Industrial production fault tracing analysis system based on digital twin

Abstract

The present application provides an industrial production fault tracing analysis system based on digital twins, which relates to the field of fault analysis technology. The system includes: a fault collaborative coupling analysis module for constructing a collaborative production equipment group in an industrial production workshop; a fault diffusion analysis module for generating multiple fault diffusion networks; an anomaly identification analysis module for establishing an anomaly identification model; a digital twin modeling module for constructing a fault diffusion twin; a fault tracing module for locating the risk fault source and the fault propagation node under a preset duration zone. The present application can solve the technical problem in the prior art that the accuracy of fault analysis is poor due to the lack of fault tracing and diffusion analysis of collaborative equipment. By performing fault diffusion analysis on the collaborative production equipment group and constructing a fault diffusion twin, the analysis of the fault diffusion and tracing process is realized, thereby improving the accuracy of fault tracing and preventing the technical effect of missed fault reports.

Description

Technical Field

[0001] The present application relates to the technical field of fault analysis, and in particular to an industrial production fault tracing and analysis system based on digital twins. Background Art

[0002] In today's industrial production environment, ensuring the stable operation of the production line and reducing downtime are crucial to improving production efficiency and reducing costs. With the development of virtual simulation technology, digital twin technology has been widely used in industrial production. Digital twin is a technology that creates a virtual copy of a physical entity to achieve real-time monitoring and analysis of the state, performance, and behavior of a physical entity.

[0003] However, the existing digital twin technology still has certain limitations in the fault tracing and analysis of industrial production. It is more used for real-time monitoring of faults and lacks fault tracing and diffusion analysis of collaborative equipment, resulting in poor accuracy of fault analysis and easy omissions.

[0004] In summary, the prior art has a technical problem in that the accuracy of fault analysis is poor due to the lack of fault tracing and diffusion analysis of collaborative equipment. Summary of the invention

[0005] The purpose of this application is to provide an industrial production fault tracing and analysis system based on digital twins to solve the technical problem in the prior art that the accuracy of fault analysis is poor due to the lack of fault tracing and diffusion analysis of collaborative equipment.

[0006] In view of the above problems, the present application provides an industrial production fault tracing and analysis system based on digital twins, the system comprising: a fault collaborative coupling analysis module, used to construct a collaborative production equipment group in an industrial production workshop, wherein the collaborative production equipment group is a set of production equipment with a fault collaborative coupling relationship in the industrial production workshop; a fault diffusion analysis module, used to collect historical fault transaction data sets of the collaborative production equipment group, perform fault diffusion analysis, and generate multiple fault diffusion networks, wherein any fault diffusion network includes a fault source, a fault feature-oriented network, and a fault continuous propagation node; an anomaly identification analysis module, used to perform anomaly identification analysis based on the fault feature-oriented network, and establish an anomaly identification model for the collaborative production equipment group; a digital twin modeling module, used to perform digital twin modeling based on the multiple fault diffusion networks, and construct a fault diffusion twin, and the fault diffusion twin is connected to the anomaly identification model; a fault tracing module, used to input the abnormal point set into the fault diffusion twin when the abnormal identification model outputs an abnormal point set, and locate the risk fault source and the fault propagation node under a preset duration zone.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] Construct a collaborative production equipment group in an industrial production workshop, wherein the collaborative production equipment group is a set of production equipment with a fault collaborative coupling relationship in the industrial production workshop; collect the historical fault transaction data set of the collaborative production equipment group, perform fault diffusion analysis, and generate multiple fault diffusion networks, wherein any fault diffusion network includes a fault source, a fault feature-oriented network, and a fault continuous propagation node; perform abnormal identification analysis based on the fault feature-oriented network, and establish an abnormal identification model for the collaborative production equipment group; perform digital twin modeling based on the multiple fault diffusion networks, and construct a fault diffusion twin, which is connected to the abnormal identification model; when the abnormal identification model outputs an abnormal point set, the abnormal point set is input into the fault diffusion twin, and the risk fault source and the fault propagation node under the preset continuous time zone are located. Therefore, by performing fault diffusion analysis on the collaborative production equipment group in the industrial production workshop, establishing a fault diffusion network, and using digital twin technology to construct a fault diffusion twin, the analysis of the fault diffusion and tracing process is realized, and the accuracy of fault tracing is improved, while preventing the technical effect of fault omission.

[0009] In addition, the above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0011] Figure 1 This is a schematic diagram of the structure of the industrial production fault tracing analysis system based on digital twins in this application;

[0012] Figure 2 This is a flow chart of generating multiple fault diffusion networks in the industrial production fault tracing and analysis system based on digital twins in this application.

[0013] Description of reference numerals:

[0014] Fault collaborative coupling analysis module 11, fault diffusion analysis module 12, anomaly identification and analysis module 13, digital twin modeling module 14, fault tracing module 15. DETAILED DESCRIPTION

[0015] This application solves the technical problem of poor accuracy of fault analysis in the prior art due to the lack of fault tracing and diffusion analysis of collaborative equipment by providing an industrial production fault tracing analysis system based on digital twins. By performing fault diffusion analysis on the collaborative production equipment group in the industrial production workshop, establishing a fault diffusion network, and using digital twin technology to construct a fault diffusion twin, the fault diffusion and tracing process is analyzed, thereby improving the accuracy of fault tracing and preventing the underreporting of faults.

[0016] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0017] Please see attached Figure 1 , the present application provides an industrial production fault tracing analysis system based on digital twins, the system comprising:

[0018] The fault collaborative coupling analysis module 11 is used to construct a collaborative production equipment group in an industrial production workshop, wherein the collaborative production equipment group is a set of production equipment having a fault collaborative coupling relationship in the industrial production workshop.

[0019] Specifically, the fault coordination coupling relationship refers to a set of production equipment in an industrial production workshop. When one of the equipment fails, it may affect the normal operation of other equipment because these equipment are interdependent in the production line. For example, the semi-finished product produced by one equipment may be the raw material of another equipment. If the former fails, the latter will also be abnormal due to lack of raw materials. Analyze the impact that each equipment failure in the industrial production workshop may have on other equipment, including the blockage or interruption of material flow, energy flow and information flow, to obtain a set of production equipment with a fault coordination coupling relationship, establish a coordinated production equipment group, and provide a basis for subsequent fault tracing analysis.

[0020] The fault diffusion analysis module 12 is used to collect the historical fault transaction data set of the collaborative production equipment group, perform fault diffusion analysis, and generate multiple fault diffusion networks, wherein any fault diffusion network includes a fault source, a fault feature-oriented network, and a fault continuous propagation node.

[0021] Specifically, any device in the collaborative production equipment group may fail during operation, and the spread of the failure may affect the production efficiency and safety of the entire equipment group. Therefore, it is necessary to first collect the historical failure transaction data set of the collaborative production equipment group, that is, the relevant data of all failures that occurred in the equipment group in the past period of time.

[0022] Fault diffusion analysis is the process of analyzing the propagation and impact of faults in a group of devices. By analyzing fault data, the source of the fault, i.e., the fault source, and the characteristic diffusion direction and diffusion characteristics of the fault, i.e., the diffusion characteristics of temperature, vibration, etc., can be identified. For example, the fault source causes the temperature of the device to rise, and the temperature is transferred to other devices, i.e., the fault feature-oriented network. The fault source may be a specific device or component, and the fault feature-oriented network describes the diffusion characteristics of the fault characteristics starting from the fault source.

[0023] A fault propagation node refers to a device that, if not handled in time after a fault occurs, will cause other devices to fail within a certain period of time. The identification of fault propagation nodes is crucial to preventing and controlling the spread of faults. Multiple fault propagation networks are generated based on the fault source, fault feature-oriented network, and fault propagation nodes. Each fault propagation network is the result of an analysis of a specific fault source, including a fault source, a fault feature-oriented network, and a fault propagation node.

[0024] The abnormality identification and analysis module 13 is used to perform abnormality identification and analysis based on the fault feature-oriented network and establish an abnormality identification model for the collaborative production equipment group.

[0025] Specifically, the fault feature-oriented network describes the characteristic diffusion direction and diffusion characteristics of the fault, that is, the diffusion characteristics of temperature, vibration, etc. For example, the fault source causes the temperature of the equipment to rise, and the temperature is transferred to other equipment. This network is usually composed of nodes and edges, where the nodes represent the equipment or components to which the temperature, vibration, etc. diffuse, and the edges represent the diffusion paths of the faults. The weight of the edges represents the characteristics of the temperature, vibration, etc. diffused to other equipment or components. The goal of abnormal identification analysis is to establish a model that can identify abnormal behavior in a collaborative production equipment group. First, features are extracted from the fault feature-oriented network, including the diffusion characteristics of each diffusion node (equipment or component), such as temperature, vibration, etc. Based on this, an existing machine learning model is selected, such as a support vector machine, a decision tree, a random forest or a neural network, etc., and the extracted diffusion characteristics of each diffusion node (equipment or component) are used as training data to construct an abnormal identification model. The training of the intelligent learning model is a common technical means used by those skilled in the art and will not be expanded here.

[0026] Deploy the trained anomaly recognition model to the collaborative production equipment group to monitor the status of the equipment in real time and identify potential abnormal behaviors.

[0027] The digital twin modeling module 14 is used to perform digital twin modeling based on the multiple fault diffusion networks to construct a fault diffusion twin, and the fault diffusion twin is connected to the abnormality identification model.

[0028] Specifically, a digital twin is a virtual model that is an exact replica of a physical entity. In the event of an abnormality in a collaborative production equipment group, the digital twin model will simulate the fault diffusion state of the equipment group, thereby providing timely warnings, preventing fault omissions caused by fault diffusion, and improving the accuracy of fault tracing.

[0029] First, the data of multiple fault diffusion networks are integrated into a unified model. The integrated data is used to build a digital twin model to generate a fault diffusion twin, which will simulate the actual operation of the equipment group, including the propagation and impact of the fault. The constructed fault diffusion twin is connected to the anomaly recognition model. In this way, when the anomaly recognition model detects an anomaly, the fault diffusion twin can perform fault tracing and fault diffusion analysis.

[0030] The fault tracing module 15 is used to input the abnormal point set into the fault diffusion twin when the abnormal identification model outputs the abnormal point set, and locate the risk fault source and the fault propagation node under the preset duration zone.

[0031] Specifically, when the anomaly recognition model outputs an abnormal point set, the points in the abnormal point set represent the locations where failures may occur in the collaborative production equipment group. Next, these abnormal point sets are input into the fault diffusion twin to locate the risk fault source and the fault propagation node under the preset duration zone.

[0032] The fault diffusion twin uses its internal simulation and prediction capabilities to analyze the propagation path and possible impact range of the fault in the equipment group based on the input abnormal point set. Exemplarily, by analyzing the abnormal point set, the fault diffusion twin can identify the initial source of the fault, that is, the risk fault source. This is the starting point of the fault propagation. The fault diffusion twin further analyzes the propagation of the fault within the preset duration time zone and identifies the fault propagation nodes. These nodes may be devices or components that are continuously affected after the fault occurs and may cause the fault to spread further. Among them, the preset duration time zone is set by professional and technical personnel in this field.

[0033] Taking the risk fault source and the fault propagation node under the preset duration zone as the fault tracing analysis result, reminder information or early warning information can be automatically generated to assist staff in maintenance, reduce production losses and improve production safety.

[0034] Furthermore, the fault collaborative coupling analysis module 11 is also used for:

[0035] Connect to the data sharing system of the industrial production workshop and retrieve the historical workshop fault record data set; perform fault homology aggregation analysis on the historical workshop fault record data set to generate a first homology fault data set; extract the faulty equipment in the first homology fault data set to establish the collaborative production equipment group.

[0036] Furthermore, the fault collaborative coupling analysis module 11 is also used for:

[0037] The historical workshop fault record data set includes multiple historical workshop fault record data, and any historical workshop fault record data includes faulty equipment and fault factors; clustering is performed based on the fault factors to generate multiple clusters of fault record data; cluster association aggregation is performed on the multiple clusters of fault record data to generate the first homologous fault data set.

[0038] Furthermore, the fault collaborative coupling analysis module 11 is also used for:

[0039] Extracting a first cluster of fault record data and a second cluster of fault record data; performing device similarity identification on the first cluster of fault record data and the second cluster of fault record data to generate a first inter-cluster similarity index; if the first inter-cluster similarity index is greater than a preset threshold, aggregating the first cluster of fault record data and the second cluster of fault record data to generate the first homologous fault data set.

[0040] Specifically, after connecting to the data sharing system of the industrial production workshop, historical workshop fault record data sets can be retrieved. These data sets contain fault records that occurred in various equipment in the workshop over a period of time in the past, and are important resources for analyzing fault modes and trends. The data sharing system is a platform used by existing industrial production workshops to store historical data and download data. The historical workshop fault record data set is subjected to fault homology aggregation analysis, that is, records with the same fault source are classified to identify those faults caused by the same root cause, even if these faults may occur in different devices or at different time points. Through this aggregation, the first homology fault data set can be generated. This data set contains all identified fault events with the same fault source. Having the same fault source means that the faults of different devices are caused by the same fault source, so these devices are associated, that is, the failure of any device may cause the failure of other devices.

[0041] Next, the faulty equipment is extracted from the first homologous fault data set. These equipment are the entities where the fault occurred, and a collaborative production equipment group is established. This facilitates the subsequent analysis of how the fault spreads in the equipment group, how to effectively prevent and control the fault, and improves the accuracy of fault tracing analysis. At the same time, it is convenient to analyze the impact of fault spread, facilitate timely fault maintenance, and ensure industrial production safety.

[0042] The steps of performing fault homology aggregation analysis on the historical workshop fault record data set are as follows: The historical workshop fault record data set contains multiple historical workshop fault record data, each record includes faulty equipment and fault factors. Faulty equipment refers to the specific equipment where the fault occurs, while the fault factor describes the root cause of the fault, that is, the fault source equipment. Clustering is performed based on fault factors, that is, fault records with similar fault factors are grouped together using existing clustering algorithms, thereby generating multi-cluster fault record data. Clustering is an unsupervised learning technique that can group data points into multiple clusters so that data points in the same cluster are similar to each other, while data points in different clusters are as different as possible.

[0043] The purpose of cluster-association aggregation of multi-cluster fault record data is to identify those fault records that may have associated fault factors, although they have different fault factors, and to discover the deep connection between faults, thereby generating the first homologous fault data set. This data set contains all identified fault events with the same or similar root causes, which is convenient for analyzing the causes and diffusion mechanisms of faults, so as to take effective measures for fault tracing and diffusion analysis, and facilitate timely early warning and maintenance.

[0044] The specific steps of performing inter-cluster association aggregation on the multiple clusters of fault record data to generate the first homologous fault data set include: randomly extracting any two clusters of fault record data from the multiple clusters of fault record data, and recording them as the first cluster of fault record data and the second cluster of fault record data. Further, the two clusters of fault record data are subjected to equipment similarity identification to determine whether the faulty devices in the two clusters of fault records are similar, that is, whether they may fail due to the same or similar fault reasons. Equipment similarity identification can be accomplished by comparing the features of the faulty devices in the two clusters, which may include the type of equipment, frequency of use, maintenance history, fault mode, etc. By analyzing these features, the first inter-cluster similarity index can be calculated, which quantifies the degree of similarity of the faulty devices between the first cluster and the second cluster.

[0045] If the similarity index between the first clusters is greater than the preset threshold, it means that the faulty devices in the two clusters of fault record data are similar enough and there may be common faults. In this case, the first cluster of fault record data and the second cluster of fault record data can be aggregated to generate the first common fault data set. This data set will contain all fault records in the two clusters, thus providing a more comprehensive data basis for subsequent fault analysis and prevention.

[0046] Further, as attached Figure 2 As shown, the fault diffusion analysis module 12 is also used for:

[0047] The historical fault transaction data set is divided into fault source identification data, and multiple fault diffusion transaction data sets corresponding to the multiple fault sources are generated; the first fault source of the first fault diffusion transaction data set is extracted, and the first fault source is used as the central node to perform fault feature diffusion fitting to generate a first fault feature-oriented network, wherein the first fault feature-oriented network includes multiple device nodes, and the multiple device nodes have sensing features that can identify faults; the first fault source is used as the central node, and based on the first fault diffusion transaction data set, the fault continuous fermentation node is connected to generate a first fault continuous propagation node; the first fault source, the first fault feature-oriented network and the first fault continuous propagation node are combined to generate a first fault diffusion network, and the network is added to the multiple fault diffusion networks.

[0048] Furthermore, the fault diffusion analysis module 12 is also used for:

[0049] Taking the first fault source as the central node, based on the first fault diffusion transaction data set, and with the continuous increase of the fault fermentation time as a constraint, a fault node time sequence is constructed; wherein any node in the fault node time sequence contains at least one device; and the first fault continuous propagation node is generated using the fault node time sequence.

[0050] Specifically, the historical fault transaction data set is divided by fault source identification, that is, the historical fault transaction data set is classified according to the fault source. In this way, multiple fault diffusion transaction data sets corresponding to multiple fault sources can be generated, and each data set contains all fault events related to a specific fault source. Next, the corresponding fault source is extracted from the first fault diffusion transaction data set and recorded as the first fault source. This fault source is the starting point of fault propagation. With the first fault source as the central node, fault feature diffusion fitting is performed, that is, the process of fault propagation from the fault source to other devices is analyzed, and the diffusion path of the fault feature is simulated. In this way, a first fault feature-guided network can be generated. This network contains multiple device nodes, and each device node has sensing features that can identify faults. These features may include changes in parameters such as temperature, vibration, and current.

[0051] Then, taking the first fault source as the central node, based on the first fault diffusion transaction data set, the fault continuous fermentation node connection is performed to identify the device nodes that are still continuously affected after the fault occurs and may cause the fault to spread further, that is, the first fault continuous propagation node. Finally, the first fault source, the first fault feature-oriented network and the first fault continuous propagation node are combined to generate the first fault diffusion network. This network comprehensively describes how the fault propagates between devices through a specific path starting from the fault source, and which devices may be continuously affected by the fault. The generated first fault diffusion network is added to multiple fault diffusion networks for more comprehensive analysis and modeling. The method of generating the first fault diffusion network is used to generate multiple fault diffusion networks corresponding to different fault sources, which provides support for subsequent digital twin modeling, facilitates fault tracing and diffusion analysis, and timely fault repair, and improves the accuracy of fault tracing analysis.

[0052] The step of generating the first fault continuous propagation node specifically includes: taking the first fault source as the central node, based on the first fault diffusion transaction data set, constructing the fault node time series, that is, analyzing the propagation and continuous fermentation of the fault in time. The construction of this time series is constrained by the continuous increase of the fault continuous fermentation time, that is, when analyzing the fault propagation, focus on those time nodes and devices that are continuously affected after the fault occurs and cause the fault to spread over time.

[0053] In the fault node sequence, each node represents a specific time point and contains at least one device. At what point in time are these devices affected by the fault source, causing themselves to fail? By analyzing these nodes, the path of the continued spread of the fault can be identified. Specifically, the time is divided into a series of continuous time intervals. In each time interval, the spread of the fault from the fault source to other devices is analyzed, and the affected devices are recorded. For devices that fail in continuous time intervals or whose fault characteristics continue to strengthen, they are marked as fault continued fermentation nodes. All identified fault continued fermentation nodes are connected in chronological order to form a fault node sequence. This time series network shows the propagation and continuous impact of the fault in the device group over time. All nodes in the fault node sequence are used as the first fault continued propagation nodes to facilitate subsequent fault diffusion analysis, ensure the accuracy of fault analysis, and prevent underreporting caused by fault diffusion.

[0054] Furthermore, the digital twin modeling module 14 is also used for:

[0055] A twin network topology structure is constructed based on the multiple fault diffusion networks; modeling data is generated for the multiple fault diffusion networks, and fault diffusion and traceability reverse verification are performed using the modeling data and the twin network topology structure to generate the fault diffusion twin, which has a communication port; based on the communication port, the fault diffusion twin is connected to the abnormality identification model.

[0056] Specifically, a twin network is a digital model that simulates systems and processes in the physical world, which can reflect the state and behavior of physical systems in real time. To build a twin network topology based on multiple fault diffusion networks, first take each device in the multiple fault diffusion networks as a node and establish a corresponding digital copy in the twin network. According to the fault feature-oriented network and the fault continuous propagation node in the multiple fault diffusion networks, establish the connection relationship between the nodes, and thus obtain the twin network topology.

[0057] Next, the modeling data and twin network topology are used to verify the fault diffusion and traceability. Modeling data is usually obtained based on historical fault data, equipment performance data, and operation records, and is used to train the fault diffusion twin so that it can accurately predict the fault diffusion path. The modeling data and twin network topology are used to verify the fault diffusion and traceability to ensure that the fault diffusion twin can accurately reflect the fault diffusion in the actual system and can effectively trace the source of the fault.

[0058] The verification process includes: using the fault diffusion twin to simulate the diffusion process of the fault in the twin network, and through the reverse analysis of the model, trying to find the source of the specific fault. Through the above steps, a fault diffusion twin can be generated. This twin has a communication port and can exchange data and information with external systems. Such a fault diffusion twin can monitor the production environment in real time, predict potential failures, and help enterprises take preventive measures to improve production efficiency and safety. Based on the communication port, the fault diffusion twin is connected to the anomaly identification model to facilitate timely fault tracing analysis, thereby achieving faster and more accurate fault response.

[0059] Furthermore, the fault tracing module 15 is also used for:

[0060] The first abnormal point in the abnormal point set is input into the fault diffusion twin, and the source is traced back to generate the risk fault source; the first abnormal point set is used as the diffusion starting node, and the fault diffusion analysis is performed in the preset duration zone through the fault diffusion twin, and the fault propagation node is output.

[0061] When the first abnormal point in the abnormal point set is input into the fault diffusion twin, the fault diffusion twin is used to trace the source and identify the risk fault source that caused the abnormal point. Tracing the source is a reverse analysis process that determines the most likely fault source by analyzing the association between the abnormal point and other device nodes in the twin network and historical fault data.

[0062] Taking the first abnormal point set as the diffusion starting node, the fault diffusion analysis is performed in the preset duration zone through the fault diffusion twin, that is, simulating the fault starting from the abnormal point and spreading to other device nodes within a specific time. Set the abnormal point set as the starting point of fault diffusion in the fault diffusion twin. Then, according to the connection relationship and fault diffusion rules of the device nodes in the fault diffusion twin, simulate the diffusion process of the fault in the preset duration zone. Finally, record the device nodes affected by the fault during the simulation process, which are the fault propagation nodes. Through the above steps, risk fault sources and fault propagation nodes can be generated to realize fault tracing and diffusion analysis, improve the accuracy of fault analysis, and ensure the operation safety of workshop equipment.

[0063] In summary, the industrial production fault tracing analysis system based on digital twins provided by this application has the following technical effects:

[0064] Construct a collaborative production equipment group in an industrial production workshop, wherein the collaborative production equipment group is a set of production equipment with a fault collaborative coupling relationship in the industrial production workshop; collect the historical fault transaction data set of the collaborative production equipment group, perform fault diffusion analysis, and generate multiple fault diffusion networks, wherein any fault diffusion network includes a fault source, a fault feature-oriented network, and a fault continuous propagation node; perform abnormal identification analysis based on the fault feature-oriented network, and establish an abnormal identification model for the collaborative production equipment group; perform digital twin modeling based on the multiple fault diffusion networks, and construct a fault diffusion twin, which is connected to the abnormal identification model; when the abnormal identification model outputs an abnormal point set, the abnormal point set is input into the fault diffusion twin, and the risk fault source and the fault propagation node under the preset continuous time zone are located. Therefore, by performing fault diffusion analysis on the collaborative production equipment group in the industrial production workshop, establishing a fault diffusion network, and using digital twin technology to construct a fault diffusion twin, the analysis of the fault diffusion and tracing process is realized, and the accuracy of fault tracing is improved, while preventing the technical effect of fault omission.

[0065] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. 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 application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0066] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. The industrial production fault tracing analysis system based on digital twins is characterized by: include: A fault collaborative coupling analysis module is used to construct a collaborative production equipment group in an industrial production workshop, wherein the collaborative production equipment group is a set of production equipment with a fault collaborative coupling relationship in the industrial production workshop; A fault diffusion analysis module is used to collect the historical fault transaction data set of the collaborative production equipment group, perform fault diffusion analysis, and generate multiple fault diffusion networks, wherein any fault diffusion network includes a fault source, a fault feature-oriented network, and a fault continuous propagation node; An abnormality identification and analysis module, used to perform abnormality identification and analysis based on the fault feature-oriented network and establish an abnormality identification model for the collaborative production equipment group; A digital twin modeling module, used for performing digital twin modeling based on the multiple fault diffusion networks to construct a fault diffusion twin, wherein the fault diffusion twin is connected to the anomaly identification model; A fault tracing module is used to input the abnormal point set into the fault diffusion twin when the abnormal recognition model outputs the abnormal point set, and locate the risk fault source and the fault propagation node under the preset duration zone; Wherein, the fault diffusion analysis module is also used for: Perform fault source identification and division on the historical fault transaction data set to generate multiple fault diffusion transaction data sets corresponding to multiple fault sources; Extracting a first fault source of a first fault diffusion transaction data set, taking the first fault source as a central node, performing fault feature diffusion fitting, and generating a first fault feature-oriented network, wherein the first fault feature-oriented network includes a plurality of device nodes, and the plurality of device nodes have sensing features that can identify faults; Taking the first fault source as a central node, and based on the first fault diffusion transaction data set, connecting the fault continuous fermentation nodes to generate a first fault continuous propagation node; generating a first fault diffusion network by combining the first fault source, the first fault feature-oriented network and the first fault continuous propagation node, and adding the first fault diffusion network to the multiple fault diffusion networks; Wherein, the digital twin modeling module is also used for: constructing a twin network topology structure based on the multiple fault diffusion networks; Generate modeling data for the multiple fault diffusion networks, perform fault diffusion and source tracing and reverse verification using the modeling data and the twin network topology structure, and generate the fault diffusion twin, wherein the fault diffusion twin has a communication port; Based on the communication port, the fault diffusion twin is connected to the anomaly identification model.

2. The industrial production fault tracing analysis system based on digital twins according to claim 1 is characterized in that: The fault collaborative coupling analysis module is also used for: Connecting to the data sharing system of the industrial production workshop to retrieve the historical workshop fault record data set; Performing a fault homology aggregation analysis on the historical workshop fault record data set to generate a first homology fault data set; The faulty equipment in the first homologous fault data set is extracted to establish the collaborative production equipment group.

3. The industrial production fault tracing analysis system based on digital twins according to claim 2 is characterized in that: The fault collaborative coupling analysis module is also used for: The historical workshop fault record data set includes a plurality of historical workshop fault record data, and any historical workshop fault record data includes a faulty device and a fault factor; Clustering is performed based on the fault factors to generate multiple clusters of fault record data; The multi-cluster fault record data is aggregated by cluster association to generate the first homologous fault data set.

4. The industrial production fault tracing analysis system based on digital twins according to claim 3 is characterized in that: The fault collaborative coupling analysis module is also used for: Extracting a first cluster of fault record data and a second cluster of fault record data; Performing device similarity identification on the first cluster of fault record data and the second cluster of fault record data to generate a first inter-cluster similarity index; If the first inter-cluster similarity index is greater than a preset threshold, the first cluster fault record data and the second cluster fault record data are aggregated to generate the first homologous fault data set.

5. The industrial production fault tracing analysis system based on digital twins according to claim 1 is characterized in that: The fault diffusion analysis module is also used for: Taking the first fault source as the central node, based on the first fault diffusion transaction data set, and taking the continuous increase of the fault fermentation time as a constraint, construct a fault node time series; Wherein, any node in the fault node sequence includes at least one device; A first fault continuous propagation node is generated according to the fault node time sequence.

6. The industrial production fault tracing analysis system based on digital twins according to claim 1 is characterized in that: The fault tracing module is also used for: Inputting the first abnormal point in the abnormal point set into the fault diffusion twin, tracing back the source, and generating the risk fault source; Taking the first abnormal point set as the diffusion starting node, the fault diffusion analysis is performed in the preset duration zone through the fault diffusion twin, and the fault propagation node is output.

Images