CN112162519A - Compound machine tool digital twin monitoring system - Google Patents

Abstract

Translated fromChinese

本发明公开了一种复合型机床数字孪生监控系统，涉及数字孪生技术领域，该系统通过机床信息模块创建数字孪生监控系统，并基于OPC‑UA传输接口实现多平台、多数据、多接口通信，构建人机交互模块；多领域数据采集模块采用不同类型的传感器实时采集多源异构数据，基于信息融合技术对采集数据进行处理形成孪生数据，并转发至建模计算模块；建模计算模块通过孪生数据的驱动，结合约束、预测、决策等规则，形成复合型机床的数字孪生体；个性化决策模块通过实时重构优化机床监控孪生模型，对实体机床设备实时监控与管理。本发明简化了对复合型机床运行的监控过程，提高了机床系统的监控精度，实现了对复合型机床运行的主动式预测性维护。

The invention discloses a composite machine tool digital twin monitoring system, which relates to the technical field of digital twins. The system creates a digital twin monitoring system through a machine tool information module, and realizes multi-platform, multi-data and multi-interface communication based on an OPC-UA transmission interface. Build a human-computer interaction module; the multi-domain data acquisition module uses different types of sensors to collect multi-source heterogeneous data in real time, processes the collected data based on information fusion technology to form twin data, and forwards it to the modeling calculation module; the modeling calculation module passes Driven by twin data, combined with constraints, prediction, decision-making and other rules, a digital twin of compound machine tools is formed; the personalized decision-making module optimizes the machine tool monitoring twin model through real-time reconstruction, and monitors and manages physical machine tools in real time. The invention simplifies the monitoring process of the operation of the compound machine tool, improves the monitoring precision of the machine tool system, and realizes the active predictive maintenance of the operation of the compound machine tool.

Description

Compound machine tool digital twin monitoring system

Technical Field

The invention belongs to the technical field of digital twinning, and particularly relates to a composite type machine tool digital twinning monitoring system.

Background

The digital twin nature concept is a quasi-real time digitized mirror of a physical entity or process. Compared with the traditional life cycle management and simulation technology, the digital twin has the characteristics of two-way, continuous and open, and new capacity is added to the physical entity through means of virtual-real interaction feedback, data fusion analysis, decision iteration optimization and the like. For the manufacturing field, the digital twin may be defined as "a dynamic model that digitally describes manufacturing elements such as people, products, equipment, and processes, and synchronously updates optimization as the development state of objects, working conditions, product geometry, etc. change", reflecting the full life cycle process of the corresponding entity equipment.

The method for analyzing and mining the data is a big data technology, but the big data technology is mainly focused on the data analysis and mining of a machine tool entity system, and operating system data in a virtual world is ignored. With the system of the machine tool being more intelligent, the problems of real-time monitoring of data and running states and the like are more prominent; the data-driven method collects data by installing a large number of sensors, thereby obtaining the operating state and information of the machine tool. However, some key components cannot be equipped with sensors, resulting in great difficulty in data acquisition and further mining. Therefore, the development of the intelligent monitoring technology of the compound machine tool has limited the integration, intelligentization and flexibility process of the machine tool to a certain extent, and becomes one of the bottleneck technologies affecting the reliability and precision development of the compound machine tool system, and a breakthrough of the theoretical method and technology for monitoring and maintaining the machine tool operation system is urgently needed.

Disclosure of Invention

The invention aims to provide a digital twin monitoring system of a composite machine tool, thereby overcoming the defects of the traditional composite machine tool monitoring system.

In order to achieve the above object, the present invention provides a digital twin monitoring system for a composite machine tool, comprising:

the machine tool information module is used for constructing, storing and managing information models of components related in an actual composite machine tool system, and the information models comprise a geometric model, a functional information model, a rule model, a behavior model, a performance prediction model and a control logic model;

the multi-field data acquisition module receives data types such as position, speed, current, rotating speed, load, motor load and the like in real time through multiple channels, respectively stores relevant data information of a main shaft, a feed shaft, a cutter and a machining program, and transmits the data information to the data processing module and the modeling calculation module;

and the data processing module is used for receiving the running state data information transmitted by the real-time multi-field data acquisition module, and deeply processing the acquired numerical information by using a big data technology so as to eliminate noise components and redundant information in the signal and reconstruct a new data set. Then, when time and frequency domain statistical characteristics are extracted from the reconstructed data set, and the component signals decomposed by the characteristic extraction technology are quantized by using indexes such as energy values, entropy values, spectral kurtosis and the like; finally, dimensionality reduction is carried out on the high-dimensional features, and preparation is made for accurate modeling and predictive maintenance of an operating system;

and the modeling calculation module is used for receiving the running state data information transmitted by the real-time multi-field data acquisition module, fusing the data as driving data, and associating constraint rules, prediction rules, decision rules and the like together to jointly form a digital twin body of the manufacturing and processing equipment, wherein the digital twin body exists in the whole life product cycle of the manufacturing and processing equipment and can dynamically, truly and real-timely reflect the real state of the manufacturing and processing equipment in a physical layer.

The human-computer interaction module is used for receiving the machine tool running state data sent by the real-time data acquisition module, processing and analyzing the running state data of the actual production line and obtaining the actual running state data of each component related to the actual running state; establishing a virtual monitoring system corresponding to an actual machine tool system, namely: the digital twin monitoring system is realized by mapping a geometric model corresponding to a part involved in actual operation in a machine tool information model module, the geometric model binds a part function information model, a rule model, a behavior model, a control logic model and a motion control display program in an attribute form, the motion control display program of each part extracts actual operation state data corresponding to the part, then the actual operation state data corresponding to the part is subjected to deep processing to obtain virtual operation state monitoring data, the geometric model is controlled according to the virtual row state monitoring data to complete simulation action, and dynamic simulation of the composite machine tool operation state monitoring system is realized. Meanwhile, visually expressing the state data corresponding to the part on an interface in a chart form;

and the personalized service decision module is used for dynamically tracking and reflecting the latest state of the equipment entity by the machine tool physical entity through the digital twin body, generating corresponding decision information through simulation, and monitoring and optimizing a machine tool physical system by using the generated decision information so as to finally realize the fusion and intelligent monitoring of the physical information and the virtual information of the manufacturing and processing equipment.

Further, the machine tool information module comprises a geometric model, a functional information model, a rule model, a behavior model and a control logic model.

Further, the multi-domain data acquisition module comprises:

data acquisition, including multi-source physical field data, model generation data and virtual-real fusion data, and acquiring numerical information by using different sensors according to different types of physical fields, for example, a temperature sensor is used in a temperature field, a hall sensor is used in an electromagnetic field, a pressure sensor is used in a structural field, and the like;

data management, namely generating multiplication data information by using directly acquired numerical information, finally enabling the data information to be in a mesh structure, and backing up and storing the data;

and data transmission, namely receiving data types such as position, speed, current, rotating speed, load, motor load and the like in real time by multiple channels, transmitting data information to the data processing module and the modeling calculation module, and realizing communication interaction among the modules.

Furthermore, the data processing module receives the running state data information transmitted by the real-time multi-field data acquisition module, and deeply processes the acquired numerical information by using a big data technology so as to eliminate noise components and redundant information in the signals and reconstruct a new data set. Then, when time and frequency domain statistical characteristics are extracted from the reconstructed data set, and the component signals decomposed by the characteristic extraction technology are quantized by using indexes such as energy values, entropy values, spectral kurtosis and the like; finally, dimensionality reduction is carried out on the high-dimensional features, and preparation is made for accurate modeling and predictive maintenance of an operating system;

furthermore, the modeling calculation module receives the running state data information transmitted by the real-time multi-field data acquisition module, constructs a self-adaptive model changing along with the running environment, and can accurately monitor the performance of parts and the whole machine of the machine tool; injecting a fault mode in the historical maintenance data of the machine tool into the three-dimensional physical model and the performance model to construct a fault model which can be used for fault diagnosis and prediction; combining the historical operating data of the machine tool with a performance model and fusing a data driving method to construct a performance prediction model, and predicting the performance and the residual life of the whole machine; the local linearization model and the machine tool running state environment model are fused and a control optimization model is constructed, so that the optimization of the machine tool control performance can be realized, and the machine tool can play better performance in the running process.

Further, the personalized service decision module is debugged and verified by adopting a virtual PLC and a virtual prototype, and is transferred to a real machine tool system for butt joint after being matured. And then, the virtual prototype system and the real physical system are synchronized in real time based on the physical PLC. The machine tool physical entity and the virtual model carry out data information interaction, an information layer is transmitted through the data mapping dictionary, an interface used for transmission is an OPC-UA interface, the interface can unify model data under the condition of ensuring that communication data are not lost, a complex data model is supported, and communication of multiple platforms, multiple data and multiple interfaces can be realized.

Furthermore, the simulation platform adopts an Untiy real-time three-dimensional interactive virtual content construction platform, can perform accurate physical simulation, and can also customize a development editing interface so as to facilitate secondary development.

Further, the fault diagnosis and prediction comprises the following steps:

s1, classifying according to the type of the physical field according to the system modeling information;

s2, normalizing the modeling data of one type of physical field;

s3, setting the output of the neural network as the fault type and fault degree of the physical field, and training the normalized data through the neural network to obtain the trained neural network;

s4, repeating S2-S3 to obtain trained neural networks corresponding to all physical fields;

and S5, classifying the system modeling information acquired in real time through S1, and constructing a fault model to perform fault diagnosis and prediction through the correspondingly trained neural network in the corresponding type S4 by combining fault modes in the historical maintenance data of the machine tool to obtain diagnosis and prediction results.

Further, the physical field adopted for fault diagnosis and prediction is selected and fused according to the operation condition of the machine tool system.

Further, the monitoring process comprises the steps of:

a1, classifying according to the type of the operating physical field, and drawing a data curve when corresponding production equipment operates normally;

a2, classifying historical fault information according to the type of production equipment, and drawing a corresponding historical data curve;

a3, comparing the historical data curve with the corresponding data curve in normal operation, judging the contact ratio, and if the historical data curve is basically normal with the corresponding data curve in normal operation, the production equipment corresponding to the historical data curve works normally; otherwise, the fault occurs;

a4, overhauling the corresponding equipment according to the fault, marking the fault type on the corresponding real-time data curve, and forming a data curve with fault analysis to replace the data curve in A1;

a5, repeating A1-A4 to obtain a data curve with fault analysis;

a6, classifying the real-time working condition information according to the type of production equipment, and drawing a corresponding real-time data curve;

a7, comparing the real-time data curve with the corresponding data curve with fault analysis, judging the contact ratio, and when the real-time data curve and the corresponding data curve with fault analysis are basically normal, the production equipment corresponding to the real-time data curve works normally; otherwise, a real-time diagnosis result is obtained for the occurrence of the fault, and the intelligent monitoring of the operation process of the machine tool is further realized.

Compared with the prior art, the invention has the following beneficial effects:

the invention provides a composite machine tool digital twin monitoring system, which is characterized in that a digital twin monitoring system is established through a machine tool information module, multi-platform, multi-data and multi-interface communication is realized based on an OPC-UA transmission interface, and a man-machine interaction module of the composite machine tool twin monitoring system is established; the multi-field data acquisition module adopts different sensors to acquire multi-source heterogeneous data in the operation process in real time according to different physical field types, processes the acquired data based on an information fusion technology to form twin data, and forwards the twin data to the modeling calculation module; the modeling calculation module is driven by twin data of the operation of the composite machine tool and is combined with rules such as constraint rules, prediction rules, decision rules and the like to jointly form a digital twin body of the composite machine tool, and the real state of equipment in a physical layer can be dynamically reflected in real time; and the personalized decision module is used for realizing real-time monitoring and management of the entity machine tool equipment by reconstructing and optimizing the machine tool monitoring twin model in real time. The invention simplifies the monitoring process of the operation process of the composite machine tool, improves the monitoring precision of a machine tool system, improves the accuracy of decision information and realizes the active predictive maintenance of the operation of the composite machine tool.

Drawings

In order to more clearly illustrate the technical solution of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only one embodiment of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on the drawings without creative efforts.

Fig. 1 is a schematic structural diagram of a composite type machine tool digital twin monitoring system provided by the invention.

Fig. 2 is a schematic structural diagram of a machine tool information module of a composite machine tool digital twin monitoring system provided by the invention.

Fig. 3 is a schematic structural diagram of a multi-domain data acquisition module of the composite machine tool digital twin monitoring system provided by the invention.

Fig. 4 is a schematic structural diagram of a data processing module of the composite machine tool digital twin monitoring system provided by the invention.

Fig. 5 is a schematic structural diagram of a modeling calculation module of the composite machine tool digital twin monitoring system provided by the invention.

Fig. 6 is a schematic structural diagram of a human-computer interaction module of the composite machine tool digital twin monitoring system provided by the invention.

Fig. 7 is a schematic structural diagram of a personalized service decision module of the composite machine tool digital twin monitoring system provided by the invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in fig. 1, the digital twin monitoring system of the compound machine tool provided by the invention comprises: the system comprises a machine tool information module, a multi-field data acquisition module, a data processing module, a modeling calculation module, a human-computer interaction module and a personalized service decision module.

As shown in fig. 2, the machine tool information module is configured to construct, store, and manage information models of components involved in an actual composite machine tool system, where the information models include a geometric model, a functional information model, a rule model, a behavior model, a performance prediction model, and a control logic model.

As shown in fig. 3, the multi-domain data acquisition module receives data types such as position, speed, current, rotation speed, load, motor load and the like in real time through multiple channels, respectively stores data information related to a main shaft, a feed shaft, a tool and a machining program, and transmits the data information to the data processing module and the modeling calculation module.

As shown in fig. 4, the data processing module receives the operation state data information transmitted by the real-time multi-domain data acquisition module, and performs deep processing on the acquired numerical information by using a big data technology to eliminate noise components and redundant information in the signal and reconstruct a new data set. Then, when time and frequency domain statistical characteristics are extracted from the reconstructed data set, and the component signals decomposed by the characteristic extraction technology are quantized by using indexes such as energy values, entropy values, spectral kurtosis and the like; and finally, carrying out dimensionality reduction on the high-dimensional features, and preparing for accurate modeling and predictive maintenance of an operating system.

As shown in fig. 5, the modeling calculation module receives the operation state data information transmitted by the real-time multi-domain data acquisition module, fuses the data as driving data, and associates constraint rules, prediction rules, decision rules and the like together to form a digital twin body of the manufacturing and processing equipment, wherein the digital twin body exists in the whole life product cycle of the manufacturing and processing equipment and can dynamically, truly and real-timely reflect the real state of the manufacturing and processing equipment in the physical layer.

As shown in fig. 6, the human-computer interaction module receives the machine tool running state data sent by the real-time data acquisition module, and processes and analyzes the running state data of the actual production line to obtain the actual running state data of each component involved in the actual running state; establishing a virtual monitoring system corresponding to an actual machine tool system, namely: the digital twin monitoring system is realized by mapping a geometric model corresponding to a part involved in actual operation in a machine tool information model module, the geometric model binds a part function information model, a rule model, a behavior model, a control logic model and a motion control display program in an attribute form, the motion control display program of each part extracts actual operation state data corresponding to the part, then the actual operation state data corresponding to the part is subjected to deep processing to obtain virtual operation state monitoring data, the geometric model is controlled according to the virtual row state monitoring data to complete simulation action, and dynamic simulation of the composite machine tool operation state monitoring system is realized. And simultaneously, visually expressing the state data corresponding to the part on the interface in a chart form.

As shown in fig. 7, in the personalized service decision module, the machine tool physical entity dynamically tracks and reflects the latest state of the equipment entity through the digital twin body, generates corresponding decision information through simulation, and monitors and optimizes the machine tool physical system by using the generated decision information, thereby finally realizing the fusion and intelligent monitoring of the physical information and the virtual information of the manufacturing and processing equipment.

The machine tool information module comprises a geometric model, a functional information model, a rule model, a behavior model, a performance prediction model and a control logic model.

The information of the geometric model comprises the overall dimension, geometric tolerance, assembly position relation, color information, central coordinate point component material information and component matching relation;

the information of the rule model comprises a constraint rule, a prediction rule and a decision rule;

the information of the behavior model comprises dynamics constraint, dynamic scene import and control logic definition;

with continued reference to fig. 1, the multi-domain data acquisition module includes different types of sensors for acquiring data types such as a multi-channel real-time receiving position, speed, current, rotation speed, load, motor load, and the like, the acquisition contents mainly include multi-source physical field data, model generation data, and virtual-real fusion data, and according to the difference of the types of the physical fields, different sensors are used to acquire numerical information, for example, a temperature sensor is used in a temperature field, a hall sensor is used in an electromagnetic field, and a pressure sensor is used in a structural field.

The tasks of the data processing module include:

receiving running state data information transmitted by a real-time multi-field data acquisition module, and carrying out deep processing on the acquired numerical value information by using a big data technology so as to eliminate noise components and redundant information in signals and reconstruct a new data set;

when the statistical characteristics of a frequency domain are extracted from the reconstructed data set, the component signals decomposed by the characteristic extraction technology are quantized by using indexes such as an energy value, an entropy value and a spectral kurtosis;

and finally, carrying out dimensionality reduction on the high-dimensional features, and preparing for accurate modeling and predictive maintenance of an operating system.

The modeling calculation module comprises fusion of multidimensional models of an operating system mechanism model, a data driving model, a fault prediction model and an abnormal event diagnosis model, and is combined with a constraint rule, a prediction rule, a decision rule and the like to form a digital twin body of the manufacturing and processing equipment together.

The human-computer interaction module is used for receiving the machine tool running state data sent by the real-time data acquisition module, processing and analyzing the running state data of the actual production line and obtaining the actual running state data of each component related to the actual running state; establishing a virtual monitoring system corresponding to an actual machine tool system, namely: the digital twin monitoring system is realized by mapping a geometric model corresponding to a part involved in actual operation in a machine tool information model module, the geometric model binds a part function information model, a rule model, a behavior model, a control logic model and a motion control display program in an attribute form, the motion control display program of each part extracts actual operation state data corresponding to the part, then the actual operation state data corresponding to the part is subjected to deep processing to obtain virtual operation state monitoring data, the geometric model is controlled according to the virtual row state monitoring data to complete simulation action, and dynamic simulation of the composite machine tool operation state monitoring system is realized. Meanwhile, visually expressing the state data corresponding to the part on an interface in a chart form;

the machine tool physical entity dynamically tracks and reflects the latest state of the equipment entity through the digital twin body of the machine tool physical entity, generates corresponding decision information through simulation, monitors and optimizes a machine tool physical system by utilizing the generated decision information, and finally realizes the fusion of the physical information and the virtual information of the manufacturing and processing equipment and intelligent monitoring

The monitoring mechanism of the digital twin monitoring system comprises a fault diagnosis prediction part and an intelligent monitoring part, and is based on the driving of massive twin data, and utilizes a big data technology combined with algorithms such as deep learning and neural network to extract characteristic data of interactive data, so that the accurate predictive maintenance of a machine tool system is realized, and an accurate decision can be provided for the system.

The fault diagnosis and prediction comprises the following steps:

s1, classifying according to the type of the physical field according to the system modeling information;

s2, normalizing the modeling data of one type of physical field;

s3, setting the output of the neural network as the fault type and fault degree of the physical field, and training the normalized data through the neural network to obtain the trained neural network;

s4, repeating S2-S3 to obtain trained neural networks corresponding to all physical fields;

and S5, classifying the system modeling information acquired in real time through S1, and constructing a fault model to perform fault diagnosis and prediction through the correspondingly trained neural network in the corresponding type S4 by combining fault modes in the historical maintenance data of the machine tool to obtain diagnosis and prediction results.

Further, the physical field adopted for fault diagnosis and prediction is selected and fused according to the operation condition of the machine tool system.

The intelligent monitoring process comprises the following steps:

a1, classifying according to the type of the operating physical field, and drawing a data curve when corresponding production equipment operates normally;

a2, classifying historical fault information according to the type of production equipment, and drawing a corresponding historical data curve;

a3, comparing the historical data curve with the corresponding data curve in normal operation, judging the contact ratio, and if the historical data curve is basically normal with the corresponding data curve in normal operation, the production equipment corresponding to the historical data curve works normally; otherwise, the fault occurs;

a4, overhauling the corresponding equipment according to the fault, marking the fault type on the corresponding real-time data curve, and forming a data curve with fault analysis to replace the data curve in A1;

a5, repeating A1-A4 to obtain a data curve with fault analysis;

a6, classifying the real-time working condition information according to the type of production equipment, and drawing a corresponding real-time data curve;

a7, comparing the real-time data curve with the corresponding data curve with fault analysis, judging the contact ratio, and when the real-time data curve and the corresponding data curve with fault analysis are basically normal, the production equipment corresponding to the real-time data curve works normally; otherwise, a real-time diagnosis result is obtained for the occurrence of the fault, and the intelligent monitoring of the operation process of the machine tool is further realized.

The working principle of the digital twin monitoring system of the compound machine tool is explained in detail so that the technical personnel in the field can understand the invention more:

the multisource data acquisition module adopts different sensors to acquire numerical information according to different types of physical fields, for example, a temperature sensor is adopted in a temperature field, a Hall sensor is adopted in an electromagnetic field, a pressure sensor is adopted in a structural field, and the like, and data information is sent to the data processing module through a transmission interface OPC-UA; the data processing module processes the received real-time data, such as cleaning, data mining, data association, data noise reduction, feature extraction and the like, so as to eliminate noise components and redundant information in the signals, reconstruct a new data set and transmit the new data set to the modeling calculation module; the modeling calculation module carries out secondary mining on the reconstructed data set to form fusion of multidimensional models of a system mechanism model, a data driving model, a fault prediction model and an abnormal event diagnosis model, and digital twins of manufacturing and processing equipment are formed together by utilizing a constraint rule, a prediction rule, a decision rule and the like in a correlation mode; the digital twin body is corrected and optimized according to real-time data; the digital twin monitoring system combines the digital twin correction optimization model with a historical fault mode and an operating environment condition, forms an early warning signal and decision information according to a monitoring result, and simultaneously stores the early warning signal and the decision information to the server; the method for fault diagnosis and prediction can be selected according to the requirement.

The above disclosure is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of changes or modifications within the technical scope of the present invention, and shall be covered by the scope of the present invention.

Claims (10)

Translated fromChinese

1.一种复合型机床数字孪生监控系统，其特征在于包括：机床信息模块、多领域数据采集模块、数据处理模块、建模计算模块、人机交互模块、个性化服务决策模块；其中：1. a composite machine tool digital twin monitoring system, is characterized in that comprising: machine tool information module, multi-domain data acquisition module, data processing module, modeling calculation module, human-computer interaction module, personalized service decision module; Wherein:机床信息模块，用于构建、存储、管理实际复合型机床系统中所涉及部件的信息模型，所述信息模型包括几何模型、功能信息模型、规则模型、行为模型、性能预测模型、控制逻辑模型；The machine tool information module is used to construct, store and manage the information models of the components involved in the actual compound machine tool system, the information models include geometric models, functional information models, rule models, behavior models, performance prediction models, and control logic models;多领域数据采集模块，多通道实时接收位置、速度、电流、旋转速度、负载与电机负载等数据类型，并分别存储主轴、进给轴、刀具和加工程序相关数据信息，并将数据信息传输到数据处理模块和建模计算模块；Multi-domain data acquisition module, multi-channel real-time reception of position, speed, current, rotation speed, load and motor load and other data types, and store the spindle, feed axis, tool and processing program related data information, and transmit the data information to Data processing module and modeling calculation module;数据处理模块，接收实时多领域数据采集模块传输的运行状态数据信息，并对采集到的数值信息利用大数据技术进行深加工，以消除信号中存在噪声成分和冗余信息，重新构造新的数据集；The data processing module receives the operating status data information transmitted by the real-time multi-domain data acquisition module, and uses the big data technology to further process the collected numerical information to eliminate noise components and redundant information in the signal, and reconstruct a new data set ;之后，从重构后的数据集中提取时、频域统计特征，并利用能量值、熵值、谱峭度等指标对特征提取技术分解后的分量信号进行量化；最后对高维特征进行维数约简，为精准建模以及对运行系统预测性维护做准备；After that, the time and frequency domain statistical features are extracted from the reconstructed data set, and the component signals decomposed by the feature extraction technology are quantified by using indicators such as energy value, entropy value, and spectral kurtosis. Reduction in preparation for accurate modeling and predictive maintenance of operational systems;建模计算模块，接收实时多领域数据采集模块传输的运行状态数据信息，并将这些数据融合作为驱动数据，利用约束规则、预测规则、决策规则等关联在一起，共同形成制造加工设备的数字孪生体，该数字孪生体存在于制造加工设备的整个生命产品周期，并可以动态、真实、实时地反映物理层中制造加工设备的真实状态；The modeling calculation module receives the operating status data information transmitted by the real-time multi-domain data acquisition module, and integrates these data as driving data, and uses constraint rules, prediction rules, decision rules, etc. to associate them together to form a digital twin of manufacturing and processing equipment. The digital twin exists in the entire life product cycle of the manufacturing and processing equipment, and can dynamically, truly and in real time reflect the real state of the manufacturing and processing equipment in the physical layer;人机交互模块，接收实时数据采集模块发送的机床运行状态数据，并对实际生产线的运行状态数据进行处理分析，得到实际运行状态中涉及的每个部件的实际运行状态数据；建立一个与实际机床系统相对应的虚拟监控系统，即：数字孪生监控系统，所述数字孪生监控系统通过映射机床信息模型模块中实际运行中涉及部件所对应的几何模型实现，所述几何模型以属性的形式绑定了部件功能信息模型、规则模型、行为模型、控制逻辑模型和运动控制显示程序，每个部件的运动控制显示程序提取该部件对应的实际运行状态数据后，对该部件对应的实际运行状态数据进行深度处理，得到虚拟运行状态监控数据，根据虚拟行状态监控数据控制几何模型完成仿真动作，实现对复合型机床运行状态监控系统的动态仿真；The human-computer interaction module receives the machine tool running status data sent by the real-time data acquisition module, processes and analyzes the running status data of the actual production line, and obtains the actual running status data of each component involved in the actual running status; The virtual monitoring system corresponding to the system, namely: the digital twin monitoring system, the digital twin monitoring system is realized by mapping the geometric models corresponding to the components involved in the actual operation in the machine tool information model module, and the geometric models are bound in the form of attributes The function information model, rule model, behavior model, control logic model and motion control display program of each component are obtained. After the motion control display program of each component extracts the actual operating state data corresponding to the component, Through in-depth processing, the virtual running state monitoring data is obtained, and the geometric model is controlled according to the virtual row state monitoring data to complete the simulation action, and the dynamic simulation of the running state monitoring system of the compound machine tool is realized;同时，将该部件对应的状态数据以图表形式可视化表达在界面上；At the same time, the state data corresponding to the component is visually expressed on the interface in the form of a chart;个性化服务决策模块，机床物理实体通过其数字孪生体动态地跟踪及反映设备实体的最新状态，并通过仿真模拟产生相应的决策信息，并利用产生的决策信息对机床物理系统进行监控、优化，最终实现制造加工设备的物理信息与虚拟信息融合及智能化监控。Personalized service decision-making module, the physical entity of the machine tool dynamically tracks and reflects the latest state of the equipment entity through its digital twin, and generates corresponding decision-making information through simulation, and uses the generated decision-making information to monitor and optimize the machine tool physical system, Finally, the fusion of physical information and virtual information and intelligent monitoring of manufacturing and processing equipment are realized.2.根据权利要求1所述的一种复合型机床数字孪生监控系统，其特征在于：所述机床信息模块包括几何模型、功能信息模型、规则模型、行为模型、控制逻辑模型。2 . The composite machine tool digital twin monitoring system according to claim 1 , wherein the machine tool information module includes a geometric model, a function information model, a rule model, a behavior model, and a control logic model. 3 .3.根据权利要求1所述的一种复合型机床数字孪生监控系统，其特征在于：所述多领域数据采集模块包括：3. A kind of compound machine tool digital twin monitoring system according to claim 1, is characterized in that: described multi-domain data acquisition module comprises:数据采集，包括多来源的物理场数据、模型生成数据及虚实融合数据，根据物理场类型的不同，采用不同的传感器获取数值信息，如温度场采用温度传感器，电磁场采用霍尔传感器，结构场采用压力传感器等；Data acquisition, including multi-source physical field data, model generation data, and virtual-real fusion data. According to different types of physical fields, different sensors are used to obtain numerical information, such as temperature sensors for temperature fields, Hall sensors for electromagnetic fields, and structural fields. pressure sensor, etc.;数据管理，利用直接采集到的数值信息可以产生繁衍数据信息，最终使得数据信息呈网状结构，并将数据进行备份、存储；Data management, using the directly collected numerical information can generate reproduction data information, and finally make the data information in a mesh structure, and back up and store the data;数据传输，多通道实时接收位置、速度、电流、旋转速度、负载与电机负载等数据类型，并将数据信息传输到数据处理模块和建模计算模块，实现各模块间的通信交互。Data transmission, multi-channel real-time reception of data types such as position, speed, current, rotation speed, load and motor load, and transmission of data information to the data processing module and modeling calculation module to realize the communication and interaction between the modules.4.根据权利要求1所述的一种复合型机床数字孪生监控系统，其特征在于：所述数据处理模块，接收实时多领域数据采集模块传输的运行状态数据信息，并对采集到的数值信息利用大数据技术进行深加工，以消除信号中存在噪声成分和冗余信息，重新构造新的数据集；之后，从重构后的数据集中提取时、频域统计特征，并利用能量值、熵值、谱峭度等指标对特征提取技术分解后的分量信号进行量化；最后对高维特征进行维数约简，为精准建模以及对运行系统预测性维护做准备。4. a kind of compound machine tool digital twin monitoring system according to claim 1, is characterized in that: described data processing module, receives the running state data information transmitted by the real-time multi-domain data acquisition module, and compares the collected numerical information Use big data technology for deep processing to eliminate noise components and redundant information in the signal, and reconstruct a new data set; after that, time and frequency domain statistical features are extracted from the reconstructed data set, and the energy value and entropy value are used. The component signals decomposed by the feature extraction technology are quantified using indicators such as spectral kurtosis, spectral kurtosis, etc.; finally, the dimension of high-dimensional features is reduced to prepare for accurate modeling and predictive maintenance of the operating system.5.根据权利要求1所述的一种复合型机床数字孪生监控系统，其特征在于：所述建模计算模块，接收实时多领域数据采集模块传输的运行状态数据信息，构建随运行环境变化的自适应模型，可精准监测机床的部件和整机性能；将机床历史维修数据中的故障模式注入三维物理模型和性能模型，构建出故障模型，可用于故障诊断与预测；将机床历史运行数据与性能模型结合并融合数据驱动方法，构建出性能预测模型，预测整机性能和剩余寿命；将局部线性化模型与机床运行状态环境模型融合并构建控制优化模型，可实现机床控制性能寻优，使机床在运行过程中发挥更好的性能。5. a kind of compound machine tool digital twin monitoring system according to claim 1, is characterized in that: described modeling calculation module, receives the running state data information transmitted by the real-time multi-domain data acquisition module, constructs the change with the running environment. The self-adaptive model can accurately monitor the performance of the components and the whole machine; inject the failure mode in the historical maintenance data of the machine tool into the three-dimensional physical model and performance model, and construct a failure model, which can be used for fault diagnosis and prediction; The performance model combines and integrates the data-driven method to construct a performance prediction model to predict the performance and remaining life of the whole machine; the local linearization model and the machine tool operating state environment model are integrated to build a control optimization model, which can realize the optimization of machine tool control performance and make Machine tools perform better during operation.6.根据权利要求1所述的一种复合型机床数字孪生监控系统，其特征在于：所述个性化服务决策模块，采用虚拟PLC、虚拟原型进行调试验证，成熟之后迁移到真实机床系统做对接，之后，基于物理PLC让虚拟原型系统与真实物理系统进行实时同步；6. A kind of compound machine tool digital twin monitoring system according to claim 1, is characterized in that: described personalized service decision-making module, adopts virtual PLC, virtual prototype to carry out debugging and verification, and migrates to real machine tool system to do docking after maturity , and then synchronize the virtual prototype system with the real physical system in real time based on the physical PLC;机床物理实体与虚拟模型进行数据信息交互，通过数据映射字典传入信息层，传输所用的接口为 OPC-UA接口，该接口可在保证不丢失通信数据的情况下统一模型数据，并支持复杂的数据模型，可实现多平台、多数据、多接口的通信。The physical entity of the machine tool and the virtual model exchange data information, and the data is transferred to the information layer through the data mapping dictionary. The interface used for transmission is the OPC-UA interface. This interface can unify model data without losing communication data and support complex The data model can realize multi-platform, multi-data and multi-interface communication.7.根据权利要求1所述的一种复合型机床数字孪生监控系统，其特征在于：所述模拟仿真平台，采用Untiy实时三维交互虚拟内容构建平台，可以进行精确物理模拟，也可以定制开发编辑接口以便二次开发。7. a kind of compound machine tool digital twin monitoring system according to claim 1, is characterized in that: described simulation simulation platform, adopts Untiy real-time three-dimensional interactive virtual content construction platform, can carry out accurate physical simulation, also can customize development and editing Interface for secondary development.8.根据权利要求5所述的一种复合型机床数字孪生监控系统，其特征在于：所述故障诊断与预测包括以下步骤：8. A kind of compound machine tool digital twin monitoring system according to claim 5, is characterized in that: described fault diagnosis and prediction comprise the following steps:S1、根据所述系统建模信息按照物理场的类型进行分类；S1, classify according to the type of physical field according to the system modeling information;S2、对其中一类的物理场的建模数据进行归一化处理；S2, normalize the modeling data of one of the physical fields;S3、设定神经网络的输出为该类物理场的故障类型及故障程度，将归一化后的数据通过神经网络进行训练，得到训练过的神经网络；S3. Set the output of the neural network as the failure type and failure degree of the physical field, and train the normalized data through the neural network to obtain a trained neural network;S4、重复S2-S3得到所有物理场对应的训练过的神经网络；S4. Repeat S2-S3 to obtain the trained neural network corresponding to all physical fields;S5、将实时采集的系统建模信息通过S1分类后，结合机床历史维修数据中的故障模式，构建出故障模型通过相应的类型的S4中相应训练过的神经网络进行故障诊断与预测，得到诊断与预测结果。S5. After classifying the system modeling information collected in real time through S1, combined with the failure mode in the historical maintenance data of the machine tool, a failure model is constructed to perform fault diagnosis and prediction through the corresponding neural network trained in S4 of the corresponding type, and the diagnosis is obtained. with predicted results.9.根据权利要求5所述的一种复合型机床数字孪生监控系统，其特征在于：所述故障故障诊断与预测采用的物理场根据机床系统运行状况进行选取与融合。9 . The composite machine tool digital twin monitoring system according to claim 5 , wherein the physical field used for fault diagnosis and prediction is selected and fused according to the operating conditions of the machine tool system. 10 .10.根据权利要求1所述的一种复合型机床数字孪生监控系统，其特征在于：所述监控过程包括以下步骤：10. A compound machine tool digital twin monitoring system according to claim 1, wherein the monitoring process comprises the following steps:A1、根据运行物理场的类型进行分类，绘制相应生产设备正常运行时的数据曲线；A1. Classify according to the type of operating physical field, and draw the data curve of the corresponding production equipment during normal operation;A2、将历史的故障信息按照生产设备的类型进行分类，绘制相应的历史数据曲线；A2. Classify the historical fault information according to the type of production equipment, and draw the corresponding historical data curve;A3、将历史数据曲线与相应的正常运行时的数据曲线进行对比，判断重合度，当所述历史数据曲线与相应的正常运行时的数据曲线基本正常，则对应历史数据曲线的生产设备为正常工作；否则为出现故障；A3. Compare the historical data curve with the corresponding data curve during normal operation to determine the degree of coincidence. When the historical data curve and the corresponding data curve during normal operation are basically normal, the production equipment corresponding to the historical data curve is normal work; otherwise, there is a failure;A4、根据故障对相应设备进行检修，并将故障类型在相应的实时数据曲线上进行标记，形成具有故障分析的数据曲线替换A1中的数据曲线；A4. Repair the corresponding equipment according to the fault, mark the fault type on the corresponding real-time data curve, and form a data curve with fault analysis to replace the data curve in A1;A5、重复A1-A4，得到具有故障分析的数据曲线；A5. Repeat A1-A4 to obtain a data curve with failure analysis;A6、将实时的工况信息按照生产设备的类型进行分类，绘制相应的实时数据曲线；A6. Classify the real-time working condition information according to the type of production equipment, and draw the corresponding real-time data curve;A7、将实时数据曲线与相应的具有故障分析的数据曲线进行对比，判断重合度，当所述实时数据曲线与相应的具有故障分析的数据曲线基本正常，则对应实时数据曲线的生产设备为正常工作；否则为出现故障，得到实时诊断结果，进而实现对机床运行过程的智能监控。A7. Compare the real-time data curve with the corresponding data curve with fault analysis to determine the degree of coincidence. When the real-time data curve and the corresponding data curve with fault analysis are basically normal, the production equipment corresponding to the real-time data curve is normal work; otherwise, there is a fault, and real-time diagnosis results are obtained, thereby realizing intelligent monitoring of the running process of the machine tool.

Images