Method for constructing dynamic simulation model verification system of digital-real combination test

Abstract

The invention discloses a method for constructing a verification system of a dynamic simulation model of a digital-real combination test. The method comprises the following steps of: acquiring operation data of a simulation target and constructing a digital twin virtual scene based on a preset simulation model; building a real test scene and acquiring test data in the real test scene; a data interaction module is established to construct a test data interaction relationship between the two scenes, and real-time data of a simulation target in any scene is utilized to guide a simulation target in the other scene to make the same action feedback; and constructing a verification module to synchronously obtain feedback parameters fed back by the two scenes, and finally performing consistency verification on the simulation model based on the feedback parameters. According to the invention, through data interaction and guidance between the digital twin virtual scene and the real test scene, a closed-loop digital-real combined simulation verification system is formed. The system can be continuously self-optimized, so that the practicability and the reliability of a simulation model are improved, the dependence on a physical test is reduced, and the cost and the time consumption are reduced.

Description

Technical Field

[0001] The invention belongs to the technical field of model verification, and in particular relates to a method for constructing a dynamic simulation model verification system for a digital-realistic combined experiment. Background Art

[0002] In modern engineering and scientific research, with the significant increase in computing power and the advancement of simulation technology, simulation models have become key tools for designing, testing and analyzing complex systems. These models can simulate a wide range of phenomena from mechanical structures to electronic circuits to biological systems, providing engineers and scientists with the ability to deeply understand system behavior. They not only accelerate the innovation process, but also reduce R&D costs and risks.

[0003] The core value of simulation models lies in their ability to replicate real-world conditions and events with high accuracy, thereby providing decision makers with reliable predictions and analysis. However, the realization of this value is highly dependent on the accuracy and reliability of the model. Existing simulation model verification methods, although they have made progress in some areas, still face a series of challenges: 1. They often rely on static data and cannot fully reflect the dynamic characteristics of the system. 2. Insufficient real-time update capabilities make it difficult to adapt to rapidly changing environments and conditions. 3. Under changing operating conditions, existing methods find it difficult to maintain consistency between the model and the actual system behavior. These limitations not only affect the credibility of the simulation results, but also restrict the application potential of simulation technology in a wider range of scenarios.

[0004] In order to overcome these challenges and improve the verification efficiency and accuracy of the simulation model, this paper proposes a method for constructing a dynamic simulation model verification system for digital-realistic combined experiments. This method achieves synchronization and consistency in the closed-loop interaction of relevant data between real and virtual scenes through real-time data interaction and a two-way guidance mechanism, significantly improving the accuracy and practicality of model prediction. Summary of the invention

[0005] In view of this, the purpose of the present invention is to propose a method for constructing a digital-real combination test dynamic simulation model verification system, which can form a closed-loop digital-real combination simulation verification system through data interaction and guidance between the digital twin virtual scene and the real test scene. This system can continuously optimize itself, improve the practicality and reliability of the simulation model, reduce dependence on physical experiments, and reduce costs and time consumption.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention provides a method for constructing a dynamic simulation model verification system for a digital-realistic combined experiment, comprising:

[0008] Collect the operation data of the simulation target and build a digital twin virtual scene based on the preset simulation model;

[0009] Build a realistic test scenario for the simulation target and obtain test data from the realistic test scenario;

[0010] Establish a data interaction module to construct the test data interaction relationship between the real test scene and the virtual scene, and use the real-time data of the simulated target in the virtual scene to guide the simulated target in the real test scene to make the same action feedback or use the real-time data of the simulated target in the real test scene to guide the simulated target in the virtual scene to make the same action feedback;

[0011] A verification module is constructed to synchronously obtain feedback parameters from real test scenarios and virtual scenarios, and the consistency of dynamic data of the simulation model is verified based on the feedback parameters.

[0012] Preferably, collecting the operation data of the simulation target and building a digital twin virtual scene based on a preset simulation model includes:

[0013] Collect various operation data generated by the simulation target in actual operation through deployed sensors and data acquisition equipment;

[0014] Preprocess various operation data separately and store them in the database; the preprocessing includes classification, cleaning, denoising, standardization and normalization;

[0015] Based on the simulation model to be verified, the basic digital twin virtual scene is constructed using the pre-processed operation data;

[0016] Use operational data to calibrate the simulation model to ensure that the digital twin virtual scene accurately reflects the behavior of the simulation target.

[0017] Preferably, building a realistic test scenario of the simulation target and obtaining test data in the realistic test scenario includes:

[0018] Build a realistic test environment according to the operation requirements, operating environment and test needs of the simulation target;

[0019] Install test equipment based on the test environment, including all devices that can universally represent the simulation target subject and various types of sensors, data acquisition systems and actuators;

[0020] Determine the test parameters that need to be controlled, monitored and collected during the test, and determine the sampling rules, path channels, classification processing and storage rules for data collection based on the specific conditions of the test environment;

[0021] After the realistic test scenario is built, the test is performed under controlled conditions according to the test requirements, and the test data of the simulation target in actual operation is collected.

[0022] Preferably, the data interaction module performs the following steps:

[0023] Establish the test data interaction relationship between the real test scene and the virtual scene according to the preset test data interaction port, and set the data interaction architecture, interactive communication protocol, interaction frequency, interaction trigger instruction, transmission frequency and synchronization mechanism;

[0024] During the test, when the interaction condition is triggered, the real-time data of the simulation target in the virtual scene is used to guide the simulation target in the actual test scene to make the same action feedback, or the real-time data of the simulation target in the actual test scene is used to guide the simulation target in the virtual scene to make the same action feedback, wherein the trigger condition includes an interaction trigger instruction actively sent by the user, an interaction trigger instruction triggered with a preset time frequency, and an interaction trigger instruction triggered with a preset trigger condition;

[0025] When the data interaction action starts, the data collected from the real test scene and the virtual scene are annotated to generate feedback parameters, where the annotation content includes timestamp annotation, annotation of the data belonging to the initiator / guided end, and data classification at the initiator / guided end.

[0026] Preferably, the verification module performs the following steps:

[0027] Synchronously obtain feedback parameters from the real test scene and the virtual scene, analyze the feedback parameters, determine the timestamp of the feedback parameters, the label of the data belonging to the initiator / guided end, and the data classification of the data at the initiator / guided end;

[0028] The two sets of feedback parameters are aligned based on the timestamp annotation of the feedback parameters, and the starting and ending positions of the data interaction action on the feedback parameters sorted by the time series are determined based on the annotation of the data belonging to the initiator / guided end, and the two sets of feedback parameters are sliced ​​according to the starting and ending positions;

[0029] Based on the slicing results and the annotation of the data belonging to the initiator / guided end, determine that the feedback parameters corresponding to the real test scene and the virtual scene in each slice belong to the initiator / guided end, and perform binary classification on the slicing results;

[0030] The consistency of the simulation model is verified using the feedback parameters corresponding to the binary classification results and the feedback parameters corresponding to when no data interaction is performed.

[0031] Preferably, performing consistency verification of dynamic data of the simulation model using feedback parameters corresponding to the binary classification results includes:

[0032] According to the binary classification result, when the simulation target in the virtual scene is used as the initiator and the simulation target in the real test scene is used as the guided end, the feedback parameters of the virtual scene in the corresponding slicing result and the feedback parameters corresponding to the real test scene are the first type of feedback parameter pair;

[0033] According to the binary classification result, when the simulation target in the real test scene is used as the initiator and the simulation target in the virtual scene is used as the guided end, the feedback parameters of the virtual scene in the corresponding slicing result and the feedback parameters corresponding to the real test scene are the second type of feedback parameter pair;

[0034] Determine the feedback parameters of the virtual scene when no data interaction action is performed and the feedback parameters corresponding to the real test scene as the third type of feedback parameter pair;

[0035] Based on the first type of feedback parameter pair, dynamic data consistency verification is performed on the simulation model in the virtual scene relative to the simulation target in the actual test scene to obtain a first verification result;

[0036] Based on the second type of feedback parameter pair, dynamic data consistency verification is performed on the simulation model in the virtual scene relative to the simulation target in the actual test scene to obtain a second verification result;

[0037] Based on the third type of feedback parameter pair, dynamic data consistency verification is performed on the simulation model in the virtual scene relative to the simulation target in the actual test scene to obtain a third verification result;

[0038] According to the first verification result and the third verification result, the error curve corresponding to the third verification result and the error curve in the first verification result are subjected to a parallel trend test using a multi-time point double difference method. If the test result shows that the two curves are in a parallel trend or an asymptotic trend, the accuracy verification of the simulation model is passed; if the test result shows that the two curves are in a dissimilar trend, the accuracy verification is not passed.

[0039] According to the second verification results and the third verification results, the error curve corresponding to the third verification result and the error curve in the second verification result are subjected to a parallel trend test using the multi-time-point double difference method. If the test result shows that the two curves are in a parallel trend or an asymptotic trend, the accuracy verification of the simulation model passes; if the test result shows that the two curves are in a dissimilar trend, the accuracy verification fails.

[0040] The present invention has achieved at least the following beneficial effects:

[0041] 1. By constructing digital twin virtual scenes and real test scenes, and realizing data interaction and guidance between the two, a closed-loop digital-real combination simulation verification system is formed. This system can continuously optimize itself, improve the practicality and reliability of the simulation model, and reduce dependence on physical experiments, reducing costs and time consumption.

[0042] 2. By comprehensively considering the feedback parameters of different interaction directions and states, a comprehensive evaluation of the simulation model performance is achieved. This approach not only improves the accuracy of model verification, but also provides a solid foundation for the continuous improvement and application of simulation models through multi-dimensional error analysis and trend inspection.

[0043] Other advantages, objectives and features of the present invention will be described in the following description and will be apparent to those skilled in the art to some extent, or those skilled in the art may be taught from the practice of the present invention. The objectives and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to make the purpose, technical solution and beneficial effects of the present invention clearer, the present invention provides the following drawings for illustration:

[0045] Figure 1 The present invention is a flowchart of a method for constructing a dynamic simulation model verification system for a digital-realistic test in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0047] The present invention provides a method for constructing a dynamic simulation model verification system for a digital-realistic combined test, referring to Figure 1 ,include:

[0048] Step S1, collecting the operation data of the simulation target and building a digital twin virtual scene based on a preset simulation model;

[0049] Step S2: construct a realistic test scenario of the simulation target and obtain test data in the realistic test scenario;

[0050] Step S3, establishing a data interaction module to construct a test data interaction relationship between the real test scene and the virtual scene, using the real-time data of the simulation target in the virtual scene to guide the simulation target in the real test scene to make the same action feedback or using the real-time data of the simulation target in the real test scene to guide the simulation target in the virtual scene to make the same action feedback;

[0051] Step S4: construct a verification module to synchronously obtain feedback parameters fed back by the actual test scene and the virtual scene, and verify the consistency of dynamic data of the simulation model based on the feedback parameters.

[0052] The working principle and beneficial effects of the above technical solution are as follows: by synchronously acquiring the feedback parameters of the real test scene and the virtual scene, the present invention realizes the dynamic data consistency verification of the simulation model. This synchronous verification mechanism ensures the high-precision comparison between the simulation results and the actual test data, and improves the accuracy of the model prediction. The established data interaction module allows real-time data exchange and action guidance between the two scenes, so that the virtual scene can respond to the changes of the real test scene in real time, and vice versa. This real-time interaction capability significantly improves the dynamic adaptability and flexibility of the simulation system. By constructing a digital twin virtual scene and a real test scene, and realizing data interaction and guidance between the two, a closed-loop digital-real combination simulation verification system is formed. This system can continuously optimize itself, improve the practicality and reliability of the simulation model, and reduce the dependence on physical experiments, reducing costs and time consumption. The present invention constructs a complete digital-real combination simulation verification process, which not only improves the accuracy and efficiency of the simulation, but also provides continuous optimization and updating for the simulation model through real-time data interaction and dynamic verification, ensuring the effectiveness and adaptability of the model throughout the life cycle. The realization of this process is of great significance for the application of simulation technology in engineering design, testing and decision support.

[0053] In a preferred embodiment, collecting the operation data of the simulation target and building a digital twin virtual scene based on a preset simulation model includes:

[0054] Collect various operation data generated by the simulation target in actual operation through deployed sensors and data acquisition equipment;

[0055] Preprocess various operation data separately and store them in the database; the preprocessing includes classification, cleaning, denoising, standardization and normalization;

[0056] Based on the simulation model to be verified, the basic digital twin virtual scene is constructed using the pre-processed operation data;

[0057] Use operational data to calibrate the simulation model to ensure that the digital twin virtual scene accurately reflects the behavior of the simulation target.

[0058] In a preferred embodiment, building a realistic test scenario of a simulation target and obtaining test data in the realistic test scenario includes:

[0059] Build a realistic test environment according to the operation requirements, operating environment and test needs of the simulation target;

[0060] Install test equipment based on the test environment, including all devices that can universally represent the simulation target subject and various types of sensors, data acquisition systems and actuators;

[0061] Determine the test parameters that need to be controlled, monitored and collected during the test, and determine the sampling rules, path channels, classification processing and storage rules for data collection based on the specific conditions of the test environment;

[0062] After the realistic test scenario is built, the test is performed under controlled conditions according to the test requirements, and the test data of the simulation target in actual operation is collected.

[0063] The working principle and beneficial effects of the above technical solution are as follows: by carefully building a realistic test environment according to the specific operation requirements and test requirements of the simulation target, and installing universal equipment and multiple sensors representing the main body of the simulation target, the present invention ensures that the collected test data can accurately reflect the actual operation of the simulation target. By determining the test parameters that need to be controlled, monitored and collected, and formulating data collection rules according to the specific conditions of the test environment, the adaptability of the test environment to different test requirements is improved. This flexibility allows rapid adjustment and optimization of various test scenarios. By formulating detailed data collection rules, path channels, classification processing and storage rules, the present invention realizes efficient management and systematic processing of test data. This systematic data management method not only improves the efficiency of data processing, but also ensures the integrity and traceability of the data, providing a solid foundation for subsequent data analysis and model verification. The present invention realizes an efficient and systematic data management mechanism, which not only optimizes the data collection and processing process, but also provides high-quality data support for data analysis and model verification, significantly improving the quality and efficiency of the entire simulation verification process.

[0064] In a preferred embodiment, the data interaction module performs the following steps:

[0065] Establish the test data interaction relationship between the real test scene and the virtual scene according to the preset test data interaction port, and set the data interaction architecture, interactive communication protocol, interaction frequency, interaction trigger instruction, transmission frequency and synchronization mechanism;

[0066] During the test, when the interaction condition is triggered, the real-time data of the simulation target in the virtual scene is used to guide the simulation target in the actual test scene to make the same action feedback, or the real-time data of the simulation target in the actual test scene is used to guide the simulation target in the virtual scene to make the same action feedback, wherein the trigger condition includes an interaction trigger instruction actively sent by the user, an interaction trigger instruction triggered with a preset time frequency, and an interaction trigger instruction triggered with a preset trigger condition;

[0067] When the data interaction action starts, the data collected from the real test scene and the virtual scene are annotated to generate feedback parameters, where the annotation content includes timestamp annotation, annotation of the data belonging to the initiator / guided end, and data classification at the initiator / guided end.

[0068] The working principle and beneficial effects of the above technical solution are as follows: through the preset test data interaction port and synchronization mechanism, the technical solution realizes efficient data interaction between the real test scene and the virtual scene. This synchronization ensures that the data between the two scenes can be updated in real time, and improves the response speed and interaction quality of the simulation system. The set interaction trigger instructions include multiple modes of user active sending, time frequency triggering, and preset condition triggering, so that the data interaction action can be flexibly triggered according to different test requirements and conditions. This dynamic adaptability provides a variety of interaction methods for the test and enhances the practicality of the system. In the process of data interaction, the collected data is thoroughly annotated, including timestamps, annotations of the end to which the data belongs, and data classification. This meticulous annotation method not only ensures the accurate management of the data, but also provides a clear traceability path for the post-processing and analysis of the data, greatly improving the availability of the data and the reliability of the test results. By comprehensively considering the real-time, flexibility and accuracy of data interaction of the embodiment of the present invention, the technical solution provides a comprehensive data interaction and management framework for the simulation test of digital and real combination, which is of great significance for improving the verification accuracy of the simulation model and the efficiency of the test.

[0069] In a preferred embodiment, the verification module performs the following steps:

[0070] Synchronously obtain feedback parameters from the real test scene and the virtual scene, analyze the feedback parameters, determine the timestamp of the feedback parameters, the label of the data belonging to the initiator / guided end, and the data classification of the data at the initiator / guided end;

[0071] The two sets of feedback parameters are aligned based on the timestamp annotation of the feedback parameters, and the starting and ending positions of the data interaction action on the feedback parameters sorted by the time series are determined based on the annotation of the data belonging to the initiator / guided end, and the two sets of feedback parameters are sliced ​​according to the starting and ending positions;

[0072] Based on the slicing results and the annotation of the data belonging to the initiator / guided end, determine that the feedback parameters corresponding to the real test scene and the virtual scene in each slice belong to the initiator / guided end, and perform binary classification on the slicing results;

[0073] The consistency of the simulation model is verified using the feedback parameters corresponding to the binary classification results and the feedback parameters corresponding to when no data interaction is performed.

[0074] The working principle and beneficial effects of the above technical solution are as follows: by synchronously acquiring the feedback parameters of the real test scene and the virtual scene, and performing timestamp annotation and data attribution annotation, the technical solution realizes the precise alignment of the two sets of data. This accuracy provides a solid foundation for subsequent data analysis and consistency verification. By using timestamp annotation to determine the start and end positions of the data interaction action in the time series and slicing the feedback parameters, the technical solution enhances the traceability of the data interaction action. This transparency helps to deeply understand the dynamic changes in the data interaction process and provides an intuitive reference for the adjustment and optimization of the simulation model. By binary classification of the slicing results and combining the feedback parameters when no data interaction is performed, the technical solution can efficiently verify the consistency of the simulation model. This verification method not only improves the accuracy of the verification process, but also can quickly identify the potential deviations and deficiencies of the model by comparing the data before and after the interaction, thereby providing strong data support for the improvement of the simulation model. The technical solution realizes efficient and in-depth consistency verification of the simulation model through a series of precise data synchronization, alignment, slicing and classification operations. This method not only improves the verification quality of the simulation model, but also accelerates the iterative optimization process of the model by quickly feeding back the model performance.

[0075] In a preferred embodiment, performing consistency verification of dynamic data of the simulation model using feedback parameters corresponding to the binary classification results includes:

[0076] According to the binary classification result, when the simulation target in the virtual scene is used as the initiator and the simulation target in the real test scene is used as the guided end, the feedback parameters of the virtual scene in the corresponding slicing result and the feedback parameters corresponding to the real test scene are the first type of feedback parameter pair;

[0077] According to the binary classification result, when the simulation target in the real test scene is used as the initiator and the simulation target in the virtual scene is used as the guided end, the feedback parameters of the virtual scene in the corresponding slicing result and the feedback parameters corresponding to the real test scene are the second type of feedback parameter pair;

[0078] Determine the feedback parameters of the virtual scene when no data interaction action is performed and the feedback parameters corresponding to the real test scene as the third type of feedback parameter pair;

[0079] Based on the first type of feedback parameter pair, dynamic data consistency verification is performed on the simulation model in the virtual scene relative to the simulation target in the actual test scene to obtain a first verification result;

[0080] Based on the second type of feedback parameter pair, dynamic data consistency verification is performed on the simulation model in the virtual scene relative to the simulation target in the actual test scene to obtain a second verification result;

[0081] Based on the third type of feedback parameter pair, dynamic data consistency verification is performed on the simulation model in the virtual scene relative to the simulation target in the actual test scene to obtain a third verification result;

[0082] According to the first verification result and the third verification result, the error curve corresponding to the third verification result and the error curve in the first verification result are subjected to a parallel trend test using a multi-time point double difference method. If the test result shows that the two curves are in a parallel trend or an asymptotic trend, the accuracy verification of the simulation model is passed; if the test result shows that the two curves are in a dissimilar trend, the accuracy verification is not passed.

[0083] According to the second verification results and the third verification results, the error curve corresponding to the third verification result and the error curve in the second verification result are subjected to a parallel trend test using the multi-time-point double difference method. If the test result shows that the two curves are in a parallel trend or an asymptotic trend, the accuracy verification of the simulation model passes; if the test result shows that the two curves are in a dissimilar trend, the accuracy verification fails.

[0084] The working principle and beneficial effects of the above technical solution are as follows: by distinguishing the feedback parameter pairs of the initiating end and the guided end, the technical solution can perform dynamic data consistency verification for different interaction directions. This two-way verification method improves the accuracy and reliability of the simulation model under different operating conditions. By using the multi-time point double difference method to perform parallel trend inspection on the error curve, the technical solution can accurately evaluate the performance changes of the simulation model before and after data interaction. This method not only reveals the accuracy of the model, but also identifies the deviations that may occur in the model under specific conditions. Through the comprehensive verification of three types of feedback parameter pairs, the technical solution provides a comprehensive model performance evaluation framework. The verification results of the first and second types of feedback parameter pairs enable the model to be evaluated in different interaction scenarios, while the third type of feedback parameter pairs provides benchmark data in the non-interaction state. This comprehensive evaluation method significantly improves the depth and breadth of simulation model verification and provides data support for model optimization and improvement. The technical solution realizes a comprehensive evaluation of the performance of the simulation model by comprehensively considering the feedback parameters of different interaction directions and states. This method not only improves the accuracy of model verification, but also provides a basis for the continuous improvement and application of simulation models through multi-dimensional error analysis and trend inspection.

[0085] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and details without departing from the scope defined by the claims of the present invention.

Claims

1. A method for constructing a dynamic simulation model verification system for a digital-realistic test, characterized in that: include: Collect the operation data of the simulation target and build a digital twin virtual scene based on the preset simulation model; Build a realistic test scenario for the simulation target and obtain test data from the realistic test scenario; Establish a data interaction module to construct the test data interaction relationship between the real test scene and the virtual scene, and use the real-time data of the simulated target in the virtual scene to guide the simulated target in the real test scene to make the same action feedback or use the real-time data of the simulated target in the real test scene to guide the simulated target in the virtual scene to make the same action feedback; A verification module is constructed to synchronously obtain feedback parameters from real test scenarios and virtual scenarios, and the consistency of dynamic data of the simulation model is verified based on the feedback parameters.

2. The method for constructing a dynamic simulation model verification system for a digital-realistic test according to claim 1, characterized in that: The collecting of the operation data of the simulation target and building a digital twin virtual scene based on a preset simulation model includes: Collect various operation data generated by the simulation target in actual operation through deployed sensors and data acquisition equipment; Preprocess various operation data separately and store them in the database; the preprocessing includes classification, cleaning, denoising, standardization and normalization; Based on the simulation model to be verified, the basic digital twin virtual scene is constructed using the pre-processed operation data; Use operational data to calibrate the simulation model to ensure that the digital twin virtual scene accurately reflects the behavior of the simulation target.

3. The method for constructing a dynamic simulation model verification system for a digital-realistic test according to claim 1, characterized in that: The step of building a realistic test scene of a simulation target and obtaining test data in the realistic test scene includes: Build a realistic test environment according to the operation requirements, operating environment and test needs of the simulation target; Install test equipment based on the test environment, including all devices that can universally represent the simulation target subject and various types of sensors, data acquisition systems and actuators; Determine the test parameters that need to be controlled, monitored and collected during the test, and determine the sampling rules, path channels, classification processing and storage rules for data collection based on the specific conditions of the test environment; After the realistic test scenario is built, the test is performed under controlled conditions according to the test requirements, and the test data of the simulation target in actual operation is collected.

4. The method for constructing a dynamic simulation model verification system for a digital-realistic test according to claim 1, characterized in that: The data interaction module execution includes: Establish the test data interaction relationship between the real test scene and the virtual scene according to the preset test data interaction port, and set the data interaction architecture, interactive communication protocol, interaction frequency, interaction trigger instruction, transmission frequency and synchronization mechanism; During the test, when the interaction condition is triggered, the real-time data of the simulation target in the virtual scene is used to guide the simulation target in the actual test scene to make the same action feedback, or the real-time data of the simulation target in the actual test scene is used to guide the simulation target in the virtual scene to make the same action feedback, wherein the trigger condition includes an interaction trigger instruction actively sent by the user, an interaction trigger instruction triggered with a preset time frequency, and an interaction trigger instruction triggered with a preset trigger condition; When the data interaction action starts, the data collected from the real test scene and the virtual scene are annotated to generate feedback parameters, where the annotation content includes timestamp annotation, annotation of the data belonging to the initiator / guided end, and data classification at the initiator / guided end.

5. The method for constructing a dynamic simulation model verification system for a digital-realistic test according to claim 4, characterized in that: Verification module execution includes: Synchronously obtain feedback parameters from the real test scene and the virtual scene, analyze the feedback parameters, determine the timestamp of the feedback parameters, the label of the data belonging to the initiator / guided end, and the data classification of the data at the initiator / guided end; The two sets of feedback parameters are aligned based on the timestamp annotation of the feedback parameters, and the starting and ending positions of the data interaction action on the feedback parameters sorted by the time series are determined based on the annotation of the data belonging to the initiator / guided end, and the two sets of feedback parameters are sliced ​​according to the starting and ending positions; Based on the slicing results and the annotation of the data belonging to the initiator / guided end, determine that the feedback parameters corresponding to the real test scene and the virtual scene in each slice belong to the initiator / guided end, and perform binary classification on the slicing results; The consistency of the simulation model is verified using the feedback parameters corresponding to the binary classification results and the feedback parameters corresponding to when no data interaction is performed.

6. A method for constructing a dynamic simulation model verification system for a digital-realistic test according to claim 5, characterized in that: The verification of the consistency of dynamic data of the simulation model using the feedback parameters corresponding to the binary classification results includes: According to the binary classification result, when the simulation target in the virtual scene is used as the initiator and the simulation target in the real test scene is used as the guided end, the feedback parameters of the virtual scene in the corresponding slicing result and the feedback parameters corresponding to the real test scene are the first type of feedback parameter pair; According to the binary classification result, when the simulation target in the real test scene is used as the initiator and the simulation target in the virtual scene is used as the guided end, the feedback parameters of the virtual scene in the corresponding slicing result and the feedback parameters corresponding to the real test scene are the second type of feedback parameter pair; Determine the feedback parameters of the virtual scene when no data interaction action is performed and the feedback parameters corresponding to the real test scene as the third type of feedback parameter pair; Based on the first type of feedback parameter pair, dynamic data consistency verification is performed on the simulation model in the virtual scene relative to the simulation target in the actual test scene to obtain a first verification result; Based on the second type of feedback parameter pair, dynamic data consistency verification is performed on the simulation model in the virtual scene relative to the simulation target in the actual test scene to obtain a second verification result; Based on the third type of feedback parameter pair, dynamic data consistency verification is performed on the simulation model in the virtual scene relative to the simulation target in the actual test scene to obtain a third verification result; According to the first verification result and the third verification result, the error curve corresponding to the third verification result and the error curve in the first verification result are subjected to a parallel trend test using a multi-time point double difference method. If the test result shows that the two curves are in a parallel trend or an asymptotic trend, the accuracy verification of the simulation model is passed; if the test result shows that the two curves are in a dissimilar trend, the accuracy verification is not passed. According to the second verification results and the third verification results, the error curve corresponding to the third verification result and the error curve in the second verification result are subjected to a parallel trend test using the multi-time-point double difference method. If the test result shows that the two curves are in a parallel trend or an asymptotic trend, the accuracy verification of the simulation model passes; if the test result shows that the two curves are in a dissimilar trend, the accuracy verification fails.

Images