄像头获取的运输环境感知数据，系统可以识别和捕获异常的运输轨迹片段，有助于及时发现和识别出存在问题或异常行为的轨迹片段。基于获取的异常轨迹片段，系统进行异常运输行为的识别，异常行为可以包括车辆违规驾驶、交通事故、路况异常等情况，通过识别这些异常行为，可以帮助监管部门或企业管理者及时采取措施，减少事故风险。对识别出的异常运输行为数据进行分类和归纳，形成异常运输模式分类数据，有助于系统进一步分析和理解不同类型的异常行为，为后续的处理和管理提供更有针对性的信息。对异常运输模式分类数据进行异常标签处理，生成异常运输轨迹分类标签，标签可以用于对异常轨迹进行分类和归纳，便于后续的数据分析和管理。The present invention uses the transportation environment perception data acquired by the vehicle-mounted camera, and the system can identify and capture abnormal transportation trajectory fragments, which helps to timely discover and identify trajectory fragments with problems or abnormal behaviors. Based on the acquired abnormal trajectory fragments, the system identifies abnormal transportation behaviors. Abnormal behaviors may include vehicle illegal driving, traffic accidents, abnormal road conditions, etc. By identifying these abnormal behaviors, it can help regulatory authorities or corporate managers take timely measures to reduce accident risks. Classifying and summarizing the identified abnormal transportation behavior data to form abnormal transportation mode classification data helps the system further analyze and understand different types of abnormal behaviors, and provide more targeted information for subsequent processing and management. Abnormal label processing is performed on the abnormal transportation mode classification data to generate abnormal transportation trajectory classification labels. The labels can be used to classify and summarize abnormal trajectories, which is convenient for subsequent data analysis and management.
作为本发明的一个实例，参考图3所示，在本实例中所述步骤S3包括：As an example of the present invention, referring to FIG. 3 , in this example, step S3 includes:
步骤S31：利用车载摄像头获取运输环境感知数据；Step S31: using the vehicle-mounted camera to obtain transportation environment perception data;
本发明实施例中，通过选择在车辆上合适的位置安装摄像头，以确保能够有效地捕捉车辆周围的运输环境情况。常见的安装位置包括车辆前部、后部、侧面和车厢内部等位置。配置摄像头的参数，如分辨率、帧率、曝光等，以获得清晰、稳定的图像。确保摄像头能够适应不同光照条件和环境变化，保证在各种情况下都能够获取可靠的数据。摄像头应能够实时捕捉周围环境的图像数据，并传输至处理系统，涉及到视频流的处理和传输，要确保传输的效率和稳定性，以便系统能够及时地获取和处理数据。对从摄像头获取的图像数据进行处理和分析，以提取有用的信息和特征，包括目标检测、车辆识别、道路标志识别、行人识别等任务，以获得关于运输环境的感知数据。In an embodiment of the present invention, a camera is installed at a suitable position on the vehicle to ensure that the transportation environment around the vehicle can be effectively captured. Common installation locations include the front, rear, side and interior of the vehicle. Configure the parameters of the camera, such as resolution, frame rate, exposure, etc., to obtain clear and stable images. Ensure that the camera can adapt to different lighting conditions and environmental changes to ensure that reliable data can be obtained under various circumstances. The camera should be able to capture image data of the surrounding environment in real time and transmit it to the processing system, which involves the processing and transmission of the video stream. The efficiency and stability of the transmission must be ensured so that the system can obtain and process data in a timely manner. The image data obtained from the camera is processed and analyzed to extract useful information and features, including tasks such as target detection, vehicle recognition, road sign recognition, pedestrian recognition, etc., to obtain perception data about the transportation environment.
步骤S32：基于运输环境感知数据对异常轨迹片段进行异常运输行为识别，生成异常运输行为数据；Step S32: identifying abnormal transportation behaviors of abnormal trajectory segments based on the transportation environment perception data, and generating abnormal transportation behavior data;
本发明实施例中，通过从感知数据中提取特征，特征具体是关于车辆、行人、道路、交通标志等方面的信息，用于描述环境和运输行为的特征。特征提取具体使用传统的计算机视觉方法，也可以使用深度学习技术，例如卷积神经网络（CNN）等。使用机器学习算法或深度学习模型对提取的特征进行分析，以检测异常的运输行为，涉及到分类、聚类、异常检测等技术，以区分正常行为和异常行为。识别出的异常行为需要进行标记，以便后续的处理和分析。标记具体基于时间戳、空间位置等信息，将异常行为与具体的轨迹片段关联起来。In an embodiment of the present invention, features are extracted from the perception data, and the features are specifically information about vehicles, pedestrians, roads, traffic signs, etc., which are used to describe the characteristics of the environment and transportation behavior. Feature extraction specifically uses traditional computer vision methods, and deep learning techniques such as convolutional neural networks (CNNs) may also be used. The extracted features are analyzed using machine learning algorithms or deep learning models to detect abnormal transportation behaviors, involving classification, clustering, anomaly detection and other technologies to distinguish between normal and abnormal behaviors. The identified abnormal behaviors need to be marked for subsequent processing and analysis. The marking is specifically based on information such as timestamps and spatial locations to associate abnormal behaviors with specific trajectory segments.
步骤S33：对异常运输行为数据进行异常模式分类，生成异常运输模式分类数据；Step S33: classifying the abnormal transport behavior data into abnormal patterns to generate abnormal transport pattern classification data;
本发明实施例中，通过从异常运输行为数据中提取特征，特征具体是关于行为类型、行为持续时间、行为频率、行为时段等方面的信息。特征工程的过程需要根据具体情况设计并应用不同的特征提取方法和技术。选择合适的异常模式分类算法，如聚类算法、分类算法等，用于对异常运输行为数据进行分类。常用的算法包括K均值聚类、层次聚类、支持向量机（SVM）、决策树、神经网络等。使用选定的算法对准备好的特征进行模型训练。需要进行模型参数调优、交叉验证等步骤来优化模型的性能。使用训练好的模型对异常运输行为数据进行分类，将其归类为不同的异常模式。分类的结果包括各种异常行为的类别标签，以及每个类别的概率或置信度。对分类结果进行评估，包括准确率、召回率、精确度等指标的评估。分析分类结果，解释每个异常模式的特征和规律，为后续的异常轨迹分类标签生成提供参考。In the embodiment of the present invention, features are extracted from abnormal transportation behavior data, and the features are specifically information about behavior type, behavior duration, behavior frequency, behavior time period, etc. The process of feature engineering needs to design and apply different feature extraction methods and techniques according to specific circumstances. Select a suitable abnormal pattern classification algorithm, such as a clustering algorithm, a classification algorithm, etc., to classify abnormal transportation behavior data. Common algorithms include K-means clustering, hierarchical clustering, support vector machine (SVM), decision tree, neural network, etc. Use the selected algorithm to train the model for the prepared features. Model parameter tuning, cross-validation and other steps are required to optimize the performance of the model. Use the trained model to classify the abnormal transportation behavior data and classify it into different abnormal patterns. The classification results include category labels for various abnormal behaviors, as well as the probability or confidence of each category. Evaluate the classification results, including the evaluation of indicators such as accuracy, recall, and precision. Analyze the classification results, explain the characteristics and laws of each abnormal pattern, and provide a reference for the subsequent generation of abnormal trajectory classification labels.
步骤S34：对异常运输模式分类数据进行异常标签处理，生成异常运输轨迹分类标签。Step S34: performing abnormal label processing on the abnormal transportation mode classification data to generate abnormal transportation trajectory classification labels.
本发明实施例中，通过根据步骤S33生成的异常运输模式分类数据，为每个异常模式分配一个标签或类别，标签具体是预先定义的，也可以根据具体的分类结果进行动态生成。将异常运输模式分类数据中的异常模式标签与原始轨迹数据进行关联，以便将异常行为标记应用到具体的轨迹片段上，涉及到时间戳或空间位置等信息的匹配和关联。根据关联的异常模式标签，将异常运输行为的标记应用到相应的轨迹片段上，具体通过在轨迹数据中添加额外的属性或标记来实现。对于同一段轨迹存在多个异常模式标签的情况，需要进行标签整合和冲突解决，具体根据一定的优先级或规则对多个异常标签进行合并或选择。In an embodiment of the present invention, a label or category is assigned to each abnormal mode according to the abnormal transportation mode classification data generated in step S33. The label is specifically pre-defined, and can also be dynamically generated according to the specific classification results. The abnormal mode label in the abnormal transportation mode classification data is associated with the original trajectory data so that the abnormal behavior mark is applied to the specific trajectory segment, which involves the matching and association of information such as timestamps or spatial positions. According to the associated abnormal mode label, the mark of the abnormal transportation behavior is applied to the corresponding trajectory segment, which is specifically achieved by adding additional attributes or tags to the trajectory data. In the case where there are multiple abnormal mode labels for the same trajectory, label integration and conflict resolution are required, and multiple abnormal labels are specifically merged or selected according to certain priorities or rules.
优选的，步骤S32包括以下步骤：Preferably, step S32 includes the following steps:
步骤S321：对运输环境感知数据进行生物感知标记，生成运输环境生物数据；Step S321: biometrically marking the transport environment sensing data to generate transport environment biological data;
步骤S322：对运输环境感知数据进行非生物感知标记，生成运输环境非生物数据；Step S322: performing non-biological perception marking on the transport environment perception data to generate transport environment non-biological data;
步骤S323：对异常轨迹片段进行运输车辆位置感知，生成运输车辆位置信息数据；通过运输环境感知数据对运输车辆位置信息数据进行感知范围分析，生成运输车辆感知范围数据；Step S323: sensing the position of the transport vehicle for the abnormal trajectory segment to generate transport vehicle position information data; performing a sensing range analysis on the transport vehicle position information data through the transport environment sensing data to generate transport vehicle sensing range data;
步骤S324：利用范围入侵检测公式根据运输环境生物数据和运输环境非生物数据对运输车辆感知范围数据进行入侵检测，生成范围生物入侵检测值和范围非生物入侵检测值；Step S324: using a range intrusion detection formula to perform intrusion detection on the sensing range data of the transport vehicle according to the transport environment biological data and the transport environment non-biological data, and generating a range biological intrusion detection value and a range non-biological intrusion detection value;
步骤S325：对范围生物入侵检测值进行第一异常行为标记，生成第一异常运输行为数据；对范围非生物入侵检测值进行第二异常行为标记，生成第二异常运输行为数据；将第一异常运输行为数据和第二异常运输行为数据进行数据整合，生成异常运输行为数据；Step S325: performing a first abnormal behavior mark on the range biological invasion detection value to generate first abnormal transportation behavior data; performing a second abnormal behavior mark on the range non-biological invasion detection value to generate second abnormal transportation behavior data; integrating the first abnormal transportation behavior data and the second abnormal transportation behavior data to generate abnormal transportation behavior data;
其中范围入侵检测公式具体如下：The range intrusion detection formula is as follows:
； ;
式中，表示为范围入侵检测值，表示为检测时间上限，表示为生物数据权重参数，表示为生物数据样本数量，表示为第个生物数据样本的权重，表示为第个生物数据样本的检测结果，表示为非生物数据权重参数，表示为非生物数据样本数量，表示为第个非生物数据样本的权重，表示为第个非生物数据样本的检测结果，表示为非生物数据非线性调节参数。In the formula, Expressed as a range of intrusion detection values, It represents the upper limit of the detection time. Expressed as biological data weight parameter, Expressed as the number of biological data samples, Expressed as The weight of biological data samples, Expressed as The test results of biological data samples, Expressed as the non-biological data weight parameter, Expressed as the number of non-biological data samples, Expressed as The weight of non-biological data samples, Expressed as The test results of non-biological data samples, Represented as a non-linear adjustment parameter for non-biological data.
本发明通过将运输环境感知数据进行生物和非生物感知标记，可以将环境中的不同类型信息进行区分和分类，有助于更精确地识别异常运输行为。将环境中的生物感知数据和非生物感知数据结合起来进行异常行为识别，可以提供更全面的信息，从而使得异常行为的识别更加准确和可靠。结合生物感知数据和非生物感知数据进行异常行为识别，可以融合多维度的信息，包括生物特征（如行人、动物等）和非生物特征（如车辆、路况等），有助于提高识别的准确性和鲁棒性。通过对生物和非生物感知数据的同时考虑，可以更全面地感知和理解运输环境，从而更好地发现和识别异常情况，提高了系统对环境的感知能力。结合生物和非生物感知数据进行异常行为识别，可以使系统更好地适应各种复杂的交通和环境情况，从而提高了系统对多样化场景的应对能力，其中入侵检测包括对运输环境感知数据进行生物感知标记（如行人识别）和非生物感知标记（如障碍物检测），以识别潜在威胁。当检测到未经授权的进入或潜在危险时，系统会触发预警机制，通知驾驶员和监管平台，从而防止潜在安全事故，保障大件运输过程的安全。The present invention distinguishes and classifies different types of information in the environment by marking the transport environment perception data with biological and non-biological perception data, which helps to more accurately identify abnormal transport behaviors. Combining the biological perception data and non-biological perception data in the environment for abnormal behavior identification can provide more comprehensive information, thereby making the identification of abnormal behavior more accurate and reliable. Combining biological perception data and non-biological perception data for abnormal behavior identification can integrate multi-dimensional information, including biological features (such as pedestrians, animals, etc.) and non-biological features (such as vehicles, road conditions, etc.), which helps to improve the accuracy and robustness of identification. By considering both biological and non-biological perception data, the transport environment can be more comprehensively perceived and understood, so as to better discover and identify abnormal situations and improve the system's ability to perceive the environment. Combining biological and non-biological perception data for abnormal behavior identification can make the system better adapt to various complex traffic and environmental situations, thereby improving the system's ability to respond to diverse scenarios, wherein intrusion detection includes biological perception marking (such as pedestrian identification) and non-biological perception marking (such as obstacle detection) of the transport environment perception data to identify potential threats. When unauthorized entry or potential danger is detected, the system will trigger an early warning mechanism to notify the driver and the supervision platform, thereby preventing potential safety accidents and ensuring the safety of the large-scale transportation process.
作为本发明的一个实例，参考图4所示，在本实例中所述步骤S32包括：As an example of the present invention, referring to FIG. 4 , in this example, step S32 includes:
步骤S321：对运输环境感知数据进行生物感知标记，生成运输环境生物数据；Step S321: biometrically marking the transport environment sensing data to generate transport environment biological data;
本发明实施例中，通过使用车载摄像头等传感器设备获取运输环境感知数据，包括视频图像、雷达数据、红外线图像等。对获取的感知数据进行生物感知特征提取，例如人类行人的检测、动物的识别等。生物感知特征具体包括形态、颜色、运动轨迹等方面的信息。对提取的生物感知特征进行标记或分类，以标识出感知数据中的生物信息，具体通过目标检测、实例分割等计算机视觉技术来实现。In the embodiment of the present invention, sensor devices such as vehicle-mounted cameras are used to obtain transportation environment perception data, including video images, radar data, infrared images, etc. The acquired perception data is subjected to biological perception feature extraction, such as human pedestrian detection and animal recognition. The biological perception features specifically include information on morphology, color, motion trajectory, etc. The extracted biological perception features are marked or classified to identify biological information in the perception data, which is specifically achieved through computer vision technologies such as target detection and instance segmentation.
本发明实施例中，更为重要地是，对运输环境感知数据进行相邻视频帧比较，得到运输环境运动物体数据；根据运输环境运动物体数据进行生物感知标记，生成运输环境生物数据；In the embodiment of the present invention, more importantly, the transport environment perception data is compared with adjacent video frames to obtain the transport environment moving object data; biological perception marking is performed according to the transport environment moving object data to generate the transport environment biological data;
其中运输环境运动物体数据包括第一运输环境运动物体数据以及第二运输环境运动物体数据，相邻视频帧比较的步骤包括：The transport environment moving object data includes first transport environment moving object data and second transport environment moving object data, and the step of comparing adjacent video frames includes:
对运输环境感知数据中的当前运输环境感知数据以及上一帧运输环境感知数据进行卷积计算，分别得到当前运输环境感知卷积层数据以及上一帧运输环境感知卷积层数据；对当前运输环境感知卷积层数据以及上一帧运输环境感知卷积层数据进行自注意力加权计算，分别得到当前运输环境感知自注意力加权数据以及上一帧运输环境感知自注意力加权数据；对当前运输环境感知自注意力加权数据以及上一帧运输环境感知自注意力加权数据进行聚类计算，分别得到当前运输环境感知聚类数据以及上一帧运输环境感知聚类数据；对当前运输环境感知聚类数据以及上一帧运输环境感知聚类数据进行相似性计算，得到运输环境感知相似性数据；根据运输环境感知相似性数据对运输环境感知数据中的当前运输环境感知数据以及上一帧运输环境感知数据进行匹配，得到运输环境感知匹配数据；根据运输环境感知匹配数据进行结构变化计算，得到运输环境感知结构变化数据；确定运输环境感知结构变化数据大于或等于预设的运输环境感知结构变化阈值数据时，将运输环境感知结构变化数据对应的运输环境感知数据（既运输环境感知结构变化数据对应的运输环境感知数据中的区域数据）确定为第一运输环境运动物体数据；其中结构变化计算通过结构变化计算公式进行计算，结构变化计算公式具体为：Perform convolution calculation on the current transport environment perception data and the previous frame of transport environment perception data in the transport environment perception data to obtain the current transport environment perception convolution layer data and the previous frame of transport environment perception convolution layer data respectively; perform self-attention weighted calculation on the current transport environment perception convolution layer data and the previous frame of transport environment perception convolution layer data to obtain the current transport environment perception self-attention weighted data and the previous frame of transport environment perception self-attention weighted data respectively; perform clustering calculation on the current transport environment perception self-attention weighted data and the previous frame of transport environment perception self-attention weighted data to obtain the current transport environment perception clustering data and the previous frame of transport environment perception clustering data respectively; perform similarity calculation on the current transport environment perception clustering data and the previous frame of transport environment perception clustering data The method comprises the following steps: performing a structure change calculation based on the transportation environment perception matching data to obtain the transportation environment perception structure change data; performing a structure change calculation based on the transportation environment perception matching data to obtain the transportation environment perception structure change data; and determining the transportation environment perception structure change data corresponding to the transportation environment perception structure change data (i.e., the area data in the transportation environment perception data corresponding to the transportation environment perception structure change data) as the first transportation environment moving object data; wherein the structure change calculation is performed by a structure change calculation formula, and the structure change calculation formula is specifically as follows:
； ;
为运输环境感知结构变化数据，为匹配感知点序次项，为匹配感知点数量项，为运输环境感知匹配数据，为第时刻运输环境感知匹配数据对应的第个匹配感知点数据，为第时刻运输环境感知匹配数据对应的第个匹配感知点数据，为微小变化项，为运输环境运动频率数据，为调整因子数据; To sense structural change data for the transportation environment, To match the order of perception points, is the number of matching perception points, Perceive and match data for transportation environment, For the The transportation environment perception matching data corresponding to the Matching perception point data, For the The transportation environment perception matching data corresponding to the Matching perception point data, is a small change term, For the transport environment movement frequency data, is the adjustment factor data;
获取车辆行驶速度数据以及车辆行驶偏移角度数据；根据车辆行驶速度数据、车辆行驶偏移角度数据以及预设的运输环境感知数据对应的帧采集参数数据进行图像偏移尺度计算，得到图像偏移尺度数据，其中图像偏移尺度数据包括图像像素偏移数据以及图像像素变形数据；根据图像偏移尺度数据对运输环境感知数据中的当前运输环境感知数据进行变形处理，得到当前运输环境感知变形数据；将当前运输环境感知变形数据与上一帧运输环境感知数据进行相减，得到运输环境感知变化数据；对运输环境感知变化数据进行数据复杂率计算，得到运输环境感知变化复杂率数据；根据运输环境感知变化复杂率数据生成第二运输环境运动物体数据；Acquire vehicle travel speed data and vehicle travel offset angle data; perform image offset scale calculation according to the vehicle travel speed data, the vehicle travel offset angle data and the frame acquisition parameter data corresponding to the preset transport environment perception data to obtain image offset scale data, wherein the image offset scale data includes image pixel offset data and image pixel deformation data; perform deformation processing on the current transport environment perception data in the transport environment perception data according to the image offset scale data to obtain current transport environment perception deformation data; subtract the current transport environment perception deformation data from the previous frame of transport environment perception data to obtain transport environment perception change data; perform data complexity rate calculation on the transport environment perception change data to obtain transport environment perception change complexity rate data; generate second transport environment moving object data according to the transport environment perception change complexity rate data;
其中数据复杂率计算具体为：对运输环境感知变化数据进行层次聚类计算，得到运输环境感知层次聚类数据；根据运输环境感知层次聚类数据以及运输环境感知数据进行分割处理，得到运输环境感知区域数据；对运输环境感知区域数据进行数据复杂率计算，得到运输环境感知变化复杂率数据。The data complexity rate calculation is specifically as follows: perform hierarchical clustering calculation on the transport environment perception change data to obtain the transport environment perception hierarchical clustering data; perform segmentation processing on the transport environment perception hierarchical clustering data and the transport environment perception data to obtain the transport environment perception area data; perform data complexity rate calculation on the transport environment perception area data to obtain the transport environment perception change complexity rate data.
步骤S322：对运输环境感知数据进行非生物感知标记，生成运输环境非生物数据；Step S322: performing non-biological perception marking on the transport environment perception data to generate transport environment non-biological data;
本发明实施例中，通过使用车载摄像头、激光雷达、超声波传感器等设备获取运输环境感知数据，包括视频图像、雷达数据、距离传感数据等。对获取的感知数据进行非生物感知特征提取，例如车辆的检测、道路的识别、交通标志的检测等。非生物感知特征具体包括形状、颜色、纹理、位置等方面的信息。对提取的非生物感知特征进行标记或分类，以标识出感知数据中的非生物信息，具体通过目标检测、图像分割、图像分类等计算机视觉技术来实现。In the embodiment of the present invention, the transport environment perception data including video images, radar data, distance sensor data, etc. are obtained by using on-board cameras, laser radars, ultrasonic sensors and other devices. Non-biological perception features are extracted from the acquired perception data, such as vehicle detection, road recognition, traffic sign detection, etc. Non-biological perception features specifically include information on shape, color, texture, position, etc. The extracted non-biological perception features are marked or classified to identify the non-biological information in the perception data, which is specifically achieved through computer vision technologies such as target detection, image segmentation, and image classification.
步骤S323：对异常轨迹片段进行运输车辆位置感知，生成运输车辆位置信息数据；通过运输环境感知数据对运输车辆位置信息数据进行感知范围分析，生成运输车辆感知范围数据；Step S323: sensing the position of the transport vehicle for the abnormal trajectory segment to generate transport vehicle position information data; performing a sensing range analysis on the transport vehicle position information data through the transport environment sensing data to generate transport vehicle sensing range data;
本发明实施例中，通过使用相应的轨迹分析算法或者机器学习模型对运输车辆的轨迹数据进行分析，识别出异常轨迹片段，比如突然的加速、减速、不规律的路线等。针对异常轨迹片段所在的位置，使用车载传感器、GPS等设备感知运输车辆的实时位置，生成相应的位置信息数据。获取运输环境感知数据，这包括道路交通状况、天气信息、道路条件等。这些数据可以通过传感器、交通监控系统、天气预报等渠道获取。将运输车辆的位置信息数据与运输环境感知数据进行关联，分析异常轨迹片段所在位置的运输环境情况。基于运输环境感知数据，对运输车辆的位置信息数据进行感知范围分析，包括确定异常轨迹片段的影响范围，以及潜在的危险区域。将感知范围分析的结果整理成数据形式，生成运输车辆的感知范围数据。In an embodiment of the present invention, the trajectory data of the transport vehicle is analyzed by using a corresponding trajectory analysis algorithm or a machine learning model to identify abnormal trajectory segments, such as sudden acceleration, deceleration, irregular routes, etc. For the location of the abnormal trajectory segment, the real-time location of the transport vehicle is sensed using on-board sensors, GPS and other devices to generate corresponding location information data. The transport environment perception data is obtained, which includes road traffic conditions, weather information, road conditions, etc. These data can be obtained through sensors, traffic monitoring systems, weather forecasts and other channels. The location information data of the transport vehicle is associated with the transportation environment perception data to analyze the transportation environment conditions at the location of the abnormal trajectory segment. Based on the transportation environment perception data, the location information data of the transport vehicle is subjected to a perception range analysis, including determining the impact range of the abnormal trajectory segment and the potential danger zone. The results of the perception range analysis are organized into data form to generate the perception range data of the transport vehicle.
步骤S324：利用范围入侵检测公式根据运输环境生物数据和运输环境非生物数据对运输车辆感知范围数据进行入侵检测，生成范围生物入侵检测值和范围非生物入侵检测值；Step S324: using a range intrusion detection formula to perform intrusion detection on the sensing range data of the transport vehicle according to the transport environment biological data and the transport environment non-biological data, and generating a range biological intrusion detection value and a range non-biological intrusion detection value;
本发明实施例中，通过根据生物数据和非生物数据，制定范围入侵检测的公式。这个公式具体是基于统计模型、机器学习算法或者专家经验规则，用于评估运输车辆感知范围内生物和非生物环境的变化情况，从而确定是否存在入侵情况。将步骤S323中生成的运输车辆感知范围数据作为输入，结合步骤1中获取的生物数据和非生物数据，准备进行入侵检测。根据制定的范围入侵检测公式，对运输车辆感知范围内的生物和非生物数据进行计算，得到范围生物入侵检测值和范围非生物入侵检测值。根据计算得到的入侵检测值，判断是否存在入侵情况。如果范围生物入侵检测值或范围非生物入侵检测值超出了设定的阈值，则认为存在入侵，并生成相应的入侵警报或记录。In an embodiment of the present invention, a formula for range intrusion detection is formulated based on biological data and non-biological data. This formula is specifically based on a statistical model, a machine learning algorithm, or an expert experience rule, and is used to evaluate changes in the biological and non-biological environments within the perception range of the transport vehicle, thereby determining whether there is an invasion. The perception range data of the transport vehicle generated in step S323 is used as input, and combined with the biological data and non-biological data obtained in step 1, intrusion detection is prepared. According to the formulated range intrusion detection formula, the biological and non-biological data within the perception range of the transport vehicle are calculated to obtain a range biological invasion detection value and a range non-biological invasion detection value. Based on the calculated invasion detection value, it is determined whether there is an invasion. If the range biological invasion detection value or the range non-biological invasion detection value exceeds the set threshold, it is considered that an invasion exists, and a corresponding intrusion alarm or record is generated.
步骤S325：对范围生物入侵检测值进行第一异常行为标记，生成第一异常运输行为数据；对范围非生物入侵检测值进行第二异常行为标记，生成第二异常运输行为数据；将第一异常运输行为数据和第二异常运输行为数据进行数据整合，生成异常运输行为数据；Step S325: performing a first abnormal behavior mark on the range biological invasion detection value to generate first abnormal transportation behavior data; performing a second abnormal behavior mark on the range non-biological invasion detection value to generate second abnormal transportation behavior data; integrating the first abnormal transportation behavior data and the second abnormal transportation behavior data to generate abnormal transportation behavior data;
本发明实施例中，通过根据范围生物入侵检测值，制定异常行为标记规则。这可以是基于阈值的方法，当生物入侵检测值超出预设的阈值时，标记为异常行为。另外，也可以采用机器学习算法对生物入侵检测值进行分析，识别出异常模式。根据标记的异常行为，将异常的生物入侵检测值及其相关信息整合成第一异常运输行为数据，包括异常发生的时间、位置、生物入侵类型等。同样地，根据范围非生物入侵检测值，制定异常行为标记规则，例如基于阈值或者机器学习算法。根据标记的异常行为，将异常的非生物入侵检测值及其相关信息整合成第二异常运输行为数据，包括异常发生的时间、位置、非生物入侵类型等。将第一异常运输行为数据和第二异常运输行为数据进行整合，可以按照时间顺序或者位置进行合并，确保数据的完整性和一致性。In an embodiment of the present invention, an abnormal behavior marking rule is formulated based on a range of biological invasion detection values. This can be a threshold-based method. When the biological invasion detection value exceeds a preset threshold, it is marked as abnormal behavior. In addition, a machine learning algorithm can also be used to analyze the biological invasion detection value to identify abnormal patterns. According to the marked abnormal behavior, the abnormal biological invasion detection value and its related information are integrated into the first abnormal transportation behavior data, including the time, location, type of biological invasion, etc. when the abnormality occurs. Similarly, according to the range of non-biological invasion detection values, an abnormal behavior marking rule is formulated, for example, based on a threshold or a machine learning algorithm. According to the marked abnormal behavior, the abnormal non-biological invasion detection value and its related information are integrated into the second abnormal transportation behavior data, including the time, location, type of non-biological invasion, etc. when the abnormality occurs. The first abnormal transportation behavior data and the second abnormal transportation behavior data are integrated, and they can be merged in chronological order or position to ensure the integrity and consistency of the data.
其中范围入侵检测公式具体如下：The range intrusion detection formula is as follows:
； ;
式中，表示为范围入侵检测值，表示为检测时间上限，表示为生物数据权重参数，表示为生物数据样本数量，表示为第个生物数据样本的权重，表示为第个生物数据样本的检测结果，表示为非生物数据权重参数，表示为非生物数据样本数量，表示为第个非生物数据样本的权重，表示为第个非生物数据样本的检测结果，表示为非生物数据非线性调节参数。In the formula, Expressed as a range of intrusion detection values, It represents the upper limit of the detection time. Expressed as biological data weight parameter, Expressed as the number of biological data samples, Expressed as The weight of biological data samples, Expressed as The test results of biological data samples, Expressed as the non-biological data weight parameter, Expressed as the number of non-biological data samples, Expressed as The weight of non-biological data samples, Expressed as The test results of non-biological data samples, Represented as a non-linear adjustment parameter for non-biological data.
本发明通过分析并整合了一种范围入侵检测公式，公式结合了运输环境的生物数据和非生物数据，综合考虑了不同类型的信息。这种综合多维数据的方法可以提高入侵检测的准确性和可靠性，因为不同类型的数据可以提供互补的信息。公式中的参数（如权重参数和非线性调节参数）可以根据具体需求进行调节。通过调节这些参数，可以灵活地调整不同数据类型在入侵检测中的重要性和对结果的影响程度，以适应不同场景和需求。公式通过时间积分考虑了检测时间范围内的数据，而不仅仅是单个时刻的数据。这种积分时间范围的方法可以提供对一段时间内整体入侵情况的评估，更全面地了解潜在的入侵行为。公式中的非线性调节项（ 和）可以使入侵检测的响应更加灵敏和可调。通过非线性调节，可以更好地捕捉到数据变化的细节和非线性关系，提高入侵检测的敏感性和适应性。在使用本领域常规的范围入侵检测公式时，可以得到范围入侵检测值，通过应用本发明提供的范围入侵检测公式，可以更加精确的计算出范围入侵检测值。这个复杂的范围入侵检测公式通过综合多维数据、参数调节能力、积分时间范围和非线性调节等机制，可以提高入侵检测的准确性、灵敏性和适应性，帮助有效地监测和识别潜在的运输车辆入侵行为。The present invention analyzes and integrates a range intrusion detection formula that combines biological data and non-biological data of the transportation environment and comprehensively considers different types of information. This method of integrating multi-dimensional data can improve the accuracy and reliability of intrusion detection because different types of data can provide complementary information. The parameters in the formula (such as weight parameters and non-linear adjustment parameters) can be adjusted according to specific needs. By adjusting these parameters, the importance of different data types in intrusion detection and the degree of influence on the results can be flexibly adjusted to adapt to different scenarios and needs. The formula considers the data within the detection time range through time integration, not just the data at a single moment. This method of integrating the time range can provide an assessment of the overall intrusion situation over a period of time and a more comprehensive understanding of potential intrusion behaviors. The non-linear adjustment term in the formula ( and ) can make the response of intrusion detection more sensitive and adjustable. Through nonlinear adjustment, the details of data changes and nonlinear relationships can be better captured, thereby improving the sensitivity and adaptability of intrusion detection. When using the conventional range intrusion detection formula in the field, a range intrusion detection value can be obtained. By applying the range intrusion detection formula provided by the present invention, the range intrusion detection value can be calculated more accurately. This complex range intrusion detection formula can improve the accuracy, sensitivity and adaptability of intrusion detection by integrating multidimensional data, parameter adjustment capability, integral time range and nonlinear adjustment mechanisms, thereby helping to effectively monitor and identify potential transport vehicle intrusion behaviors.
优选的，步骤S4包括以下步骤：Preferably, step S4 comprises the following steps:
步骤S41：基于异常运输行为数据进行时空关联分析，生成异常运输行为时空关联数据；Step S41: performing spatiotemporal correlation analysis based on abnormal transportation behavior data to generate spatiotemporal correlation data of abnormal transportation behavior;
步骤S42：对异常运输行为时空关联数据进行数据集划分，生成模型训练集和模型测试集；利用支持向量机算法对模型训练集进行模型训练，生成异常行驶轨迹训练模型；通过模型测试集对行驶轨迹异常训练模型进行模型优化迭代，从而生成异常行驶轨迹预测模型；Step S42: dividing the data set of the spatiotemporal correlation data of abnormal transportation behavior to generate a model training set and a model test set; using the support vector machine algorithm to train the model training set to generate an abnormal driving trajectory training model; using the model test set to perform model optimization iteration on the driving trajectory abnormality training model to generate an abnormal driving trajectory prediction model;
步骤S43：将异常运输行为时空关联数据导入至异常行驶轨迹异常检测模型中进行异常行驶轨迹预测，生成异常行驶轨迹预测数据。Step S43: Importing the spatiotemporal correlation data of abnormal transportation behavior into the abnormal driving trajectory anomaly detection model to predict the abnormal driving trajectory and generate abnormal driving trajectory prediction data.
本发明通过时空关联分析，可以发现异常运输行为在时间和空间上的分布规律和关联性，有助于更深入地理解异常行为的发生机制，为后续的模型训练和预测提供了重要参考。利用异常运输行为时空关联数据，可以划分出模型训练集和测试集，从而进行模型的训练和测试。使用支持向量机等算法对训练集进行模型训练，并通过测试集对模型进行评估和优化，以生成更准确的异常行驶轨迹预测模型。将异常运输行为时空关联数据导入到训练好的异常行驶轨迹预测模型中，可以实现对异常行驶轨迹的预测，有助于及时发现潜在的异常轨迹，并采取措施避免事故发生或加强监控。通过预测异常行驶轨迹，可以在异常行驶发生之前采取预防性措施，减少事故发生的可能性，有助于提高运输安全性，降低事故损失，保护人员和财产的安全。通过对异常行驶轨迹进行预测和及时干预，可以提高运输管理的效率和准确性，优化交通流量，减少拥堵情况的发生。The present invention can find the distribution law and correlation of abnormal transportation behavior in time and space through spatiotemporal correlation analysis, which is helpful to understand the occurrence mechanism of abnormal behavior more deeply and provide an important reference for subsequent model training and prediction. Using the spatiotemporal correlation data of abnormal transportation behavior, the model training set and the test set can be divided to train and test the model. The training set is trained by using algorithms such as support vector machines, and the model is evaluated and optimized by the test set to generate a more accurate abnormal driving trajectory prediction model. The spatiotemporal correlation data of abnormal transportation behavior is imported into the trained abnormal driving trajectory prediction model, which can realize the prediction of abnormal driving trajectory, help to discover potential abnormal trajectories in time, and take measures to avoid accidents or strengthen monitoring. By predicting abnormal driving trajectories, preventive measures can be taken before abnormal driving occurs, reduce the possibility of accidents, help improve transportation safety, reduce accident losses, and protect the safety of people and property. By predicting and timely intervening in abnormal driving trajectories, the efficiency and accuracy of transportation management can be improved, traffic flow can be optimized, and congestion can be reduced.
本发明实施例中，通过整理异常运输行为数据，并根据时间和空间信息对数据进行排序和组织。使用统计分析或地理信息系统（GIS）技术，对异常运输行为数据进行时空关联分析，具体包括时间序列分析、空间模式识别、热力图生成等方法，以发现异常行为的时空分布规律和关联性。将时空关联分析的结果可视化呈现，例如绘制时空分布图、热力图等，以便直观地理解异常行为在时空上的分布特征。将时空关联数据划分为模型训练集和模型测试集，通常采用交叉验证或留出法等方法。使用支持向量机（Support VectorMachine，SVM）算法对模型训练集进行训练。SVM是一种监督学习算法，适用于分类和回归任务。通过模型测试集对训练好的模型进行评估，并根据评估结果对模型进行优化迭代，以提高模型的准确性和泛化能力。将异常运输行为时空关联数据导入到训练好的异常行驶轨迹异常检测模型中，进行异常行驶轨迹的预测。使用训练好的模型对异常行驶轨迹进行预测，以识别出现异常情况的轨迹。生成异常行驶轨迹预测数据，并根据需要进行数据可视化或进一步的处理和分析。In an embodiment of the present invention, the abnormal transportation behavior data is sorted and the data is sorted and organized according to time and space information. Statistical analysis or geographic information system (GIS) technology is used to perform spatiotemporal correlation analysis on the abnormal transportation behavior data, specifically including time series analysis, spatial pattern recognition, heat map generation and other methods to discover the spatiotemporal distribution law and correlation of abnormal behavior. The results of the spatiotemporal correlation analysis are visualized, such as drawing a spatiotemporal distribution map, a heat map, etc., so as to intuitively understand the distribution characteristics of abnormal behavior in time and space. The spatiotemporal correlation data is divided into a model training set and a model test set, usually using methods such as cross-validation or retention. The model training set is trained using a support vector machine (SVM) algorithm. SVM is a supervised learning algorithm suitable for classification and regression tasks. The trained model is evaluated through the model test set, and the model is optimized and iterated according to the evaluation results to improve the accuracy and generalization ability of the model. The spatiotemporal correlation data of abnormal transportation behavior is imported into the trained abnormal driving trajectory anomaly detection model to predict the abnormal driving trajectory. Use the trained model to predict abnormal driving trajectories to identify trajectories with abnormal conditions. Generate abnormal driving trajectory prediction data and perform data visualization or further processing and analysis as needed.
优选的，步骤S5包括以下步骤：Preferably, step S5 comprises the following steps:
步骤S51：对正常轨迹片段进行车辆信息监管，生成正常轨迹车辆行驶监管数据；Step S51: Perform vehicle information supervision on the normal trajectory segment to generate normal trajectory vehicle driving supervision data;
步骤S52：基于异常轨迹片段进行车辆预警，生成异常轨迹车辆行驶预警数据；Step S52: Perform vehicle warning based on the abnormal trajectory segment and generate abnormal trajectory vehicle driving warning data;
步骤S53：将正常轨迹车辆行驶监管数据和异常轨迹车辆行驶预警数据进行数据集成，生成行驶轨迹监管预警架构；Step S53: integrating the normal trajectory vehicle driving supervision data and the abnormal trajectory vehicle driving warning data to generate a driving trajectory supervision warning framework;
步骤S54：利用异常行驶轨迹预测数据对行驶轨迹监管预警架构进行预警规则动态调整，生成预警规则调整数据；根据预警规则调整数据对行驶轨迹监管预警架构进行监管及预警推送，以执行行驶轨迹的大件运输监管及预警作业。Step S54: dynamically adjust the warning rules of the driving trajectory supervision and warning framework using the abnormal driving trajectory prediction data to generate warning rule adjustment data; supervise and push warnings to the driving trajectory supervision and warning framework based on the warning rule adjustment data to perform large-scale transportation supervision and warning operations on the driving trajectory.
本发明通过监管正常轨迹，可以建立起对车辆正常行驶的基准，一旦有异常情况出现，就可以及时预警，提高对车辆行驶安全的监管效果。将正常轨迹监管数据和异常轨迹预警数据进行集成，生成行驶轨迹监管预警架构，集成可以让监管者在一个系统内综合查看车辆的行驶情况，提高监管的效率和准确性。利用异常行驶轨迹预测数据对监管预警架构进行预警规则动态调整，这意味着系统能够根据实时数据对预警规则进行调整，更加灵活地适应不同情况下的监管需求，提高了监管的智能化水平。根据预警规则调整数据对监管预警架构进行监管及预警推送，以执行行驶轨迹的大件运输监管及预警作业，步骤保证了监管信息能够及时传达给相关人员，以便他们能够采取必要的行动，确保车辆行驶的安全和顺利，其中具体的预警类型包括：超载预警：检测车辆是否超出规定载重量。超速预警：监控车辆速度，防止超速行驶。行驶轨迹偏离预警：检测车辆是否偏离预定路线。环境风险预警：基于激光雷达和摄像头数据，检测周围环境中的潜在风险。入侵预警：基于生物和非生物感知标记，识别未经授权的进入和潜在威胁。这些预警类型通过统一的描述，明确了每种预警的功能和目的，确保大件运输过程的全面安全监控。The present invention can establish a benchmark for normal vehicle driving by supervising the normal trajectory. Once an abnormal situation occurs, a warning can be issued in time to improve the supervision effect of vehicle driving safety. The normal trajectory supervision data and the abnormal trajectory warning data are integrated to generate a driving trajectory supervision warning framework. The integration allows the supervisor to comprehensively view the driving conditions of the vehicle in one system, thereby improving the efficiency and accuracy of supervision. The supervision warning framework is dynamically adjusted by using the abnormal driving trajectory prediction data, which means that the system can adjust the warning rules according to real-time data, more flexibly adapt to the supervision needs under different circumstances, and improve the intelligent level of supervision. According to the warning rule adjustment data, the supervision warning framework is supervised and the warning is pushed to perform the large-scale transportation supervision and warning operations of the driving trajectory. The steps ensure that the supervision information can be conveyed to the relevant personnel in time so that they can take necessary actions to ensure the safety and smoothness of the vehicle driving. The specific warning types include: Overload warning: detect whether the vehicle exceeds the specified load. Overspeed warning: monitor the vehicle speed to prevent speeding. Driving trajectory deviation warning: detect whether the vehicle deviates from the predetermined route. Environmental risk warning: based on laser radar and camera data, detect potential risks in the surrounding environment. Intrusion warning: Based on biological and non-biological sensing markers, unauthorized entry and potential threats are identified. These warning types clarify the function and purpose of each warning through a unified description to ensure comprehensive safety monitoring of the large-scale transportation process.
本发明实施例中，通过收集正常轨迹片段的车辆信息，包括车辆速度、行驶路径、行驶时间等，通过车载设备或GPS系统实时监管车辆的行驶情况。对监管数据进行处理和整理，将其存储到数据库或数据仓库中，以便后续的分析和查询。基于异常轨迹片段，利用预先设定的规则或机器学习算法进行异常检测，例如检测超速、急刹车、偏离路线等异常情况，并生成异常轨迹车辆行驶预警数据。对预警数据进行处理，包括时间戳、车辆ID、异常类型等信息，并将其存储到数据库或数据仓库中。将正常轨迹车辆行驶监管数据和异常轨迹车辆行驶预警数据进行集成，形成行驶轨迹监管预警架构，整合监管和预警功能。设计监管预警架构的数据结构和流程，确保数据的一致性和完整性。利用异常行驶轨迹预测数据，对行驶轨迹监管预警架构中的预警规则进行动态调整，包括根据模型输出的异常概率对预警规则的权重进行调整。根据预测模型输出的异常概率，生成预警规则调整数据，包括异常类型、预警阈值等信息。根据预警规则调整数据，对行驶轨迹监管预警架构进行监管和预警推送，例如向相关部门发送预警信息、触发监控设备进行实时监控等。In an embodiment of the present invention, by collecting vehicle information of normal trajectory segments, including vehicle speed, driving path, driving time, etc., the vehicle driving situation is monitored in real time through the vehicle-mounted device or GPS system. The supervision data is processed and sorted, and stored in a database or data warehouse for subsequent analysis and query. Based on the abnormal trajectory segments, abnormal detection is performed using pre-set rules or machine learning algorithms, such as detecting abnormal situations such as speeding, sudden braking, and deviation from the route, and abnormal trajectory vehicle driving warning data is generated. The warning data is processed, including information such as timestamp, vehicle ID, and abnormal type, and stored in a database or data warehouse. The normal trajectory vehicle driving supervision data and the abnormal trajectory vehicle driving warning data are integrated to form a driving trajectory supervision and warning architecture, integrating supervision and warning functions. The data structure and process of the supervision and warning architecture are designed to ensure the consistency and integrity of the data. The abnormal driving trajectory prediction data is used to dynamically adjust the warning rules in the driving trajectory supervision and warning architecture, including adjusting the weight of the warning rules according to the abnormal probability output by the model. According to the abnormal probability output by the prediction model, warning rule adjustment data is generated, including information such as abnormal type and warning threshold. Adjust data according to the warning rules, supervise and push warnings to the driving trajectory supervision and warning architecture, such as sending warning information to relevant departments, triggering monitoring equipment for real-time monitoring, etc.
本发明提供了一种基于行驶轨迹的大件运输监管及预警装置，所述基于行驶轨迹的大件运输监管及预警装置包括：The present invention provides a large-scale transport supervision and early warning device based on a driving trajectory, and the large-scale transport supervision and early warning device based on a driving trajectory comprises:
GPS定位装置、北斗定位装置、车载传感器、位置感知设备、车载摄像头、激光雷达和超声波传感器；GPS positioning devices, Beidou positioning devices, vehicle-mounted sensors, location awareness devices, vehicle-mounted cameras, lidar and ultrasonic sensors;
以及至少一个处理器；and at least one processor;
与所述至少一个处理器通信连接的存储器及多种传感器模块，传感器模块包括北斗定位模块、车载传感器模块、环境感知模块以及数据采集模块；A memory and a plurality of sensor modules communicatively connected to the at least one processor, wherein the sensor modules include a Beidou positioning module, a vehicle-mounted sensor module, an environment perception module, and a data acquisition module;
其中，所述存储器存储有可被所述至少一个处理器执行的计算机程序，所述计算机程序被所述至少一个处理器执行，以使所述至少一个处理器能够执行如上所述的任一种基于行驶轨迹的大件运输监管及预警方法。Among them, the memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute any of the large-scale transportation supervision and early warning methods based on driving trajectories as described above.
本发明还提供了一种基于行驶轨迹的大件运输监管及预警系统，用于执行如上所述的基于行驶轨迹的大件运输监管及预警方法，该基于行驶轨迹的大件运输监管及预警系统包括：The present invention also provides a large-scale transport supervision and early warning system based on driving trajectory, which is used to execute the large-scale transport supervision and early warning method based on driving trajectory as described above. The large-scale transport supervision and early warning system based on driving trajectory includes:
数据处理模块，用于获取大件运输车辆实时轨迹数据；对大件运输车辆实时轨迹数据进行数据预处理，生成标准大件运输车辆轨迹数据；The data processing module is used to obtain the real-time trajectory data of large-scale transport vehicles; perform data preprocessing on the real-time trajectory data of large-scale transport vehicles to generate standard large-scale transport vehicle trajectory data;
轨迹片段划分模块，用于对标准大件运输车辆轨迹数据进行运输行驶轨迹分段，得到大件运输轨迹行为片段；对大件运输轨迹行为片段进行轨迹偏差计算，得到轨迹偏差数据；将行驶轨迹偏差数据和预设的行驶轨迹偏差程度阈值进行对比，生成异常轨迹片段和正常轨迹片段；The trajectory segmentation module is used to segment the standard large-scale transport vehicle trajectory data into large-scale transport trajectory behavior segments; calculate the trajectory deviation of the large-scale transport trajectory behavior segments to obtain trajectory deviation data; compare the driving trajectory deviation data with a preset driving trajectory deviation degree threshold to generate abnormal trajectory segments and normal trajectory segments;
异常行为识别模块，用于获取运输环境感知数据；基于运输环境感知数据对异常轨迹片段进行异常运输行为识别，生成异常运输行为数据；对异常运输行为数据进行异常标签处理，生成异常运输轨迹分类标签；The abnormal behavior recognition module is used to obtain the transport environment perception data; based on the transport environment perception data, the abnormal transport behavior of the abnormal trajectory segment is recognized to generate the abnormal transport behavior data; the abnormal transport behavior data is processed with abnormal labels to generate abnormal transport trajectory classification labels;
异常轨迹预测模块，用于基于异常运输行为数据进行时空关联分析，生成异常运输行为时空关联数据；对异常运输行为时空关联数据进行异常行驶轨迹预测，生成异常行驶轨迹预测数据；The abnormal trajectory prediction module is used to perform spatiotemporal correlation analysis based on abnormal transportation behavior data to generate spatiotemporal correlation data of abnormal transportation behavior; perform abnormal driving trajectory prediction on the spatiotemporal correlation data of abnormal transportation behavior to generate abnormal driving trajectory prediction data;
监管及预警模块，用于对正常轨迹片段进行车辆信息监管，生成正常轨迹车辆行驶监管数据；基于异常轨迹片段进行车辆预警，生成异常轨迹车辆行驶预警数据；将正常轨迹车辆行驶监管数据和异常轨迹车辆行驶预警数据进行数据集成，生成行驶轨迹监管预警架构；利用异常行驶轨迹预测数据对行驶轨迹监管预警架构进行监管及预警推送，以执行行驶轨迹的大件运输监管及预警作业。The supervision and warning module is used to supervise the vehicle information of normal trajectory segments and generate normal trajectory vehicle driving supervision data; to carry out vehicle warning based on abnormal trajectory segments and generate abnormal trajectory vehicle driving warning data; to integrate the normal trajectory vehicle driving supervision data and the abnormal trajectory vehicle driving warning data to generate a driving trajectory supervision and warning framework; to use the abnormal driving trajectory prediction data to supervise and push warnings to the driving trajectory supervision and warning framework, so as to perform large-scale transportation supervision and warning operations on the driving trajectory.
本发明的有益效果在于通过获取大件运输车辆的实时轨迹数据并进行处理分析，能够实现对大件运输过程的实时监管和异常预警，及时发现问题并采取措施。利用运输环境感知数据对异常轨迹片段进行异常运输行为识别，能够准确识别出异常的运输行为，帮助预防事故和损失。通过时空关联分析，可以深入理解异常运输行为的发生机制和规律，有助于提高对异常情况的理解和应对能力。基于异常运输行为数据进行时空关联分析，并生成异常行驶轨迹预测数据，可以提前预警潜在的异常情况，有效减少事故发生的可能性。将正常轨迹车辆行驶监管数据与异常轨迹车辆行驶预警数据进行数据集成，形成综合的行驶轨迹监管预警架构，使监管工作更加全面和有效。通过监管和预警，能够提升大件运输的安全性和效率，保障货物和人员的安全，同时优化运输流程，降低成本。因此，本发明通过数据处理、异常检测、环境感知和时空关联分析，提高了监管预警的准确性和效率。The beneficial effect of the present invention is that by acquiring the real-time trajectory data of large-scale transport vehicles and processing and analyzing them, it is possible to realize real-time supervision and abnormal warning of the large-scale transportation process, discover problems in time and take measures. By using the transportation environment perception data to identify abnormal transportation behaviors of abnormal trajectory segments, it is possible to accurately identify abnormal transportation behaviors and help prevent accidents and losses. Through spatiotemporal correlation analysis, the occurrence mechanism and law of abnormal transportation behaviors can be deeply understood, which helps to improve the understanding and response capabilities of abnormal situations. Based on the abnormal transportation behavior data, spatiotemporal correlation analysis is performed, and abnormal driving trajectory prediction data is generated, which can warn potential abnormal situations in advance and effectively reduce the possibility of accidents. The normal trajectory vehicle driving supervision data and the abnormal trajectory vehicle driving warning data are integrated to form a comprehensive driving trajectory supervision and warning framework, making the supervision work more comprehensive and effective. Through supervision and warning, the safety and efficiency of large-scale transportation can be improved, the safety of goods and personnel can be guaranteed, and the transportation process can be optimized and the cost can be reduced. Therefore, the present invention improves the accuracy and efficiency of supervision and warning through data processing, anomaly detection, environmental perception and spatiotemporal correlation analysis.
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 |
|---|---|---|---|
| CN202410695403.6ACN118278841B (en) | 2024-05-31 | 2024-05-31 | Large-piece transportation supervision and early warning method, system and device based on driving trajectory |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410695403.6ACN118278841B (en) | 2024-05-31 | 2024-05-31 | Large-piece transportation supervision and early warning method, system and device based on driving trajectory |
| Publication Number | Publication Date |
|---|---|
| CN118278841Atrue CN118278841A (en) | 2024-07-02 |
| CN118278841B CN118278841B (en) | 2024-10-08 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410695403.6AActiveCN118278841B (en) | 2024-05-31 | 2024-05-31 | Large-piece transportation supervision and early warning method, system and device based on driving trajectory |
| Country | Link |
|---|---|
| CN (1) | CN118278841B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118469426A (en)* | 2024-07-13 | 2024-08-09 | 江西财经大学 | A data processing method and system based on Internet of Things |
| CN118736512A (en)* | 2024-09-04 | 2024-10-01 | 沈阳盘古网络技术有限公司 | A method and system for automatically identifying mining transport vehicles |
| CN119024381A (en)* | 2024-10-22 | 2024-11-26 | 湖南湘科智慧科技有限公司 | Real-time behavior trajectory big data analysis and processing method based on Beidou positioning |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018122585A1 (en)* | 2016-12-30 | 2018-07-05 | 同济大学 | Method for urban road traffic incident detecting based on floating-car data |
| WO2021077760A1 (en)* | 2019-10-23 | 2021-04-29 | 江苏智通交通科技有限公司 | Abnormal driving early warning method on basis of reasonable driving range of vehicle at intersection |
| CN115188187A (en)* | 2022-07-05 | 2022-10-14 | 浙江嘉兴数字城市实验室有限公司 | A roadside perception data quality monitoring system and method based on vehicle-road collaboration |
| CN116010885A (en)* | 2022-12-21 | 2023-04-25 | 重庆邮电大学 | Method and system for detecting abnormal space-time data of vehicle under long-sequence condition |
| CN116659519A (en)* | 2023-07-31 | 2023-08-29 | 安徽领云物联科技有限公司 | Intelligent park travelling locus tracking system based on artificial intelligence |
| CN117576636A (en)* | 2024-01-17 | 2024-02-20 | 新汽有限公司 | Logistics transportation process monitoring method and system |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018122585A1 (en)* | 2016-12-30 | 2018-07-05 | 同济大学 | Method for urban road traffic incident detecting based on floating-car data |
| WO2021077760A1 (en)* | 2019-10-23 | 2021-04-29 | 江苏智通交通科技有限公司 | Abnormal driving early warning method on basis of reasonable driving range of vehicle at intersection |
| CN115188187A (en)* | 2022-07-05 | 2022-10-14 | 浙江嘉兴数字城市实验室有限公司 | A roadside perception data quality monitoring system and method based on vehicle-road collaboration |
| CN116010885A (en)* | 2022-12-21 | 2023-04-25 | 重庆邮电大学 | Method and system for detecting abnormal space-time data of vehicle under long-sequence condition |
| CN116659519A (en)* | 2023-07-31 | 2023-08-29 | 安徽领云物联科技有限公司 | Intelligent park travelling locus tracking system based on artificial intelligence |
| CN117576636A (en)* | 2024-01-17 | 2024-02-20 | 新汽有限公司 | Logistics transportation process monitoring method and system |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118469426A (en)* | 2024-07-13 | 2024-08-09 | 江西财经大学 | A data processing method and system based on Internet of Things |
| CN118469426B (en)* | 2024-07-13 | 2024-09-27 | 江西财经大学 | A data processing method and system based on Internet of Things |
| CN118736512A (en)* | 2024-09-04 | 2024-10-01 | 沈阳盘古网络技术有限公司 | A method and system for automatically identifying mining transport vehicles |
| CN119024381A (en)* | 2024-10-22 | 2024-11-26 | 湖南湘科智慧科技有限公司 | Real-time behavior trajectory big data analysis and processing method based on Beidou positioning |
| Publication number | Publication date |
|---|---|
| CN118278841B (en) | 2024-10-08 |
| Publication | Publication Date | Title |
|---|---|---|
| CN118278841B (en) | Large-piece transportation supervision and early warning method, system and device based on driving trajectory | |
| US11380105B2 (en) | Identification and classification of traffic conflicts | |
| CN118486152B (en) | Security alarm information data interaction system and method | |
| CN118469426B (en) | A data processing method and system based on Internet of Things | |
| JP2023546672A (en) | Rail feature identification system | |
| CN118053260B (en) | Production workshop construction operation safety early warning system and method based on the Internet of Things | |
| CN119672932B (en) | Abnormal behavior security early warning system based on multi-video image analysis | |
| CN119399446A (en) | A smart community target recognition and tracking method based on big data | |
| CN118366312B (en) | Traffic detection system and method | |
| CN120088989B (en) | Traffic road network multi-mode sensing abnormal event early warning method and system | |
| Hu et al. | Detecting socially abnormal highway driving behaviors via recurrent graph attention networks | |
| Patel et al. | A framework for proactive safety evaluation of intersection using surrogate safety measures and non-compliance behavior | |
| CN119671283A (en) | A behavior analysis and safety risk prevention and control system for electrical workers | |
| CN119917970A (en) | Abnormal event detection and alarm method for highways based on deep learning | |
| CN118447720B (en) | Vision-based anti-collision method and system for automatic driving vehicle | |
| CN119380306A (en) | Road abnormality prediction and warning method, device, equipment and computer-readable storage medium | |
| CN118863775A (en) | Integrated prevention and control service platform based on multi-department collaboration | |
| CN119007131A (en) | Road side digital video monitoring method and system based on deep learning | |
| CN118606667A (en) | Intelligent traffic scene recognition method driven by spatiotemporal element association and safety requirements | |
| Das et al. | Vehicular propagation velocity forecasting using open CV | |
| Anand et al. | Advanced Traffic monitoring: Precise vehicle counting, speed estimation, color recognition and Type classification utilizing ROC curve analysis | |
| CN117315943B (en) | Monitoring analysis and early warning method and system for overrun transportation violations | |
| CN119600823B (en) | Unmanned toll station traffic flow monitoring method and system based on data fusion | |
| CN117590863B (en) | Unmanned aerial vehicle cloud edge end cooperative control system of 5G security rescue net allies oneself with | |
| CN119478860B (en) | Large transport vehicle identification method and system based on radar and vision fusion |
| Date | Code | Title | Description |
|---|---|---|---|
| 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 |