Digital twin data linkage system and method

Abstract

The invention discloses a digital twin data linkage system and method, and the system comprises a data collection module which is used for collecting entity model data corresponding to a constructed digital twin model in real time; the data processing module is used for classifying and preprocessing the entity model data collected in real time; the data linkage module is used for matching designed linkage strategies of different types of data according to the preprocessed data, and acquiring components, component combinations or scenes of the digital twin model correspondingly linked with the preprocessed data by utilizing the matched linkage strategies; associating the preprocessed data with components, component combinations or scenes of the correspondingly linked digital twin model; and the linkage display module is used for displaying the real-time state and the data linkage effect of the digital twin model in a visual interface. According to the invention, various types of data can be processed and linked more accurately, and the real-time state and the data linkage effect of the digital twin model are displayed.

Description

Technical Field

[0001] This application relates to the field of digital twin technology, and specifically relates to a digital twin data linkage system and method. Background Art

[0002] Digital twin technology has developed rapidly in recent years, especially showing great potential in the field of industrial manufacturing. This technology uses virtual models to simulate physical entities in the real world, enabling real-time monitoring and management of production lines, equipment, and process flows, and significantly improving production efficiency and decision-making accuracy. The core of digital twin lies in data collection, processing, analysis, and application display, and the effective coordination of these links is the key to ensuring the efficient operation of the digital twin system in industrial manufacturing.

[0003] In the existing technology, in order to achieve data linkage of the digital twin system, the following means are usually adopted: First, various data of production equipment are collected in real time through a sensor network, and the collected data is preliminarily processed using traditional data processing methods (such as data cleaning, normalization, etc.); Second, a predefined rule library or a simple machine learning algorithm is used to match the association relationship between the data and the digital twin model; Finally, the effect of data linkage is displayed through a visualization tool.

[0004] However, there are still many deficiencies in the existing technology in terms of data linkage. For example, the data classification is not fine enough, resulting in the inability to accurately identify the characteristics of different types of data; the linkage strategy is too simple to effectively handle complex industrial application scenarios; the data processing ability is limited, and it is difficult to achieve real-time and efficient linkage effects; the above problems limit the overall performance of the digital twin system in the field of industrial manufacturing, especially when facing large-scale and high-frequency data streams, the existing technology often seems powerless. Therefore, there is an urgent need for a new digital twin data linkage system that can process and link various types of data more precisely. Summary of the Invention

[0005] In order to process and link various types of data more precisely and display the real-time status and data linkage effect of the digital twin model, this application provides a digital twin data linkage system and method.

[0006] In a first aspect, this application provides a digital twin data linkage system, including: A data collection module for collecting in real time the entity model data corresponding to the constructed digital twin model; A data processing module for classifying and preprocessing the entity model data collected in real time; the classification of entity model data includes: event data, motion data, and status data; A data linkage module, which is used to match the linkage strategies of different types of designed data according to the preprocessed data, and use the matched linkage strategies to obtain the components, component combinations or scenarios of the digital twin model linked to the preprocessed data; associate the preprocessed data with the components, component combinations or scenarios of the corresponding linked digital twin model; the linkage strategies of different types of data include: the linkage strategy of event data includes: applying a first data linkage model, which is a model output by matching event processing rules using preset event processing rules and machine learning algorithms; the linkage strategy of motion data includes: applying a second data linkage model, which is a model output by obtaining the running trajectory and / or behavior prediction of the data representation object using machine learning algorithms; the linkage strategy of state data includes: applying a third data linkage model, which is a model output by obtaining the state change prediction of the data representation object using machine learning algorithms. A linkage display module, which is used to display the real-time state and data linkage effect of the digital twin model in the visualization interface.

[0007] By adopting the above solution, different types of entity model data are classified and preprocessed, and the corresponding linkage models constructed based on machine learning algorithms are matched and applied as linkage strategies, realizing accurate data matching and dynamic association, enabling the digital twin model to dynamically reflect the real state and behavior of the entity model, and intuitively displaying the real-time state and data linkage effect of the digital twin model on the visualization interface, enhancing the user experience.

[0008] Preferably, the data acquisition module is further used to pre-determine the generation rate of the entity model data corresponding to the constructed digital twin model, compare the entity model data generation rate with the preset rate, and when the entity model data generation rate is greater than the preset rate, integrate the designed multi-protocol data acquisition accelerator hardware to assist in supporting the data acquisition of multiple data formats and communication protocols.

[0009] By adopting the above solution, when a large amount of entity model data is generated quickly, the multi-protocol data acquisition accelerator hardware is automatically started to improve the data acquisition speed, ensuring that the data generated at high frequency can be captured and processed in a timely and accurate manner.

[0010] Preferably, the data processing module is further configured to establish communication connections with multiple types of edge computing centers, and use a type of edge computing center to perform corresponding preprocessing on the matched single-type data in a preset manner; the preset manner for a type of edge computing center to set for the matched event-type data includes: data filtering and priority sorting; the preset manner for a type of edge computing center to set for the matched motion-type data includes: data interpolation compensation and Kalman filter noise reduction; the preset manner for a type of edge computing center to set for the matched status-type data includes: sliding window statistical calculation.

[0011] By adopting the above scheme, combined with edge computing, for event-type data that focuses on the accuracy and real-time performance of responses, data filtering and priority sorting are selected to reduce the interference of invalid data and timely process high-priority events; for motion-type data that focuses on the trajectory and behavior prediction of moving objects, data compensation and filtering are selected to improve the accuracy and stability of motion trajectory and behavior prediction; for status-type data that focuses on equipment life and status monitoring, sliding window statistical calculation is performed to monitor the status change trend of the monitored object in real time and enhance the timeliness and reliability of status monitoring.

[0012] Preferably, the data linkage module is further configured to pre-classify the preprocessed data by using a scene classification model pre-constructed based on a deep learning algorithm, and cluster to obtain data belonging to different application scenarios; design different types of data linkage strategies for different application scenario data to replace the different types of data linkage strategies designed according to the preprocessed data; The data linkage strategy for event-type data designed for different application scenario data includes: applying a first data linkage sub-model, the first data linkage model includes several first data linkage sub-models, and each first data linkage sub-model is a model constructed by using a preset event processing rule and a machine learning algorithm that match a single application scenario to output a model that matches the event processing rule; the data linkage strategy for motion-type data designed for different application scenario data includes: applying a second data linkage sub-model, the second data linkage model includes several second data linkage sub-models, and each second data linkage sub-model is a model that predicts and outputs the running trajectory and / or behavior of a data representation object that matches a single application scenario by using a machine learning algorithm; the data linkage strategy for status-type data designed for different application scenario data includes: applying a third data linkage sub-model, the third data linkage model includes several third data linkage sub-models, and each third data linkage sub-model is a model that predicts and outputs the status change of a data representation object that matches a single application scenario by using a machine learning algorithm.

[0013] By adopting the above solution, considering the differences in event handling, operation prediction, and status monitoring in different application scenarios, the preprocessed data is selected for classification to effectively identify the data in different application scenarios, and different linkage strategies are respectively matched, so that each type of data can obtain the most suitable processing method, improving the processing accuracy and response speed of each type of data in its respective application scenario.

[0014] Preferably, the data linkage module is further configured to perform scenario classification on the preprocessed data according to the preprocessed data. After clustering to obtain data belonging to different application scenarios, a scenario stage classification model pre-constructed based on a deep learning algorithm is used to perform scenario stage classification on the preprocessed data, and clustering to obtain data belonging to different application scenario stages; different types of data linkage strategies designed for different application scenario stage data are matched to replace different types of data linkage strategies designed for different application scenario data; The linkage strategies for event data designed for different application scenario stage data include: applying a first data linkage unit model, the first data linkage sub-model includes a number of first data linkage unit models, and each first data linkage unit model is a model constructed by using a preset event processing rule and a machine learning algorithm that match a single application scenario stage to output a model that matches the event processing rule; the linkage strategies for motion data designed for different application scenario stage data include: applying a second data linkage unit model, the second data linkage sub-model includes a number of second data linkage unit models, and each second data linkage unit model is a model that predicts and outputs the running trajectory and / or behavior of a data representation object that matches a single application scenario stage by using a machine learning algorithm; the linkage strategies for status data designed for different application scenario stage data include: applying a third data linkage unit model, the third data linkage sub-model includes a number of third data linkage unit models, and each third data linkage unit model is a model that predicts and outputs the status change of a data representation object that matches a single application scenario stage by using a machine learning algorithm.

[0015] By adopting the above solution, in order to perform data linkage in a more refined manner, further considering multiple operation stages corresponding to different application scenarios, the data in different application scenarios is classified by stage, and appropriate linkage strategies are matched to enhance the real-time performance, accuracy, and stability of data linkage.

[0016] Preferably, the data linkage module is further configured to check for running conflicts in the same component, component combination, or scenario of the digital twin model during the process of associating the preprocessed data with the components, component combinations, or scenarios of the correspondingly linked digital twin model in chronological order of association; if there is a running conflict, optimize the matching linkage strategy and use the optimized linkage strategy to re-obtain the components, component combinations, or scenarios of the digital twin model corresponding to the preprocessed data, and continue to complete the association of the preprocessed data with the components, component combinations, or scenarios of the correspondingly linked digital twin model.

[0017] By adopting the above solution, the digital twin data linkage process is effectively detected, and when a running conflict is detected, the matching linkage strategy is automatically optimized to re-determine the data linkage object, ensuring the coordination and consistency of all parts of the digital twin model during the dynamic association process.

[0018] Preferably, it further includes: a data linkage verification module, configured to obtain the real-time data of the actual entity model and compare it with the recorded association results of the preprocessed data with the components, component combinations, or scenarios of the correspondingly linked digital twin model, and judge the response rate and accuracy rate of the completed associated components, component combinations, or scenarios. If the accuracy rate is lower than the preset accuracy rate or the corresponding rate is lower than the preset response rate for a continuous preset period, an incremental learning method is adopted to optimize the matching linkage strategy.

[0019] By adopting the above solution, the data linkage effect between the digital twin model and the actual entity model is monitored in real time, and it is judged whether the response speed and correctness of the digital twin model after data linkage meet the preset requirements, and the linkage strategy is optimized when the preset requirements are not met.

[0020] In a second aspect, the present application provides a digital twin data linkage method, including: Real-time collecting the entity model data corresponding to the constructed digital twin model; Classifying and preprocessing the real-time collected entity model data; the classification of entity model data includes: event data, motion data, and state data; According to the preprocessed data, matching the designed linkage strategies for different types of data, and using the matching linkage strategies to obtain the components, component combinations, or scenarios of the digital twin model corresponding to the preprocessed data; Associating the preprocessed data with the components, component combinations, or scenarios of the correspondingly linked digital twin model; Displaying the real-time state and data linkage effect of the digital twin model in the visualization interface.

[0021] By adopting the above - mentioned solution, it is possible to realize the real - time collection and processing of the entity model data corresponding to the digital twin model, classify and pre - process the data according to different types, and use different linkage strategies so that the pre - processed data can be accurately and dynamically associated with the components, component combinations or scenarios of the digital twin model.

[0022] In a third aspect, the present application provides a computer - readable storage medium, which includes a stored computer program. When the computer program runs, it controls the device where the computer - readable storage medium is located to execute the method as described above.

[0023] In a fourth aspect, the present application provides a computer device, which includes a memory, a processor, and a program stored on the memory and executable. When the program is executed by the processor, it realizes the steps of the method as described above.

[0024] In summary, the present application has the following beneficial effects: 1. Real - time collection and classification pre - processing of entity model data ensure the accuracy and timeliness of the data, improve the accuracy and reliability of data linkage; use machine learning algorithms to construct different types of advanced data linkage models to achieve precise matching and dynamic association of event - type, motion - type, and status - type data with the components or scenarios of the digital twin system, effectively cope with complex industrial application scenarios, and enhance the response speed and adaptability of the digital twin system; display the real - time status and data linkage effect of the digital twin model in the visualization interface, enhancing the user's intuitive perception of the system operation and operation convenience; 2. From the data collection stage to the entire process of data linkage, adaptively perform hardware integration to accelerate data collection, adaptively pre - process different types of data in combination with the edge computing center, and construct linkage strategy acquisition models for different application scenarios and different scenario stages to achieve more accurate linkage; 3. Check for running conflicts in the same component, component combination or scenario during the dynamic association process, and optimize the linkage strategy to solve the conflict problem; and verify the effectiveness of the dynamic association process according to the real - time data of the actual entity model, and optimize the linkage strategy in an incremental learning manner when necessary to achieve more effective data linkage. Description of the Drawings

[0025] Figure 1 It is a schematic structural diagram of the digital twin data linkage system in a specific embodiment; Figure 2 It is a flowchart of the digital twin data linkage method in a specific embodiment. Detailed Embodiments

[0026] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0027] As Figure 1 shown, an embodiment of the present application discloses a digital twin data linkage system, which specifically includes: a data acquisition module 1, a data processing module 2, a data linkage module 3, and a linkage display module 4.

[0028] Specifically, the data acquisition module 1 includes various types of sensors and data interfaces, and is used to collect in real time the entity model data corresponding to the constructed digital twin model, such as: sensor, PLC, ERP / MES system data.

[0029] In addition, considering that the large amount of multi-source heterogeneous data generated in real time by the industrial production entity models corresponding to some manufacturing objects is large and complex, in order to collect various entity model data more timely; the data acquisition module 1 is also used to pre-determine the entity model data generation rate corresponding to the constructed digital twin model, and compare the entity model data generation rate with a preset rate, and the preset rate can be set manually; when the entity model data generation rate is greater than the preset rate, integrate the pre-designed multi-protocol (such as: MQTT / HTTP / OPC UA, etc.) data acquisition accelerator hardware to assist in supporting the data acquisition of various data formats and communication protocols.

[0030] Specifically, the data processing module 2 is used to classify and preprocess the entity model data collected in real time; the classification of the entity model data includes: event data, motion data, status data, etc. Among them, the preprocessing includes: data format conversion, converting data in different formats into a standard format; data cleaning, data compensation and integration, etc.

[0031] In addition, in order to retain as much beneficial information as possible in the collected data information, process each type of data timely and adaptively in a suitable manner to better drive the digital twin model. The data processing module 2 is also used to establish a communication connection with multiple types of edge computing centers, and each type of edge computing center corresponds to a class of data, and uses the edge computing center belonging to a class to preprocess the matching single type of data in a corresponding preset manner; Specifically, it includes: For a type of edge computing center, the preset methods corresponding to the matched event - type data include: In addition to processing such as data synchronization and data format conversion, considering that the event - type data linkage in the digital twin system mainly focuses on the timely response and accurate processing of event - type data, data filtering (such as millimeter - level time - window filtering, Bloom Filter deduplication, etc.) and priority sorting and other processing methods are also adopted for event - type data. For a type of edge computing center, the preset methods corresponding to the matched motion - type data include: In addition to processing such as data synchronization and data format conversion, considering that the motion - type data linkage in the digital twin system mainly focuses on the timely prediction of the running trajectories or behaviors of the objects represented by motion - type data (such as personnel, equipment, materials, etc.), data interpolation compensation (such as IMU data interpolation compensation) and Kalman filter noise reduction are performed on the state - type data. For a type of edge computing center, the preset methods corresponding to the matched state - type data include: In addition to processing such as data synchronization and data format conversion, considering that the state - type data linkage in the digital twin system mainly focuses on the comprehensive monitoring and timely control of the objects represented by state - type data (such as equipment, processes, environment, etc.), sliding - window statistical calculations (such as mean / variance / covariance, etc.) are performed and compared with the preset abnormal baselines to initially obtain abnormal state data.

[0032] Specifically, the data linkage module 3 is used to match the designed linkage strategies for different types of data according to the pre - processed data, obtain the components, component combinations or scenarios of the digital twin model corresponding to the linkage of the pre - processed data by using the matched linkage strategies, and associate the pre - processed data with the components, component combinations or scenarios of the digital twin model corresponding to the linkage.

[0033] Among them, in order to drive the digital twin system, it is necessary to determine the components, component combinations or scenarios for data linkage. And to ensure the accuracy of data linkage, linkage strategies are adaptively designed for different types of data. Specifically, the linkage strategies for different types of data include: The linkage strategy for event data includes: applying the first data linkage model to match the event processing rules; the first data linkage model is a model output by matching event processing rules using preset event processing rules and machine learning algorithms; among them, the preset event processing is preset according to typical events and their processing methods in the entity model. The input of the first data linkage model is event data, and the output is the event processing rules matched by the event data; for example: sudden drop in welding torch voltage (event) - generate a warning for the welding torch device, trigger visual re-inspection of the camera device, and adjust the welding parameters of the welding device (event processing rules). Correspondingly, the linkage strategy for motion data includes: applying the second data linkage model to obtain the running trajectory and / or behavior of the data representation object; the second data linkage model is a model output by predicting the running trajectory and / or behavior of the data representation object using machine learning algorithms; the input of the second data linkage model is motion data, and the output is the predicted running trajectory and behavior of the motion data representation object. Correspondingly, the linkage strategy for status data includes: applying the third data linkage model to obtain the status change of the status data representation object; the third data linkage model is a model output by predicting the status change of the data representation object using machine learning algorithms, and the input of the third data linkage model is status data, and the output is the predicted status change of the status data representation object.

[0034] Obtaining the components, component combinations or scenarios of the digital twin model corresponding to the preprocessed data by using the matched linkage strategy includes: after processing the event data by using the event processing rules output by the first data model, identifying the components, component combinations or scenarios of the digital twin model involved in the event processing rules based on semantic parsing, such as: welding torch device components, camera device components, etc., to determine the components, component combinations or scenarios of the digital twin model corresponding to the linkage; using the predicted running trajectory and / or behavior of the motion data representation object output by the second data model, determining the components, component combinations or scenarios of the digital twin model corresponding to the linkage based on the components, component combinations or scenarios of the digital twin model involved in the same running trajectory and / or behavior of the motion data representation object in the pre-acquired historical digital twin model; correspondingly, using the predicted status change of the status data representation object output by the third data model, determining the components, component combinations or scenarios of the digital twin model corresponding to the linkage based on the components, component combinations or scenarios of the digital twin model involved in the same change of the status data representation object in the pre-acquired historical digital twin model.

[0035] Associate the preprocessed data with the components, component combinations, or scenarios of the corresponding linked digital twin model; including: generating signaling based on direct data or indirectly predicted data, attaching the signaling to the components, component combinations, or scenarios in the linked digital twin model to complete the mapping association, and completing the dynamic association according to the time change.

[0036] In addition, considering the different data requirements of actual users for the entity model data, different types of data linkage strategies designed can be further matched according to the preprocessed data and the corresponding type of data requirements, and the components, component combinations, or scenarios of the digital twin model linked to the preprocessed data can be obtained by using the matched linkage strategies; among them, the corresponding type of data requirements includes: the event class data requirements include that the event response rate reaches a preset rate and the response accuracy reaches the first preset accuracy rate, the motion class data requirements include that the accuracy rate of the running trajectory and / or behavior of the prediction object reaches the second preset accuracy rate, and the state class data requirements include that the accuracy rate of the state change of the prediction object reaches the third preset accuracy rate; the corresponding linkage strategy for setting event class data includes: applying the first data linkage model set output by matching event processing rules with different matching rates constructed by using preset event processing rules and machine learning algorithms, the corresponding linkage strategy for setting motion class data includes: applying the second data linkage model set output by predicting the running trajectory and / or behavior of the motion data representation object with different accuracies constructed by using machine learning algorithms; the corresponding linkage strategy for setting state class data includes: applying the third data linkage model set output by predicting the state change of the state data representation object with different accuracies constructed by using machine learning algorithms.

[0037] The linkage display module 4 is used to display the real-time state and data linkage effect of the digital twin model in the visualization interface. Specifically, obtain and display the running state and data of the components, component combinations, or scenarios of the digital twin model after dynamic association. In addition, it can also receive the user's display requirements and only display the components, component combinations, or scenarios that are selected or defined by the user as key and have been processed by dynamic association.

[0038] By adopting the above system, it is possible to accurately process and complete the linkage of various entity model data and the digital twin model, and timely display the real-time state and data linkage effect of the digital twin model.

[0039] A specific embodiment, considering the differences in event processing, operation prediction, and status monitoring in different application scenarios, in order to improve the accuracy and efficiency of data linkage of different types of data, the data linkage module 3 in the system is further configured to pre-classify the preprocessed data by using a scene classification model pre-constructed based on a deep learning algorithm, and cluster to obtain data belonging to different application scenarios; match different types of data linkage strategies designed for different application scenario data to replace different types of data linkage strategies designed according to the preprocessed data; taking industrial manufacturing objects as an example, the main application scenarios include: production line scenarios, production quality inspection scenarios, and warehousing scenarios.

[0040] The data linkage strategy for event data designed to match different application scenario data includes: applying a first data linkage sub-model to obtain a matching event processing rule. The first data linkage model includes a number of first data linkage sub-models, and each first data linkage sub-model is a model constructed by using a preset event processing rule and a machine learning algorithm that match a single application scenario and outputs a model that matches the event processing rule; the data linkage strategy for motion data designed to match different application scenario data includes: applying a second data linkage sub-model to obtain the predicted running trajectory and / or behavior of the data representation object that matches a single application scenario; the second data linkage model includes a number of second data linkage sub-models, and each second data linkage sub-model is a model that predicts and outputs the running trajectory and / or behavior of the data representation object that matches a single application scenario by using a machine learning algorithm; the data linkage strategy for status data designed to match different application scenario data includes: applying a third data linkage sub-model to obtain the predicted state change output of the data representation object that matches a single application scenario; the third data linkage model includes a number of third data linkage sub-models, and each third data linkage sub-model is a model that predicts and outputs the state change of the data representation object that matches a single application scenario by using a machine learning algorithm.

[0041] In addition, considering that different application scenarios handle the same event differently at different operation stages, the trajectory changes differently under the same motion data, and the monitoring intensity is different under the same state, in order to further improve the accuracy and efficiency of the linkage of different types of data, the data linkage module 3 is further configured to classify the preprocessed data according to the preprocessed data after clustering to obtain data belonging to different application scenarios, and then use the scenario stage classification model pre-constructed based on the deep learning algorithm to classify the preprocessed data into scenario stages, and cluster to obtain data belonging to different application scenario stages, and match the linkage strategies of different types of data designed for different application scenario stage data to replace the linkage strategies of different types of data designed for different application scenario data; wherein, the production line scenario stage includes: the early stage, the middle stage, and the later stage of production; the production quality inspection scenario stage includes: the initial production quality inspection stage and the re-inspection stage of production quality; the warehousing scenario stage includes: the storage stage, the transfer storage stage.

[0042] The linkage strategy of event class data designed for different application scenario stage data includes: applying the first data linkage unit model to obtain the matching event processing rule, and the first data linkage sub-model includes a number of first data linkage unit models, and each first data linkage unit model is a model constructed by using the preset event processing rule and machine learning algorithm that match a single application scenario stage to output the matching event processing rule; the linkage strategy of motion class data designed for different application scenario stage data includes: applying the second data linkage unit model to obtain the predicted running trajectory and / or behavior of the data representation object that matches a single application scenario stage; the second data linkage sub-model includes a number of second data linkage unit models, and each second data linkage unit model is a model that predicts and outputs the running trajectory and / or behavior of the data representation object that matches a single application scenario stage by using the machine learning algorithm; the linkage strategy of state class data designed for different application scenario stage data includes: applying the third data linkage unit model to obtain the predicted state change of the data representation object that matches a single application scenario stage; the third data linkage sub-model includes a number of third data linkage unit models, and each third data linkage unit model is a model that predicts and outputs the state change of the data representation object that matches a single application scenario stage by using the machine learning algorithm.

[0043] A specific embodiment comprehensively considers the mutual influence among various components and devices during the data linkage process, avoids resource conflicts and equipment damage, introduces conflict detection and optimization functions in the dynamic association process, and ensures that the digital twin data linkage system performs more robustly and efficiently in a complex industrial environment. In the system, the data linkage module 3 is further configured to, during the process of associating the preprocessed data with the components, component combinations or scenarios of the corresponding linked digital twin model, check whether there are running conflicts in the same component, the same component combination or the same scenario according to the chronological order of association; if there are running conflicts, for example, at the same time, through the dynamic association of event A data, component B in the digital twin system is turned on for a period of time, and through the dynamic association of event C data, component B in the digital twin system continues to be turned on, and there is a conflict between the two; then optimize the matching linkage strategy and use the optimized linkage strategy to re-obtain the components, component combinations or scenarios of the digital twin model corresponding to the preprocessed data, and continue to complete the association of the preprocessed data with the components, component combinations or scenarios of the corresponding linked digital twin model.

[0044] In addition, if there are still conflicts in continuing to complete the association of the preprocessed data with the components, component combinations or scenarios of the corresponding linked digital twin model, then according to the priority order of the preset data of the entity model, when performing linkage on the conflicting data, the dynamic linkage operation based on the priority data shall be the main operation.

[0045] A specific embodiment, in order to further improve the accuracy of data linkage, can define the self-evaluation and optimization of data linkage. The system further includes: The data linkage verification module 5 is configured to obtain the real-time data of the actual entity model and compare it with the recorded association results of the preprocessed data with the components, component combinations or scenarios of the corresponding linked digital twin model, and judge the response rate (event response rate) and the correct rate (the correct rate of predicting the motion trajectory and behavior and the correct rate of state change) of the completed associated components, component combinations or scenarios. If the correct rate is lower than the preset correct rate or the corresponding rate is lower than the preset response rate for a continuous preset period, the incremental learning method is adopted to optimize the matching linkage strategy.

[0046] As Figure 2 shown, the embodiment of the present application provides a digital twin data linkage method, including: S1. Real-time collect the entity model data corresponding to the constructed digital twin model.

[0047] S2. Classify and preprocess the real-time collected entity model data.

[0048] Among them, the classification of entity model data includes: event data, motion data and state data; S3. According to the preprocessed data, match the designed linkage strategies for different types of data, and use the matched linkage strategies to obtain the components, component combinations or scenarios of the digital twin model that correspond to the preprocessed data.

[0049] S4. Associate the preprocessed data with the components, component combinations or scenarios of the corresponding linked digital twin models.

[0050] S5. Display the real-time status and data linkage effect of the digital twin model in the visual interface.

[0051] In a specific embodiment, S1 in the method also includes: pre-determining the physical model data generation rate corresponding to the constructed digital twin model, and comparing the physical model data generation rate with a preset rate. When the physical model data generation rate is greater than the preset rate, an integrated multi-protocol data acquisition accelerator hardware is used to assist in supporting data acquisition in multiple data formats and communication protocols.

[0052] In a specific embodiment, S2 in the method also includes: establishing communication connections with multiple types of edge computing centers, using a type of edge computing center to pre-process the matched single type of data in a corresponding preset manner; a type of edge computing center sets a preset method for matching event type data, including data filtering and priority sorting; a type of edge computing center sets a preset method for matching motion type data, including data interpolation compensation and Kalman filtering noise reduction; a type of edge computing center sets a preset method for matching state type data, including sliding window statistical calculation.

[0053] In a specific embodiment, S3 in the method also includes: before the linkage strategy of different types of data designed according to the preprocessed data matching, pre-classify the preprocessed data using a scene classification model pre-built based on a deep learning algorithm, and cluster to obtain data belonging to different application scenarios; the linkage strategy of different types of data designed according to the data matching of different application scenarios is used to replace the linkage strategy of different types of data designed according to the preprocessed data matching.

[0054] In a specific embodiment, S3 in the method also includes: after performing scene classification on the preprocessed data according to the preprocessed data and clustering to obtain data belonging to different application scenarios, using a scene stage classification model pre-constructed based on a deep learning algorithm to perform scene stage classification on the preprocessed data and clustering to obtain data belonging to different application scenario stages; a linkage strategy for different types of data designed for data matching at different application scenario stages is used to replace the linkage strategy for different types of data designed for data matching at different application scenario stages.

[0055] A specific embodiment, in the method, S3 further includes: during the process of associating the preprocessed data with the components, component combinations or scenarios of the corresponding linked digital twin model, checking whether there are running conflicts in the same component, the same component combination or the same scenario according to the chronological order of association; if there are running conflicts, optimizing the matching linkage strategy and using the optimized linkage strategy to re-obtain the components, component combinations or scenarios of the digital twin model corresponding to the preprocessed data, and continuing to complete the association of the preprocessed data with the components, component combinations or scenarios of the corresponding linked digital twin model.

[0056] A specific embodiment, the method further includes: obtaining the real-time data of the actual entity model, and comparing it with the recorded association results of associating the preprocessed data with the components, component combinations or scenarios of the corresponding linked digital twin model, judging the response rate and accuracy rate of the components, component combinations or scenarios for which the association is completed, if there is a continuous preset period during which the accuracy rate is lower than the preset accuracy rate or the corresponding rate is lower than the preset response rate, then adopting an incremental learning method to optimize the matching linkage strategy.

[0057] The embodiment of the present application also discloses a computer-readable storage medium.

[0058] Specifically, this computer-readable storage medium stores a computer program that can be loaded and executed by a processor, such as the digital twin data linkage method described above. This computer-readable storage medium includes, for example: various media that can store program codes such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks or optical discs.

[0059] The embodiment of the present application also discloses a computer device.

[0060] Specifically, the computer device includes a memory and a processor, and the memory stores a computer program that can be loaded and executed by the processor, such as the digital twin data linkage method described above.

[0061] The above are all the preferred embodiments of the present application. The protection scope of the present application is not limited by this. Any feature disclosed in this specification (including the abstract and drawings), unless specifically described, can be replaced by other equivalent or similar-purpose alternative features. That is, unless specifically described, each feature is only an example of a series of equivalent or similar features.

Claims

1. A digital twin data linkage system, characterized in that, Including: A data acquisition module for real-time acquisition of entity model data corresponding to the constructed digital twin model; A data processing module for classifying and preprocessing the real-time acquired entity model data; The classification of entity model data includes: event data, motion data, and status data; A data linkage module for matching the linkage strategies of different types of designed data according to the preprocessed data, and using the matched linkage strategies to obtain the components, component combinations, or scenarios of the digital twin model linked to the preprocessed data; associating the preprocessed data with the components, component combinations, or scenarios of the corresponding linked digital twin model; the linkage strategies of different types of data include: the linkage strategy of event data includes: applying a first data linkage model, which is a model output by matching event processing rules using preset event processing rules and machine learning algorithms; the linkage strategy of motion data includes: applying a second data linkage model, which is a model output by obtaining the running trajectory and / or behavior prediction of the data representation object using machine learning algorithms; the linkage strategy of status data includes: applying a third data linkage model, which is a model output by obtaining the status change prediction of the data representation object using machine learning algorithms; A linkage display module for displaying the real-time status and data linkage effect of the digital twin model in a visualization interface.

2. The digital twin data linkage system according to claim 1, characterized in that The data acquisition module is further configured to pre-determine the generation rate of entity model data corresponding to the constructed digital twin model, compare the entity model data generation rate with a preset rate, and when the entity model data generation rate is greater than the preset rate, integrate the designed multi-protocol data acquisition accelerator hardware to assist in supporting the data acquisition of multiple data formats and communication protocols.

3. The digital twin data linkage system according to claim 1, characterized in that, The data processing module is further configured to establish a communication connection with multiple types of edge computing centers, and use a type of edge computing center to perform corresponding preset preprocessing on the matched single type of data; The preset methods corresponding to the event data matched by a type of edge computing center include: data filtering and priority sorting; the preset methods corresponding to the motion data matched by a type of edge computing center include: data interpolation compensation and Kalman filter noise reduction; the preset methods corresponding to the status data matched by a type of edge computing center include: sliding window statistical calculation.

4. The digital twin data linkage system according to claim 1, wherein The data linkage module is further configured to pre-classify the preprocessed data using a scene classification model pre-constructed based on a deep learning algorithm, and cluster to obtain data belonging to different application scenarios; match the linkage strategies of different types of designed data for different application scenario data to replace the linkage strategies of different types of designed data according to the preprocessed data; The linkage strategies for event - type data designed for data matching in different application scenarios include: applying the first data linkage sub - model. The first data linkage model includes several first data linkage sub - models. Each first data linkage sub - model is a model constructed by using the preset event - handling rules and machine - learning algorithms that match a single application scenario to output a model that matches the event - handling rules; The linkage strategies for motion - type data designed for data matching in different application scenarios include: applying the second data linkage sub - model. The second data linkage model includes several second data linkage sub - models. Each second data linkage sub - model is a model that predicts and outputs the running trajectory and / or behavior of the data representation object that matches a single application scenario by using machine - learning algorithms; The linkage strategies for state - type data designed for data matching in different application scenarios include: applying the third data linkage sub - model. The third data linkage model includes several third data linkage sub - models. Each third data linkage sub - model is a model that predicts and outputs the state change of the data representation object that matches a single application scenario by using machine - learning algorithms.

5. The digital twin data linkage system according to claim 4, wherein The data linkage module is further configured to, after classifying the pre - processed data by scenario and clustering to obtain data belonging to different application scenarios, use the scenario - stage classification model pre - constructed based on the deep - learning algorithm to perform scenario - stage classification on the pre - processed data and cluster to obtain data belonging to different application - scenario stages; Linkage strategies for different types of data designed for data matching in different application - scenario stages are used to replace the linkage strategies for different types of data designed for data matching in different application scenarios; The linkage strategies for event - type data designed for data matching in different application - scenario stages include: applying the first data linkage unit model. The first data linkage sub - model includes several first data linkage unit models. Each first data linkage unit model is a model constructed by using the preset event - handling rules and machine - learning algorithms that match a single application - scenario stage to output a model that matches the event - handling rules; The linkage strategies for motion - type data designed for data matching in different application - scenario stages include: applying the second data linkage unit model. The second data linkage sub - model includes several second data linkage unit models. Each second data linkage unit model is a model that predicts and outputs the running trajectory and / or behavior of the data representation object that matches a single application - scenario stage by using machine - learning algorithms; The linkage strategies for state - type data designed for data matching in different application - scenario stages include: applying the third data linkage unit model. The third data linkage sub - model includes several third data linkage unit models. Each third data linkage unit model is a model that predicts and outputs the state change of the data representation object that matches a single application - scenario stage by using machine - learning algorithms.

6. The digital twin data linkage system according to claim 1, characterized in that The data linkage module is further configured to check for running conflicts in the same component, component combination, or scene of the digital twin model associated with the pre-processed data in the chronological order of association during the process of associating the pre-processed data with the components, component combinations, or scenes of the corresponding associated digital twin model. If there are running conflicts, the matching linkage strategy is optimized and the components, component combinations, or scenes of the digital twin model associated with the pre-processed data are re-obtained using the optimized linkage strategy, and the association between the pre-processed data and the components, component combinations, or scenes of the corresponding associated digital twin model is continued to be completed.

7. The digital twin data linkage system according to claim 1, wherein It further includes: A data linkage verification module, configured to obtain the real-time data of the actual entity model and compare it with the recorded association results of the pre-processed data with the components, component combinations, or scenes of the corresponding associated digital twin model, and judge the response rate and accuracy rate of the components, component combinations, or scenes for which the association is completed. If the accuracy rate is lower than the preset accuracy rate or the corresponding rate is lower than the preset response rate for a continuous preset period, an incremental learning method is adopted to optimize the matching linkage strategy.

8. A digital twin data linkage method applying the system according to any one of claims 1 to 7, characterized in that, It includes: Real-time acquisition of the entity model data corresponding to the constructed digital twin model; Classifying and pre-processing the real-time acquired entity model data; The classification of the entity model data includes: event data, motion data, and status data; According to the pre-processed data, match the designed linkage strategies for different types of data, and use the matching linkage strategies to obtain the components, component combinations, or scenes of the digital twin model associated with the pre-processed data; Associate the pre-processed data with the components, component combinations, or scenes of the corresponding associated digital twin model; Display the real-time status and data linkage effect of the digital twin model in the visualization interface.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein when the computer program runs, it controls the device where the computer-readable storage medium is located to execute the method according to any one of claims 1 to 7.

10. A computer device, characterized in that, The computer device includes a memory, a processor, and a program stored and executable on the memory. When the program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 7.

Images