Algorithm verification system and verification method based on digital twinning

Abstract

The application belongs to the field of vehicles, and particularly relates to an algorithm verification system and method based on digital twinning. The system comprises four subsystems, namely a digital bench system, a physical bench system, a real vehicle bench system and an interactive terminal system. The digital bench system realizes switching and fusion of multiple working conditions and multiple scenes and input of virtual sensor data through software. The real vehicle system and the physical bench system receive input of the digital bench system through a communication connection module, thereby performing testing and outputting measured data to the interactive terminal system. The system covers three types of testing, namely pure model simulation, hardware-in-the-loop sensor data testing based on digital twinning and real vehicle interactive testing based on digital twinning. The four subsystems of the application realize interactive combination based on API, can effectively test chassis domain braking and steering performance, and can effectively verify accuracy of vehicle lateral and longitudinal control algorithms.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of vehicles, and particularly relates to an algorithm verification system and method based on digital twinning. BACKGROUND

[0002] Digital twinning is a simulation process integrating multi-discipline, multi-physical quantity, multi-scale and multi-probability, which fully utilizes physical models, sensor updates, operation history and other data, and completes mapping in a virtual space, thereby reflecting the whole life cycle process of the corresponding entity equipment.

[0003] Digital twinning technology is initially applied to the health maintenance and support of aerospace vehicles. With the continuous improvement and development of this technology, it is currently widely applied in the field of engineering construction and the field of intelligent manufacturing. The purpose is to predict the response of the entity equipment and evaluate the next state through the operation analysis and sufficient information interaction of the digital mapping model.

[0004] Digital twinning is still a relatively new concept in the automotive industry. In recent years, the China Information and Communication Technology Research Institute has developed an automatic driving test system based on digital twinning, which uses Internet of Vehicles communication technology to realize sensor data uploading and virtual scene information publishing process, and has carried out road vehicle testing. A cloud-based digital twinning reference model for assisted driving has been proposed, and a digital twinning example: an advanced driver assistance system based on car-to-cloud has also been applied.

[0005] At present, in the automobile performance test bench, steering test and braking test are an important verification link. However, the existing eps (electronic power steering system) and i-booster (electronic power braking system) test benches are mostly based on given data for hardware-in-the-loop experiments, which cannot interact with real vehicles and make predictions during the experiment, and cannot obtain test data based on real vehicles, which makes the results of braking response and steering assist test in the automobile performance test bench have the problem of low accuracy, and cannot meet the requirements of high accuracy and high reality of automobile performance test. And in the algorithm verification, due to the lack of twinning data, it is difficult to realize accurate model state analysis and prediction, and complete closed-loop verification cannot be realized. Therefore, it is very important to design a test bench based on digital twinning. SUMMARY

[0006] In view of the above-mentioned deficiencies existing in the prior art, the present application provides a test bench based on digital twinning for the eps and i-booster joint simulation test bench. By using CAN bus communication technology and Internet of Vehicles communication technology, three levels of simulation tests are proposed under the consideration of vehicle lateral and longitudinal control, information interaction is realized based on digital twinning, and relatively accurate algorithm verification is achieved.

[0007] The technical solution for achieving the object of the application is: an algorithm verification system based on digital twinning, comprising a digital bench system, a physical bench system, a real vehicle system and an interactive terminal system;

[0008] The physical bench system is a combined bench for electronic power steering systems and electronic power braking systems, used for vehicle performance test benches;

[0009] The digital bench system is a mapping of the physical bench system and the real vehicle system in a digital information space, which is a digital representation of the physical bench system and the real vehicle system in a virtual space;

[0010] The real vehicle system is a real vehicle test based on specified complex multi-condition fusion, virtual scene construction and virtual sensor data input in a real space, and the real vehicle system comprises a vehicle to be tested and a communication connection module for twinning;

[0011] The interactive terminal system sends information instruction stream 1 and information instruction stream 2 to the physical bench system and the real vehicle system respectively, and the physical bench system and the real vehicle system feed back data stream 1 and data stream 2 to the digital bench system respectively;

[0012] The digital bench system realizes the switching and fusion of multiple conditions and multiple scenes, as well as the input of virtual sensor data, and the real vehicle system and the physical bench system receive the input of the digital bench system through the communication connection module to perform testing and output the measured data to the interactive terminal system.

[0013] Further, the digital bench system comprises a mapping module, a data storage module and a knowledge information module;

[0014] The mapping module comprises a physical bench mapping module and a real vehicle mapping module,

[0015] The data storage module comprises a model simulation data module, a physical bench feedback data module and a real vehicle feedback data module,

[0016] The knowledge information module comprises physical characteristic description information of the physical bench and the real vehicle.

[0017] Further, the physical bench mapping module refers to the digital model representation of the physical bench system in a virtual space, comprising a dynamics simulation module, a feedback data input module and a scene condition module;

[0018] The real vehicle mapping module respectively refers to the digital model representation of the real vehicle system in a virtual space, and the real vehicle mapping module comprises a whole vehicle dynamics simulation module, a feedback data input module and a twinning condition module.

[0019] Further, the model simulation data module, the physical bench feedback data module and the real vehicle feedback data module respectively refer to storage of model simulation data, physical bench feedback data and real vehicle feedback data;

[0020] The digital bench system obtains various types of expected outputs and actual outputs by calling an application interface to run a dynamics simulation module based on different sources of input data and different inputs of scene working conditions.

[0021] Further, data instructions are issued through an interactive terminal system to realize responses of the real vehicle system and the physical bench system, and the interactive terminal system comprises a controller module, a function module and a twin database module.

[0022] Further, the controller module comprises a GUI interface, and the GUI interface establishes control communication of real-time can information instruction transceiving and serial port information feedback by calling an application interface;

[0023] The twin database module is a database for storing twin data, and the twin data is generated, selected, corrected, optimized and fused from entity running data, sensor updating data and historical data.

[0024] Further, the function module comprises an algorithm verification module, an online simulation module and an auxiliary decision-making module.

[0025] The algorithm verification module is configured to verify an automatic driving algorithm and a vehicle lateral and longitudinal control algorithm based on a form of application interface calling.

[0026] The online simulation module is configured to realize running of a dynamics simulation module in a digital bench based on multiple input working conditions and scenes and multiple types of data input, and obtain index outputs under multiple tests.

[0027] The auxiliary decision-making module is configured to realize simulation completion and prediction information generation in the digital bench based on twin data and feedback information, issue data instructions through an interactive terminal system, and realize responses of a real vehicle and a physical bench.

[0028] Further, the physical bench system adopts a whole vehicle controller WCU as a control unit and realizes information interaction based on can bus communication technology.

[0029] A use of the above system is used for pure model simulation, hardware-in-the-loop sensor data testing based on digital twinning, or real vehicle interaction testing based on digital twinning.

[0030] A method for algorithm verification by using the above system comprises the following steps:

[0031] Step (1): Establish the ontology module and knowledge information module in the digital bench system:

[0032] Map the physical bench system and the real vehicle system, set the fusion working conditions under multiple scenarios, and run the online simulation module in the terminal interactive system according to the preset expected value and performance index;

[0033] Step (2): Establish the mapping relationship between the dynamic attributes of the ontology model and the perception data of the physical bench:

[0034] After setting the expected value and performance index, the controller module in the terminal interactive system uses the wcu control unit and the can bus communication carrier in the physical bench system to realize the real-time sending of information instruction stream 2 to the physical bench based on the corresponding communication protocol and using the can card as the carrier;

[0035] Step (3): Physical bench system starts execution:

[0036] The physical bench system receives instructions through the reserved bus interface, and then the electronic power steering system and the electronic power braking system identify their respective instructions based on their ECUs and communication protocols and execute responses; the response information is collected by the angle sensor and torque sensor, and based on the sensor information conversion module, the data is fed back to the ECU of each subsystem through the rs485 serial port;

[0037] Step (4): Collect the perception data of the physical bench and update the sensor database:

[0038] The data and information are fed back to the physical bench feedback data module through the reserved bus interface by the can card, and after updating the twin database, the information instruction stream 4 is sent to the physical bench in real time through the bus interface;

[0039] Step (5): Real vehicle test:

[0040] The digital bench system simulates the corresponding sensor data and road input parameters based on the preset multiple test working conditions, sends instruction information stream 1 to the real vehicle through the interactive terminal system based on the application interface and through the communication connection module;

[0041] Step (6): Collect and evaluate data:

[0042] The real vehicle executes the command according to the received information stream, and each sensor installed on the real vehicle collects information and sends data stream to the real vehicle feedback data module in the digital bench system; the interactive terminal system updates the twin database based on the data storage module, and sends information instruction stream 3 to the real vehicle;

[0043] Step (7): Optimize the simulation model: update the online simulation module in the interactive terminal system based on the twin database model and the ontology model;

[0044] Step (8): Data fusion: The twin database iterates the twin data based on the obtained real-time physical test bench feedback data and actual vehicle feedback data;

[0045] Step (9): Verify the algorithm:

[0046] By leveraging the fused iterative twin data, the algorithm verification module is run and the actual vehicle response is observed, thus verifying the lateral and longitudinal control algorithm based on digital twins.

[0047] Compared with the prior art, the significant advantages of this invention are:

[0048] This invention applies digital twins to traditional test benches, predicting and guiding the actual vehicle test bench based on the analysis of historical and real-time data. The original hardware-in-the-loop test is improved into three different levels of testing. The verification system of this invention can efficiently and accurately verify the control algorithm and fully simulate actual scenarios and working conditions, making the steering and braking performance tests more realistic. Attached Figure Description

[0049] Figure 1 This is a diagram of the overall framework of the test bench based on digital twins.

[0050] Figure 2 This is a diagram of the digital test bench.

[0051] Figure 3 This is a second-level framework diagram for bench-based hardware-in-the-loop sensor data testing.

[0052] Figure 4 This is a third-level framework diagram for real-vehicle interaction testing based on digital twins. Detailed Implementation

[0053] The present invention will now be described in further detail with reference to the accompanying drawings.

[0054] like Figure 1 The vehicle algorithm verification test bench system based on digital twin shown includes four subsystems: digital bench system, physical bench system, actual vehicle system, and interactive terminal system.

[0055] The digital test bench system is a mapping of the physical test bench system and the actual vehicle system in the digital information space, while the physical test bench system and the actual vehicle system are the specific manifestations of the digital test bench system in the real space.

[0056] The interactive terminal system sends information command stream 1 and information command stream 2 to the physical test bench system and the actual vehicle system. The physical test bench system and the actual vehicle system respectively feed back data stream 1 and data stream 2 to the physical test bench feedback data module and the actual vehicle feedback data module in the digital test bench data storage module.

[0057] The model simulation data module, the physical bench feedback data module and the real vehicle feedback data module in the digital bench system data storage module integrate respective data and then send data stream 3 to the twin database module in the interactive terminal system;

[0058] The twin database module sends information instruction stream 3 and information instruction stream 4 to the physical bench system and the real vehicle system;

[0059] Among them, information instruction stream 1 and information instruction stream 2 are instruction information, data stream 1 and data stream 2 are data collected by each sensor in the physical bench and the real vehicle;

[0060] Data stream 3 is integrated data of data stream 1 and data stream 2, and information instruction stream 3 and information instruction stream 4 are digital twin information sent after the twin database fuses historical data, entity running data and sensor update data.

[0061] The physical bench system is a bench in a real space, comprising a component module and a communication module. The component module comprises a brake disc, a pressure gauge, a can card, a can bus, a storage battery, a brake master cylinder, a steering structure, a torque sensor, an angle sensor, a vehicle speed sensor, etc. The communication module comprises a can bus and an RS485 serial port.

[0062] The digital bench system is a digital representation of the physical bench system and the real vehicle system in a virtual space. The digital bench system comprises a mapping module, a data storage module and a knowledge information module. The mapping module comprises a physical bench mapping module and a real vehicle mapping module. The data storage module comprises a model simulation data module, a physical bench feedback data module and a real vehicle feedback data module. The knowledge information module comprises physical property description information of the physical bench and the real vehicle, including the inherent properties, size, material of the test body and environmental factors of system operation.

[0063] The real vehicle system is a real vehicle test in a real space based on specified complex multi-working condition fusion, facing virtual scene construction and virtual sensor data input. The real vehicle system comprises a real vehicle for twinning and a communication connection module.

[0064] The interactive terminal system comprises a controller module, a function module and a twin database module.

[0065] The physical bench system comprises a joint bench of an eps system and an i-booster system, adopts a whole vehicle controller WCU as a control unit, realizes information interaction based on can bus communication technology, can test chassis performance under different working conditions based on input simulation of multi-working condition, multi-scene and sensor data, and can be used to verify the accuracy of automobile horizontal and vertical control algorithm.

[0066] The digital bench system is a digital embodiment of the physical bench system and the real vehicle system in virtual space.

[0067] The physical bench mapping module and the real vehicle mapping module respectively refer to the digital model embodiment of the physical bench system and the real vehicle system in virtual space.

[0068] The physical bench mapping module includes a dynamics simulation module, a feedback data input module, and a scene working condition module.

[0069] The real vehicle mapping module includes a whole vehicle dynamics simulation module, a feedback data input module, and a twin working condition module.

[0070] The model simulation data module, the physical bench feedback data module, and the real vehicle feedback data module respectively refer to the storage of model simulation data, physical bench feedback data, and real vehicle feedback data.

[0071] The knowledge information module includes physical characteristics description information of the physical bench and the real vehicle, including the inherent characteristics, size, material of the test body, and environmental factors of system operation.

[0072] The digital bench system, based on different sources of input data and different inputs of scene working conditions, runs the dynamics simulation module through the form of calling api (i.e., application interface, referring to the implementation of each function in the form of encapsulation by each component, and reserving the application interface, which can realize the connection and function of each system without accessing the internal part of the encapsulation body), and finally obtains each type of expected output and actual output.

[0073] The api includes digital twin api, data back annotation api, line control simulation api, scene working condition api, etc.

[0074] The real vehicle system includes a test vehicle in a real road environment and a communication connection module. The digital bench realizes the switching and fusion of multiple working conditions and multiple scenes through software such as trucksim, as well as the input of virtual sensor data. The real vehicle bench receives the input of the digital bench through the communication connection module, thereby performing testing and outputting the measured data to the interactive terminal system. The communication connection module includes the interactive flow of data stream 2 and data stream 3 through vehicle networking technology.

[0075] The interactive terminal system includes a controller module, a function module, and a twin database module.

[0076] The controller module comprises a self-developed GUI interface; the GUI interface is developed based on a Matlab / App Designer environment, a control communication module with certain real-time performance of can information instruction transceiving and serial port information feedback is established by calling an api, and the GUI interface can be used to realize data feedback, twin prediction and other functions in the functional module.

[0077] The twin database module is a database for storing twin data, and the twin data is generated, selected, corrected, optimized and fused from entity running data, sensor update data and historical data.

[0078] The entity running data, the sensor update data and the historical data are generated, selected, corrected, optimized and fused to form the twin data, and the method comprises the following steps.

[0079] Step 1: The data collected by the sensor is supplemented and selected, and the abnormal values are removed, and the appropriate values are selected;

[0080] Step 2: The real-time data sensed by the sensor is associated and integrated with the historical sensing data and the model data, and is fused into twin data;

[0081] Step 3: The real-time data, the historical sensing data and the model data are optimized, the twin data is iterated and saved to the cloud twin database.

[0082] The functional module of the interactive terminal system comprises an algorithm verification module, an online simulation module and an auxiliary decision-making module.

[0083] The algorithm verification module is configured to verify an automatic driving algorithm (such as ACC / AEB) and a vehicle longitudinal and lateral control algorithm based on an api call.

[0084] The online simulation module is configured to realize the running of a dynamics simulation module in a digital bench based on multiple input working conditions and scenes and multiple types of data input, and obtain index outputs under multiple tests.

[0085] The auxiliary decision-making module is configured to realize simulation completion and prediction information generation in the digital bench based on the twin data and feedback information, and realize the response of a real vehicle and a physical bench by issuing data instructions through the interactive terminal system.

[0086] The first degree of test is simulation in a pure virtual space, that is, based on a virtual bench model and a virtual vehicle model of a body module in a digital bench system, a knowledge information module is updated based on a physical bench and a real vehicle bench, an online simulation module and an algorithm verification module in a functional module in a terminal interactive system are run, a simulation model is run and an algorithm is verified.

[0087] The second degree of testing is a hardware-in-the-loop sensor data test based on digital twinning, that is, as shown in Figure 3 As shown in the figure, the ontology module and the knowledge information module of the digital bench system are updated according to the physical bench system, the controller module in the terminal interaction system gives information instruction stream 1, the physical bench system receives the command through the communication module and performs the operation, and sends data stream 1 to the physical bench feedback data module in the digital bench system. The terminal interaction system updates its twin database module, performs online simulation based on the data in the twin database, and sends information instruction stream 3 to the physical bench system. The digital bench realizes the update of feedback data and the setting of working condition scene based on scene working condition api and line control simulation api, and obtains simulation output by running simulation. According to the bench execution response and the simulation result, the algorithm verification module is realized.

[0088] The third degree of testing is a real vehicle interaction test based on digital twinning, that is, as shown in Figure 4 As shown in the figure, the ontology module and the knowledge information module of the digital bench system are updated according to the physical bench system and the real vehicle system, the controller module in the terminal interaction system gives information instruction stream 1 and information instruction stream 2, the physical bench system and the real vehicle system respectively receive the command through the communication module and the communication connection module and perform the operation, and send data stream 1 and data stream 2 to the physical bench feedback data module and the real vehicle feedback data module in the digital bench system. The terminal interaction system updates its twin database module, performs online simulation based on the data in the twin database, and sends information instruction stream 3 and information instruction stream 4 to the physical bench system and the real vehicle system. The digital bench realizes the update of feedback information and sensor information to the input of the online simulation module based on the digital twinning api, and obtains simulation output by running simulation based on the preset working condition. The information interaction between the real vehicle and the digital bench is carried out in real time, the prediction signal is injected into the real vehicle communication module based on the data update of the twin prediction model in the digital bench and the interaction terminal system, the auxiliary prediction is realized, and the real vehicle response is obtained. With the help of digital twinning, the vehicle lateral and longitudinal control algorithm is verified.

[0089] Embodiment

[0090] Vehicle lateral and longitudinal control algorithm verification based on digital twinning

[0091] According to the above system, a vehicle lateral and longitudinal control algorithm verification method based on digital twinning is given, the accuracy of the control algorithm and the influence on the vehicle performance are studied, including the following steps:

[0092] Step 1, establishing the ontology module and the knowledge information module in the digital bench system:

[0093] As shown in Figure 4As shown, the physical bench system and the real vehicle system are digitally mapped, the fusion setting conditions under multiple scenarios are set, the online simulation module in the terminal interaction system is run according to the preset expected value and performance index.

[0094] Step 2, establish the mapping relationship between the dynamic attributes of the ontology model and the perception data of the physical bench:

[0095] After setting the expected value and performance index, through the controller module in the terminal interaction system, with the help of the wcu control unit in the physical bench system and the can bus communication carrier, based on the corresponding communication protocol, the information instruction stream 2 is realized to the physical bench in real time by taking the can card as the carrier.

[0096] Step 3, the physical bench starts to execute:

[0097] The physical bench receives the instructions through the reserved bus interface, and then each subsystem (eps system, i-booster system) identifies and executes the response based on the respective ecu relying on the communication protocol. The corner sensor and torque sensor collect the information after the response, and based on the sensor information conversion module, realize the feedback of data to each subsystem ecu through the rs485 serial port.

[0098] Step 4, collect the perception data of the physical bench, update the sensor database:

[0099] The data and information are fed back to the physical bench feedback data module through the reserved bus interface by means of the can card, and after updating the twin database, the information instruction stream 4 is sent to the physical bench in real time by means of the bus interface.

[0100] Step 5, real vehicle test

[0101] The real driving information and environmental information in the real vehicle system are composed of real vehicle test equipment and test site.

[0102] The digital bench simulates the corresponding sensor data and road input parameters based on the preset multiple test conditions, sends the instruction information stream 1 to the real vehicle through the interactive terminal based on the api through the communication connection module.

[0103] Step 6, collect and evaluate data:

[0104] The real vehicle executes the command according to the received information stream, and each sensor (vehicle speed sensor, corner sensor, torque sensor, brake pressure sensor, etc.) installed on the real vehicle collects information and sends data stream to the real vehicle feedback data module in the digital bench.

[0105] The interactive terminal system updates the twin database based on the data storage module, and sends the information instruction stream 3 to the real vehicle.

[0106] Step 7, optimization of simulation model:

[0107] Based on the twin database model and the ontology model, update the online simulation module in the interactive terminal system.

[0108] Step 8, fusion data

[0109] The twin database iterates the twin data according to the obtained real-time physical bench feedback data and real vehicle feedback data.

[0110] Step 9, algorithm verification

[0111] With the fused iterative twin data, run the algorithm verification module and view the real vehicle response, thereby realizing the verification of the horizontal and vertical control algorithm based on digital twinning.

[0112] With the established digital bench, realize the connection of the algorithm model and the dynamics model, realize the connection of the upper computer and the lower computer and the bottom layer execution component according to the real-time control function of the bench. And according to the iterative twin data fused by the bench, realize the input of the sensor signal and the model parameter in multiple working conditions and multiple scenes, combine the algorithm module to finally output the response curve and the actual result, and the actual response of the bottom layer actuator, and complete the verification of the accuracy of the algorithm through the comparison with the preset index.

Claims

1. A digital-twin-based algorithm verification system, characterized in that, The digital bench system, the physical bench system, the real vehicle system and the interactive terminal system are included; The physical bench system is a combined bench for electronic power steering system and electronic power braking system, which is used for vehicle performance test bench; The digital bench system is a mapping of the physical bench system and the real vehicle system in the digital information space, which is a digital embodiment of the physical bench system and the real vehicle system in the virtual space; The real vehicle system is a real vehicle test in the real space based on specified complex multi-working conditions, virtual scene construction and virtual sensor data input, and the real vehicle system includes a to-be-tested vehicle and a communication connection module for twinning; The interactive terminal system sends information instruction stream 1 and information instruction stream 2 to the physical bench system and the real vehicle system respectively, and the physical bench system and the real vehicle system feed back data stream 1 and data stream 2 to the digital bench system respectively; The digital bench system realizes switching and fusion of multi-working conditions and multi-scenes, and input of virtual sensor data through software, and the real vehicle system and the physical bench system realize receiving of the input of the digital bench system through the communication connection module, so as to perform tests and output measured data to the interactive terminal system; The digital bench system includes a mapping module, a data storage module and a knowledge information module; The mapping module includes a physical bench mapping module and a real vehicle mapping module, The data storage module includes a model simulation data module, a physical bench feedback data module and a real vehicle feedback data module, The knowledge information module includes physical characteristics description information of the physical bench and the real vehicle.

2. The system of claim 1, wherein, The physical bench mapping module is a digital model embodiment of the physical bench system in the virtual space, which includes a dynamics simulation module, a feedback data input module and a scene working condition module; The real vehicle mapping module is a digital model embodiment of the real vehicle system in the virtual space, and the real vehicle mapping module includes a whole vehicle dynamics simulation module, a feedback data input module and a twinning working condition module.

3. The system of claim 2, wherein, The model simulation data module, the physical bench feedback data module and the real vehicle feedback data module respectively refer to storage of model simulation data, physical bench feedback data and real vehicle feedback data; The digital bench system obtains various types of expected output and actual output by calling an application interface based on different sources of input data and different inputs of scene working conditions.

4. The system of claim 3, wherein, The interactive terminal system includes a controller module, a function module and a twinning database module, and data instructions are issued through the interactive terminal system to realize responses of the real vehicle system and the physical bench system.

5. The system of claim 4, wherein, The controller module includes a GUI interface, which establishes a control communication with real-time can bus information instruction transceiving and serial port information feedback through calling an application interface; The twinning database module is a database for storing twinning data, and the twinning data is generated, selected, corrected, optimized and fused from entity running data, sensor updating data and historical data.

6. The system of claim 5, wherein, The function module includes an algorithm verification module, an online simulation module and an auxiliary decision-making module; The algorithm verification module is used for verifying automatic driving algorithms and vehicle lateral and longitudinal control algorithms based on calling of an application interface. The online simulation module realizes the operation of the dynamic simulation module in the digital bench based on multiple input working conditions and scenes and multiple types of data input, and obtains index outputs under multiple tests. The auxiliary decision-making module realizes simulation completion and prediction information generation in the digital bench based on twin data and feedback information, and realizes the response of the real vehicle and the physical bench through the issuance of data instructions through the interactive terminal system.

7. The system of claim 1, wherein, The physical bench system adopts a whole vehicle controller WCU as a control unit and realizes information interaction based on can bus communication technology.

8. The system of any one of claims 1-7, wherein, It is used for pure model simulation, hardware-in-the-loop sensor data testing based on digital twinning, or real vehicle interaction testing based on digital twinning.

9. A method for algorithm verification using the system of claim 7, characterized by, It comprises the following steps: Step (1): Establishing the ontology model and knowledge information module in the digital bench system: Digitally map the physical bench system and the real vehicle system, set the working conditions under the fusion of multiple scenes, and run the online simulation module in the terminal interactive system according to the preset expected value and performance index; Step (2): Establishing the mapping relationship between the dynamic attributes of the ontology model and the perception data of the physical bench: After setting the expected value and performance index, the controller module in the terminal interactive system uses the wcu control unit and the can bus communication carrier in the physical bench system, based on the corresponding communication protocol, to realize the real-time sending of information instruction stream 2 to the physical bench with the can card as the carrier; Step (3): Physical bench system starts execution: The physical bench system receives instructions through the reserved bus interface, and then the electronic power steering system and the electronic power braking system identify and execute their respective instructions based on the communication protocol relying on their respective ECUs; the response information is collected by the turn sensor and the torque sensor, and based on the sensor information conversion module, the data is fed back to the ECU of each subsystem through the RS485 serial port; Step (4): Collecting the perception data of the physical bench and updating the sensor database: Collect the perception data of the physical bench, and feed back to the physical bench feedback data module through the reserved bus interface by the can card, update the twin database, and then send information instruction stream 4 to the physical bench in real time through the bus interface; Step (5): Real vehicle testing: The digital bench system simulates corresponding sensor data and road input parameters based on preset multiple test working conditions, sends instruction information stream 1 to the real vehicle through the interactive terminal system based on the application interface and through the communication connection module; Step (6): Data collection and evaluation: The real vehicle executes the command according to the received information stream, each sensor installed on the real vehicle collects information and sends data stream to the real vehicle feedback data module in the digital bench system; the interactive terminal system updates the twin database based on the data storage module, and sends information instruction stream 3 to the real vehicle; Step (7): Optimize the simulation model: update the online simulation module in the interactive terminal system based on the twin database model and the ontology model; Step (8): Fusion data: the twin database iterates the twin data according to the obtained real-time physical bench feedback data and real vehicle feedback data; Step (9): Verify the algorithm: With the help of the integrated iterative twin data, run the algorithm verification module and view the real vehicle response to realize the verification of the horizontal and longitudinal control algorithm based on digital twin.

Images