Method and device for building equipment operation and maintenance digital twin model based on cloud and fog edge collaborative driving

Abstract

The present application relates to the technical field of high-end equipment operation and maintenance intelligentization, a cloud-fog edge collaborative driving equipment operation and maintenance digital twin model building method and device, by establishing a cloud-fog edge collaborative driving data resource integration model, determining the key indicators of the twin model, completing operation and maintenance data integration and knowledge sharing; a component twin body dynamic self-adaptive modeling mechanism and a cross-domain equipment twin body self-adaptive construction mechanism are established, and a consistency evaluation criterion of monitoring data and digital twin model data is established; based on the consensus maximization model and the game confrontation weight learning method, the equipment twin body model after scene migration is self-consistent with the operation and maintenance scene; through the integration of historical database-operation and maintenance knowledge base-model library-real-time twin database, the cloud-fog edge collaborative driving cross-domain high-end equipment operation and maintenance digital twin modeling system is efficiently run. The present application provides a key technical system for solving the autonomous controllability of the cross-domain high-end equipment operation and maintenance whole-process digital twin modeling system.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance technology for high-end equipment, specifically to a method and apparatus for building a digital twin model of equipment operation and maintenance driven by cloud-fog-edge collaboration. Background Technology

[0002] Addressing the demand for intelligent operation and maintenance of high-end equipment such as urban rail vehicles, large drones, and CNC machine tools, key challenges currently exist, including multi-source heterogeneous information across different fields, complex and variable operating conditions, and ever-increasing demands for safety and reliability. Digital twins, a key enabling technology driven by the integration of next-generation information technologies, fully utilize physical models, sensor updates, and historical data updates to integrate simulation models from multiple disciplines, physical quantities, scales, and probabilities. They feature high real-time and high-precision mapping between virtual and real spaces. Furthermore, with the deep integration of digital twin technology concepts into the entire equipment operation and maintenance process, its forward-looking intelligent technology concepts, high-fidelity visualization, faithful mapping between virtual and real spaces, and iterative optimization control provide solutions to the aforementioned challenges. Developing independently controllable digital twin industrial software for the entire intelligent operation and maintenance process of high-end equipment is an inevitable trend, meeting the needs of building intelligent digital twin systems for high-end equipment operation and maintenance.

[0003] Currently, most digital twin systems for the operation and maintenance of high-end equipment are customized for specific scenarios and are fixed and unchangeable. The developed systems have characteristics such as specific domains, specific scenarios, specific models, and specific data. When the operation and maintenance objects change or new changes are generated in the operation and maintenance scenario data, the digital twin system needs to be manually modified again. As a result, no independent and controllable core technology has been formed, which affects the formation of a key technology system for the intelligent operation and maintenance of high-end equipment. Summary of the Invention

[0004] This invention provides the following technical solution:

[0005] According to a first aspect of the present invention, the present invention claims protection for a method for building a digital twin model of equipment operation and maintenance driven by cloud-fog-edge collaboration, characterized in that it includes:

[0006] Establish a cloud-fog-edge collaborative data resource integration model, determine key indicators of the twin model, and complete the integration and knowledge sharing of operation and maintenance data;

[0007] Establish a dynamic adaptive modeling mechanism for component twins and an adaptive construction mechanism for cross-domain equipment twins; establish consistency evaluation criteria between monitoring data and digital twin model data.

[0008] Based on the consensus maximization model and the game-theoretic adversarial weight learning method, the self-consistency between the equipment twin model after scenario migration and the operation and maintenance scenario is achieved.

[0009] By integrating historical databases, operation and maintenance knowledge bases, model libraries, and real-time twin databases to create big data twins, a cross-domain high-end equipment operation and maintenance digital twin modeling system driven by cloud, fog, and edge collaboration is implemented to ensure efficient operation.

[0010] Furthermore, the establishment of a cloud-fog-edge collaborative data resource integration model, determination of key indicators for the twin model, and completion of operation and maintenance data integration and knowledge sharing also include:

[0011] The cloud refers to a computing platform with high computing power and large storage space; the edge refers to an edge platform deployed on-site with fast computing and real-time data response at the nanosecond level; and the fog refers to a computing platform between the cloud and the edge with medium computing power and medium storage space.

[0012] Based on cloud-fog-edge collaboration, a unified communication architecture for a twin database of the entire equipment operation and maintenance process is established. A parallel computing framework based on cloud computing is built, distributed data storage technology is used to store information resources, high-performance distributed processing and analysis are performed, and a data resource integration model is established.

[0013] Construct a cloud-fog-edge collaborative data information sharing architecture, and implement an edge node model training iteration mechanism based on federated learning to analyze the resource constraints, edge heterogeneity, and dynamic environment faced by edge node computing, and determine the key indicators of the twin model;

[0014] Data knowledge enhancement methods utilize transfer learning to reduce the gap between different data distributions, enabling the network to adapt to changing operating conditions based on a good modeling and prediction of the temporal characteristics of sensor data.

[0015] For massive equipment operation and maintenance databases and real-time data, knowledge extraction and mining technologies are used to extract entities, relationships and attributes from massive equipment operation and maintenance data. Knowledge fusion and processing are used to disambiguate and integrate the extracted entities and merge them with structured data. After quality assessment, reasoning verification and updating, a knowledge graph is formed to obtain the value contained in the data and complete the knowledge value-added and iteration of the data in the database.

[0016] The minimum path of a knowledge graph is defined by building data entities, relationships, and attributes, and industry knowledge graphs are built by combining historical experience.

[0017] Furthermore, the establishment of a dynamic adaptive modeling mechanism for component twins and an adaptive construction mechanism for cross-domain equipment twins, as well as the establishment of consistency evaluation criteria between monitoring data and digital twin model data, also includes:

[0018] Establish a dynamic adaptive modeling mechanism for component twins, construct a target domain composed of real-time data and a sample space with historical data as the source domain, use adversarial learning to measure the inter-domain loss, and measure the data feature distribution of the source and target domains to achieve unsupervised learning tasks using real-time data, thereby improving the accuracy of data model prediction of high-end equipment evolution data.

[0019] The model update strategy based on cloud-fog collaboration deploys training tasks in the cloud and utilizes the cloud's ample computing power to achieve rapid iterative updates of the model.

[0020] Establish a consistency evaluation criterion between component models and equipment twins, study the uncertainty quantification method of component and system digital twin models under the combined effect of prior uncertainty and model uncertainty, and establish a consistency evaluation criterion between monitoring data and digital twin model data;

[0021] Establish a common mechanism for constructing digital twin models of cross-domain high-end equipment, and develop a method for constructing equipment twins based on adaptive weighting of multi-scale component model parameters and deep reinforcement learning for cross-domain equipment. Establish an adaptive construction mechanism for cross-domain equipment twins to improve the speed of equipment self-consistent twin modeling.

[0022] Furthermore, the method of achieving self-consistency between the equipment twin model after scenario migration and the operation and maintenance scenario based on the consensus maximization model and game-theoretic adversarial weight learning method specifically includes:

[0023] The operation and maintenance scenario self-awareness of high-end equipment digital twins is achieved through prior knowledge embedding training, enabling scenario awareness and interpretation during the twin modeling process;

[0024] Based on a knowledge rule base, environmental indicator features are screened and classified. Through a knowledge intervention mechanism, the equipment digital twin achieves self-learning and self-perception of environmental variables.

[0025] Real-time online sensing with the control system is achieved through online communication, enabling the equipment twin to learn and sense environmental variables on its own.

[0026] Digital twin model iteration and simulation compensation to establish a hybrid field digital twin simulation platform;

[0027] To meet different operational and maintenance needs, the data scale and model complexity of the fusion model are adaptively determined to obtain an efficient digital twin model iteration and a proposed compensation strategy.

[0028] For equipment twin modeling mechanisms under the same scenario and different typical operation and maintenance scenarios, based on consensus maximization model and game adversarial weight learning method, the equipment twin model after scenario migration is made self-consistent with the operation and maintenance scenario.

[0029] By using self-sensing environmental variable data, the system achieves environmental variable transfer under complex and changing operating conditions based on transfer learning theory, and adapts the scenario operation and maintenance decision-making algorithm to achieve self-learning decision-making.

[0030] According to a second aspect of the present invention, the present invention claims protection for a cloud-fog edge-coordinated equipment operation and maintenance digital twin model building device, characterized in that it comprises:

[0031] The data integration and knowledge sharing module establishes a cloud-fog-edge collaborative data resource integration model, determines key indicators of the twin model, and completes operation and maintenance data integration and knowledge sharing.

[0032] The common mechanism module of the digital twin model establishes a dynamic adaptive modeling mechanism for component twins and an adaptive construction mechanism for cross-domain equipment twins, and establishes a consistency evaluation criterion between monitoring data and digital twin model data;

[0033] The adaptive modeling module for operation and maintenance scenarios, based on the consensus maximization model and the game adversarial weight learning method, achieves self-consistency between the equipment twin model after scenario migration and the operation and maintenance scenario.

[0034] The cloud-fog-edge collaborative driving module integrates twin big data through historical database, operation and maintenance knowledge base, model library and real-time twin database to execute the efficient operation of the cross-domain high-end equipment operation and maintenance digital twin modeling system driven by cloud-fog-edge collaboration.

[0035] The cloud-fog-edge collaborative driving equipment operation and maintenance digital twin model building device is used to execute the cloud-fog-edge collaborative driving equipment operation and maintenance digital twin model building method.

[0036] The multiple technical solutions provided in this invention have at least the following technical effects or advantages:

[0037] This invention addresses the challenges of coordinating and sharing maintenance knowledge due to the spatiotemporal coupling and multi-granularity, high-generalization characteristics of large-scale maintenance data in the current demand for intelligent twin models of the entire maintenance process of high-end equipment. It proposes a cloud-fog-edge collaborative-driven maintenance data integration and knowledge sharing module to solve the problem of intelligent maintenance of high-end equipment, focusing on the full-process, full-state data information of high-end equipment maintenance. To address the difficulty in unifying the architecture of twin modeling systems caused by the complex heterogeneity of cross-domain high-end equipment, this invention proposes a common mechanism module for constructing cross-domain high-end equipment digital twin models, focusing on the consistency of multi-dimensional behaviors of component models and system-level twins. Furthermore, to solve the challenge of adaptive modeling of equipment twins in high-security, complex application scenarios with multi-objective maintenance, this invention proposes an adaptive modeling module for equipment digital twin maintenance scenarios. The integration of these three modules enables the construction of a cross-domain high-end equipment maintenance full-process digital twin modeling system, providing a key technical system for achieving independent controllability of such systems.

[0038] The above description is an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the description, and to make the above content and other objects, features and advantages of the present invention more obvious and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0039] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the accompanying drawings without creative effort.

[0040] Figure 1 This is a flowchart illustrating the process of building a digital twin model for equipment operation and maintenance driven by cloud-fog-edge collaboration in this embodiment of the invention.

[0041] Figure 2 This is a schematic diagram of the twin model construction mechanism of the pantograph-contact network system of urban rail vehicles in the cloud-fog-edge collaborative driving method for building a digital twin model of equipment operation and maintenance in an embodiment of the present invention.

[0042] Figure 3 This is a structural module diagram of the equipment operation and maintenance digital twin model building device driven by cloud-fog-edge collaboration in an embodiment of the present invention. Detailed Implementation

[0043] The objectives, technical solutions, and advantages of this invention are more clearly expressed in the appendix. Figure 1 The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. The components of the embodiments of the present invention described and shown in the accompanying drawings can be arranged and designed in various different configurations.

[0044] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are all within the scope of protection of the present invention.

[0045] As attached Figure 1 As shown, the specific implementation steps of the cloud-fog-edge collaborative driving method for building a digital twin model of equipment operation and maintenance in the first embodiment of the present invention are as follows:

[0046] Step 1: Establish a cloud-fog-edge collaborative data resource integration model, determine key indicators of the twin model, and complete operation and maintenance data integration and knowledge sharing;

[0047] Step 1.1: "Cloud" refers to a computing platform with "high computing power and large storage space," "edge" refers to an edge platform deployed on-site with rapid computing and real-time data response at the nanosecond level, and "fog" refers to a computing platform between cloud and edge with medium computing power and medium storage space. Based on cloud-fog-edge collaboration, a unified communication architecture for a twin database of the entire equipment operation and maintenance process is established. A parallel computing framework based on cloud computing is built, distributed data storage technology is used to store information resources, high-performance distributed processing and analysis are performed, and a data resource integration model is established.

[0048] Step 1.2: Construct a cloud-fog-edge collaborative data information sharing architecture. Based on the federated learning edge node model (mainly including the mechanism model of equipment components, the real-time control behavior model of equipment, etc.) training iteration mechanism, realize the resource constraints, edge heterogeneity and environmental dynamics analysis faced in the edge node computing process, determine the key indicators of the twin model (mainly including the consistency of virtual and real space data, the accuracy of virtual and real interaction real-time control commands, etc.), and maximize the utilization of the computing and storage resources of the twin model data.

[0049] Step 1.3: Data knowledge enhancement methods utilize transfer learning to reduce the gap between different data distributions, enabling the network to adapt to varying operating conditions based on a good modeling and prediction of the temporal characteristics of sensor data. For massive equipment operation and maintenance databases and real-time data, knowledge extraction and mining techniques are used to extract entities, relationships, and attributes from the massive equipment operation and maintenance data. Knowledge fusion and processing are then used to disambiguate and integrate the extracted entities, and after fusing structured data, quality assessment, inference verification, and updates are performed to form a knowledge graph. This process extracts the value contained in the data, realizing knowledge value-added and iterative processing within the database. The minimum path of the knowledge graph is defined by constructing data entities, relationships, and attributes, and industry knowledge graphs are built by combining historical experience.

[0050] Step 2: Establish a dynamic adaptive modeling mechanism for component twins and an adaptive construction mechanism for cross-domain equipment twins, and establish consistency evaluation criteria between monitoring data and digital twin model data;

[0051] Step 2.1: Establish a dynamic adaptive modeling mechanism for component twins, construct a target domain composed of real-time data and a sample space using historical data as the source domain, utilize adversarial learning to measure inter-domain loss, and analyze the data feature distribution of the source and target domains. This enables unsupervised learning tasks using real-time data, improving the accuracy of high-end equipment evolution data prediction in the data model. Figure 3The model of the pantograph-overhead contact system of the urban rail vehicle shown is an adaptive modeling mechanism; the model update strategy based on cloud-fog collaboration deploys the training task in the cloud and uses the cloud's sufficient computing power to achieve rapid iterative updates of the model.

[0052] Step 2.2: Establish a consistency evaluation criterion between component models and equipment twin behavior, study the uncertainty quantification method of component and system digital twin models under the combined effect of prior uncertainty and model uncertainty, and establish a consistency evaluation criterion between monitoring data and digital twin model data.

[0053] Step 2.3: Establish a common mechanism for constructing digital twin models of cross-domain high-end equipment, and develop a method for constructing equipment twins based on adaptive weighting of multi-scale component model parameters and deep reinforcement learning for cross-domain equipment. Establish an adaptive construction mechanism for cross-domain equipment twins to improve the speed of self-consistent equipment twin modeling.

[0054] Step 3: Based on the consensus maximization model and the game adversarial weight learning method, achieve self-consistency between the equipment twin model after scenario migration and the operation and maintenance scenario;

[0055] Step 3.1: Self-awareness of operation and maintenance scenarios for high-end equipment digital twins. Through prior knowledge embedding training, scenario awareness and interpretation during the twin modeling process are achieved. Environmental indicator features are screened and classified based on a knowledge rule base. A knowledge intervention mechanism enables the equipment digital twin to self-learn and self-awareness of environmental variables. For example... Figure 2 The pantograph-overhead contact system of the urban rail vehicle shown uses images and sensors to perceive information about the equipment's operating environment and achieves real-time online perception with the control system through online communication, so as to enable the equipment twin to self-learn and self-perceive environmental variables.

[0056] Step 3.2: Digital twin model iteration and simulation compensation, establishing a hybrid field digital twin simulation platform. To address different operational needs, the data scale and model complexity of the fusion model are adaptively determined to achieve efficient digital twin model iteration and simulation compensation strategies.

[0057] Step 3.3: Equipment twin modeling mechanism under different scenarios. For different typical operation and maintenance scenarios, based on the consensus maximization model and game-theoretic adversarial weight learning method, the equipment twin model after scenario migration is made self-consistent with the operation and maintenance scenario, such as... Figure 3 As shown, by using self-sensing environmental variable data, the system achieves the transfer of environmental variables under complex changing operating conditions based on transfer learning theory, and adapts the scenario operation and maintenance decision-making algorithm to achieve self-learning decision-making.

[0058] Step 4: Integrate twin big data through historical database, operation and maintenance knowledge base, model library and real-time twin database to run the cross-domain high-end equipment operation and maintenance digital twin modeling system driven by cloud, fog and edge collaboration efficiently.

[0059] As a preferred real-time solution of the present invention, the cloud-fog-edge collaborative-driven operation and maintenance data integration and knowledge sharing module is used to establish a general information fusion of model library-knowledge base-database for the entire equipment operation and maintenance process. In order to meet the high real-time requirements of monitoring data and model data in the high-end equipment digital twin modeling process, it monitors the data distribution based on the time delay sensitivity coefficient, and uses transfer learning data fusion methods based on BLSTM deep neural networks to migrate and fuse data under different operating conditions, realize knowledge iterative value-added, and achieve efficient computing of high-end equipment operation and maintenance twin big data to support the operation of the high-end equipment intelligent operation and maintenance digital twin modeling system.

[0060] As a preferred real-time solution of this invention, the common mechanism module for constructing cross-domain high-end equipment digital twin models is used for self-consistent modeling of whole-machine twins of equipment in different fields, and for the coordination of dynamic response parameters of different component models, establishing a consistency mechanism between the dynamic response of key component models and the whole-machine twin of equipment; mainly by establishing an adjacency matrix and topological relationship transmission graph of key physical characteristics between interface-component-system, constructing a scale transmission and mapping method based on the principle of collaborative optimization and variable complexity model, establishing a behavioral consistency evaluation criterion between key component models and system-level twin models; and realizing the construction of self-consistent digital twins of equipment based on deep reinforcement learning, establishing an adaptive construction mechanism for cross-domain equipment twins, and improving the speed of self-consistent twin modeling of equipment.

[0061] As a preferred real-time solution of the present invention, the equipment digital twin operation and maintenance scenario adaptive modeling module is used to determine the current environmental characteristics, form preliminary environmental constraints, establish a scenario self-perception model of the equipment twin model, realize real-time perception of scenario environmental variables using model fusion technology, set dynamic thresholds to realize data dimensionality reduction and model dimensionality reduction of the main feature parameters and environmental constraints of multi-source heterogeneous models, construct the equipment twin inverse model based on autoencoders, generative adversarial networks, etc., and realize environmental self-consistent modeling of high-end equipment digital twin through dynamic iteration and simulation compensation of the twin model.

[0062] According to a second embodiment of the present invention, referring to the appendix Figure 3 This invention claims protection for a cloud-fog edge-driven equipment operation and maintenance digital twin model building device, characterized in that it includes:

[0063] The data integration and knowledge sharing module establishes a cloud-fog-edge collaborative data resource integration model, determines key indicators of the twin model, and completes operation and maintenance data integration and knowledge sharing.

[0064] The common mechanism module of the digital twin model establishes a dynamic adaptive modeling mechanism for component twins and an adaptive construction mechanism for cross-domain equipment twins, and establishes a consistency evaluation criterion between monitoring data and digital twin model data;

[0065] The adaptive modeling module for operation and maintenance scenarios, based on the consensus maximization model and the game adversarial weight learning method, achieves self-consistency between the equipment twin model after scenario migration and the operation and maintenance scenario.

[0066] The cloud-fog-edge collaborative driving module integrates twin big data through historical database, operation and maintenance knowledge base, model library and real-time twin database to execute the efficient operation of the cross-domain high-end equipment operation and maintenance digital twin modeling system driven by cloud-fog-edge collaboration.

[0067] The cloud-fog-edge collaborative driving equipment operation and maintenance digital twin model building device is used to execute the cloud-fog-edge collaborative driving equipment operation and maintenance digital twin model building method.

[0068] Those skilled in the art will understand that the contents disclosed herein can be varied and modified in many ways. For example, the various devices or components described above can be implemented in hardware, or in software, firmware, or a combination of some or all of the three.

[0069] This disclosure uses flowcharts to illustrate the steps of a method according to embodiments of this disclosure. It should be understood that the preceding or following steps are not necessarily performed in exact order. Instead, the steps can be processed in reverse order or simultaneously. Furthermore, other operations can be added to these processes.

[0070] Those skilled in the art will understand that all or part of the steps in the above methods can be implemented by a computer program instructing related hardware, and the program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk. Optionally, all or part of the steps in the above embodiments can also be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiments can be implemented in hardware or as a software functional module. This disclosure is not limited to any particular combination of hardware and software.

[0071] Unless otherwise defined, all terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains. It should also be understood that terms such as those defined in a common dictionary should be interpreted as having a meaning consistent with their meaning in the context of the relevant art, and not as having an idealized or highly formalized meaning, unless expressly defined herein.

[0072] The foregoing description is intended to illustrate the present disclosure and should not be construed as limiting it. While several exemplary embodiments of the present disclosure have been described, those skilled in the art will readily understand that many modifications may be made to the exemplary embodiments without departing from the novel teachings and advantages of the present disclosure. Therefore, all such modifications are intended to be included within the scope of the present disclosure as defined by the claims. It should be understood that the foregoing description is intended to illustrate the present disclosure and should not be construed as limiting it to the specific embodiments disclosed, and modifications to the disclosed embodiments and other embodiments are intended to be included within the scope of the appended claims. The present disclosure is defined by the claims and their equivalents.

[0073] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0074] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for building a digital twin model for equipment operation and maintenance driven by cloud-fog-edge collaboration, characterized in that, include: Establish a cloud-fog-edge collaborative data resource integration model, determine key indicators of the twin model, and complete the integration and knowledge sharing of operation and maintenance data; Establish a dynamic adaptive modeling mechanism for component twins and an adaptive construction mechanism for cross-domain equipment twins; establish consistency evaluation criteria between monitoring data and digital twin model data, specifically including: Establish a dynamic adaptive modeling mechanism for component twins, construct a target domain composed of real-time data and a sample space with historical data as the source domain, use adversarial learning to measure the inter-domain loss, and measure the data feature distribution of the source and target domains to achieve unsupervised learning tasks using real-time data, thereby improving the accuracy of data model prediction of high-end equipment evolution data. The model update strategy based on cloud-fog collaboration deploys training tasks in the cloud and utilizes the cloud's ample computing power to achieve rapid iterative updates of the model. Establish a consistency evaluation criterion between component models and equipment twins, study the uncertainty quantification method of component and system digital twin models under the combined effect of prior uncertainty and model uncertainty, and establish a consistency evaluation criterion between monitoring data and digital twin model data; Establish a common mechanism for constructing digital twin models of cross-domain high-end equipment, develop a method for constructing equipment twins based on adaptive weighting of multi-scale component model parameters and deep reinforcement learning for cross-domain equipment, establish an adaptive construction mechanism for cross-domain equipment twins, and improve the speed of equipment self-consistent twin modeling. Based on a consensus maximization model and a game-theoretic adversarial weight learning method, the self-consistency between the equipment twin model after scenario migration and the operation and maintenance scenario is achieved, specifically including: The operation and maintenance scenario self-awareness of high-end equipment digital twins is achieved through prior knowledge embedding training, enabling scenario awareness and interpretation during the twin modeling process; Based on a knowledge rule base, environmental indicator features are screened and classified. Through a knowledge intervention mechanism, the equipment digital twin achieves self-learning and self-perception of environmental variables. Real-time online sensing with the control system is achieved through online communication, enabling the equipment twin to learn and sense environmental variables on its own. Digital twin model iteration and simulation compensation to establish a hybrid field digital twin simulation platform; To meet different operational and maintenance needs, the data scale and model complexity of the fusion model are adaptively determined to obtain an efficient digital twin model iteration and a proposed compensation strategy. For equipment twin modeling mechanisms under the same scenario and different typical operation and maintenance scenarios, based on consensus maximization model and game adversarial weight learning method, the equipment twin model after scenario migration is made self-consistent with the operation and maintenance scenario. By using self-sensing environmental variable data, the transfer learning theory is used to realize the transfer of environmental variables under complex and changing operating conditions, and the scenario operation and maintenance decision-making algorithm is adapted to achieve self-learning decision-making. By integrating historical databases, operation and maintenance knowledge bases, model libraries, and real-time twin databases to create big data twins, a cross-domain high-end equipment operation and maintenance digital twin modeling system driven by cloud, fog, and edge collaboration is implemented to ensure efficient operation.

2. The method for building a digital twin model of equipment operation and maintenance driven by cloud-fog-edge collaboration as described in claim 1, characterized in that, The establishment of a cloud-fog-edge collaborative data resource integration model, determination of key indicators for the twin model, and completion of operation and maintenance data integration and knowledge sharing also include: The cloud refers to a computing platform with high computing power and large storage space; the edge refers to an edge platform deployed on-site with fast computing and real-time data response at the nanosecond level; and the fog refers to a computing platform between the cloud and the edge with medium computing power and medium storage space. Based on cloud-fog-edge collaboration, a unified communication architecture for a twin database of the entire equipment operation and maintenance process is established. A parallel computing framework based on cloud computing is built, distributed data storage technology is used to store information resources, high-performance distributed processing and analysis are performed, and a data resource integration model is established. Construct a cloud-fog-edge collaborative data information sharing architecture, and implement an edge node model training iteration mechanism based on federated learning to analyze the resource constraints, edge heterogeneity, and dynamic environment faced by edge node computing, and determine the key indicators of the twin model; Data knowledge enhancement methods utilize transfer learning to reduce the gap between different data distributions, enabling the network to adapt to changing operating conditions based on a good modeling and prediction of the temporal characteristics of sensor data. For massive equipment operation and maintenance databases and real-time data, knowledge extraction and mining technologies are used to extract entities, relationships and attributes from massive equipment operation and maintenance data. Knowledge fusion and processing are used to disambiguate and integrate the extracted entities and merge them with structured data. After quality assessment, reasoning verification and updating, a knowledge graph is formed to obtain the value contained in the data and complete the knowledge value-added and iteration of the data in the database. The minimum path of a knowledge graph is defined by building data entities, relationships, and attributes, and industry knowledge graphs are built by combining historical experience.

3. A device for building a digital twin model of equipment operation and maintenance driven by cloud-fog-edge collaboration, characterized in that, include: The data integration and knowledge sharing module establishes a cloud-fog-edge collaborative data resource integration model, determines key indicators of the twin model, and completes operation and maintenance data integration and knowledge sharing. The common mechanism module of the digital twin model establishes a dynamic adaptive modeling mechanism for component twins and an adaptive construction mechanism for cross-domain equipment twins, and establishes a consistency evaluation criterion between monitoring data and digital twin model data; The adaptive modeling module for operation and maintenance scenarios, based on the consensus maximization model and the game adversarial weight learning method, achieves self-consistency between the equipment twin model after scenario migration and the operation and maintenance scenario. The cloud-fog-edge collaborative driving module integrates twin big data through historical database, operation and maintenance knowledge base, model library and real-time twin database to execute the efficient operation of the cross-domain high-end equipment operation and maintenance digital twin modeling system driven by cloud-fog-edge collaboration. The cloud-fog-edge collaborative driving equipment operation and maintenance digital twin model building device is used to execute the cloud-fog-edge collaborative driving equipment operation and maintenance digital twin model building method as described in any one of claims 1 to 2.

Images