Vehicle in-the-loop test system and method based on video injection, and vehicle

Abstract

The invention discloses a whole vehicle in-the-loop test system and method based on video injection and a vehicle. The system comprises a scene simulation module, a video injection module, a control module and an intelligent driving domain controller. The scene simulation module is used for operating a plurality of virtual scenes so as to output a video stream to the video injection module; the video injection module is used for receiving the video stream and injecting the video stream into the intelligent driving domain controller in a set data format; the control module is used for resolving vehicle dynamics based on the received pose information and scene data transmitted by the scene simulation module and sending a corresponding control instruction; and the intelligent driving domain controller is used for formulating a driving strategy based on the injected video stream and the corresponding control instruction and inputting the driving strategy to the control module so as to execute the corresponding driving strategy in the virtual simulation scene. According to the invention, the vehicle function test accuracy can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle simulation, and in particular to a whole vehicle in-the-loop testing system, method and vehicle based on video injection. Background Art

[0002] In recent years, with the continuous improvement of the level of vehicle intelligence, traditional testing methods are no longer fully applicable. Advanced driver assistance / intelligent driving performs closed-loop control through perception, decision-making, and control. For the simulation test of intelligent driving vehicles, it is necessary to provide simulation of perception and simulation of the controlled actuators. In the traditional intelligent driving controller test, the domain controller of intelligent driving is used as the object to be tested, and perception simulation is realized through video darkroom, bus simulation, etc. For the controlled object, the whole vehicle is modeled (such as powertrain, braking system, steering system, body system, etc.). It is necessary to calibrate the whole vehicle dynamics model through design parameters and test parameters to ensure relatively accurate accuracy. However, due to insufficient parameter input and limitations of dynamic simulation, there is a certain deviation between the vehicle dynamics and the actual vehicle, which has a great impact on the accuracy of the test. In addition, as the main input source of perception, the number of modules increases, and performance parameters are continuously provided. The traditional dark box solution has reached a bottleneck. The video dark box technology is a device that connects the video signal of the virtual simulation scene to the display in the dark box, uses a real camera to shoot video at the display, and transmits the captured video signal to the automatic driving controller through a coaxial cable. It is easily affected by light and the display. The video dark box equipment is large and suitable for monocular cameras. Surround view cameras cannot be used. One dark box only supports one camera and the accuracy is low. At the same time, in the field of traditional power and chassis testing, road load simulation can be completed by loading dynamometers, steering simulation devices, etc., and the vehicle's dynamics and economy can be tested. However, due to the lack of perception simulation, the intelligent driving function cannot be activated, and the complete function test of the whole vehicle cannot be carried out.

[0003] Therefore, how to improve the accuracy of vehicle-in-the-loop testing is a technical problem that needs to be solved urgently. Summary of the invention

[0004] The main purpose of the present invention is to provide a whole vehicle in-the-loop testing system, method and vehicle based on video injection, which can improve the accuracy of whole vehicle function testing.

[0005] In a first aspect, the present application provides a vehicle-in-the-loop testing system based on video injection, comprising:

[0006] Scene simulation module, video injection module, control module, intelligent driving domain controller;

[0007] The scene simulation module is used to run multiple virtual scenes to output video streams to the video injection module;

[0008] The video injection module is used to receive the video stream and inject the video stream into the intelligent driving domain controller in a set data format;

[0009] A control module, which is used to solve the vehicle dynamics based on the posture information and scene data transmitted by the scene simulation module, and send corresponding control instructions;

[0010] The intelligent driving domain controller is used to formulate a driving strategy based on the injected video stream and the corresponding control instructions, and input it to the control module to execute the corresponding driving strategy in the virtual simulation scene.

[0011] In combination with the first aspect above, as an optional implementation, the system further includes a longitudinal loading module, which is used to simulate the longitudinal road load resistance of the vehicle in real time based on the received control instruction;

[0012] The steering loading device is used to simulate the lateral resistance in real time based on the received control commands.

[0013] In combination with the first aspect above, as an optional implementation, the longitudinal loading module includes:

[0014] An electric dynamometer, which is used to independently load the four wheels according to the torque command sent by the control system to simulate the longitudinal road load resistance in real time;

[0015] The torque measurement unit is used to measure the torque and speed of each wheel in real time and output it to the control module for dynamic model solution.

[0016] In combination with the first aspect above, as an optional implementation method, the steering loading device is also used to collect the lateral loading force in real time and feed it back to the control module to perform real-time calculation of the dynamic model.

[0017] In combination with the first aspect above, as an optional implementation, the control module includes:

[0018] Signal acquisition unit, dynamics model and test bench control interface;

[0019] The signal acquisition unit is used to collect the torque and speed of the four wheels and output them to the dynamic model;

[0020] The dynamic model is used to solve the torque and speed of the four wheels collected in real time, calculate the speed, acceleration and posture information of the vehicle, and output it to the scene software to drive the vehicle in the scene to complete the real-time refresh of the scene;

[0021] The bench control interface is used to send the longitudinal control instructions calculated by the dynamic model to the four dynamometers for longitudinal loading, and send the lateral control instructions calculated by the dynamic model to the steering loading device for lateral loading.

[0022] In combination with the first aspect above, as an optional implementation, the control module further includes:

[0023] The residual bus simulation unit is used to receive bus signals simulated by multiple sensors installed on the vehicle and transmit them to the intelligent driving domain controller for decision making.

[0024] In a second aspect, the present application provides a vehicle-in-the-loop testing method based on video injection, comprising:

[0025] Multiple virtual scenes run based on the scene simulation module to output video streams;

[0026] The video injection module injects the video stream into the intelligent driving domain controller in a set data format according to the received video stream;

[0027] Based on the posture information and scene data transmitted by the scene simulation module and received by the control module, the vehicle dynamics are solved and corresponding control instructions are sent;

[0028] The injected video stream and the corresponding control instructions are used as inputs of the intelligent driving domain controller, and a driving strategy is formulated by the intelligent driving domain controller and input into the control module to execute the corresponding driving strategy in a virtual simulation scene.

[0029] In combination with the above second aspect, as an optional implementation manner, the setting format includes: FPD LINK format and GMSL format.

[0030] In combination with the second aspect above, as an optional implementation method, the received video stream is processed based on the FPGA processor to complete the injection of the video stream into the intelligent driving domain controller.

[0031] In a third aspect, an embodiment of the present application further provides a vehicle, which includes the whole vehicle in-the-loop test system described in the first aspect above.

[0032] During the above vehicle-in-the-loop test process, video injection is introduced into the real vehicle test to complete the perception simulation. At the same time, steering loading equipment is added to activate the smart driving function of the smart car and complete the vehicle function test, which can improve the accuracy of the vehicle function test.

[0033] It is to be understood that the foregoing general description and the following detailed description are exemplary only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0035] Figure 1 A schematic diagram of a vehicle-in-the-loop test based on video injection provided in an embodiment of the present application;

[0036] Figure 2 A flow chart of a vehicle-in-the-loop testing method based on video injection provided in an embodiment of the present application;

[0037] Figure 3 A schematic diagram of a vehicle provided in an embodiment of the present application. DETAILED DESCRIPTION

[0038] Here, exemplary embodiments will be described in detail, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0039] Furthermore, the drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. Some of the blocks shown in the drawings are functional entities and do not necessarily correspond to physically or logically separate entities.

[0040] The embodiments of the present application are further described in detail below in conjunction with the accompanying drawings.

[0041] Vehicle-in-the-Loop (VIL) is an advanced testing method that is mainly used to evaluate the performance and functions of a car in an actual road environment. VIL testing places the entire car in a virtual or controlled environment for simulation testing, focusing on verifying the overall performance and synergy of the vehicle in complex driving scenarios.

[0042] Reference Figure 1 , Figure 1 FIG. 1 is a schematic diagram of a vehicle-in-the-loop test system based on video injection provided by the present invention, such as Figure 1 As shown:

[0043] Scene simulation module 101, video injection module 102, control module 103, intelligent driving domain controller 104;

[0044] The scene simulation module is used to run multiple virtual scenes to output video streams to the video injection module.

[0045] Specifically, as a module in the closed-loop test system of intelligent driving vehicles, the scene simulation module needs to run a large number of virtual scenes. The scenes run in the simulation system. The hardware is a high-performance workstation, and a high-performance graphics card is configured to output multiple video streams to the video injection module. The scene data information (vehicle posture information and scene data, transmitted through the CAN bus) is output to the control module through a high-speed network interface for input of dynamic solution and input of bus simulation. The scene simulation software runs in the scene simulation module, and static road networks and dynamic scenes are built on this software platform; the external parameters of the test vehicle camera, such as installation position, installation angle, etc., are set in the software, among which the distortion parameters of the fisheye lens are imported in the software, and the simulation of the fisheye lens is realized through real-time post-processing.

[0046] The video injection module is used to receive the video stream and inject the video stream into the intelligent driving domain controller in a set data format.

[0047] Specifically, the video injection module: The video injection module completes the physical simulation of the camera module, converts the sensor model in the scene simulation module into a video stream and outputs it to the intelligent driving domain controller, and converts the model data (virtual camera model simulation data) into the interface data format of various camera modules, such as FPD LINK, GMSL, etc. In general, the camera consists of front view, rear view, side view and surround view, among which the surround view is mostly a fisheye lens, which requires distortion processing. The video source for video injection comes from the scene simulation module, and the core chip of the video injection simulation circuit is implemented using FPGA. The video stream reception and image processing algorithm are completed by the FPGA processor.

[0048] It also needs to be explained that the video injection module is used to cut off the original camera module, simulate the serializer signal, and complete the injection of the video stream. The video injection board receives the video signals of different camera perspectives of the virtual simulation scene through the HDMI / DVI interface, and after internal image processing, the video signal with a specific protocol (set data format) is injected into the autonomous driving domain controller.

[0049] The control module is used to solve the vehicle dynamics based on the posture information and scene data received from the scene simulation module and send corresponding control instructions.

[0050] Specifically, the control module serves as a data processing, dynamics solving and command sending unit, and is composed of a real-time dynamics model, a signal acquisition unit, a test bench control interface, and a residual bus simulation interface.

[0051] The dynamic model runs in real time in the control module, and is solved based on the torque and speed of the four wheels collected in real time to calculate the vehicle's speed, acceleration, posture and other information, which is output to the scene software to drive the main vehicle in the scene and complete the real-time refresh of the scene.

[0052] The signal acquisition unit collects the torque and speed of the four wheels and outputs them to the dynamic model.

[0053] The test bench control interface sends the longitudinal control instructions to the four dynamometers for longitudinal loading and sends the lateral control instructions to the steering loading device for lateral loading according to the longitudinal loading instructions and lateral loading instructions calculated by the dynamic model.

[0054] The residual bus simulation module receives the real vehicle bus signal and transmits the simulated millimeter-wave radar and other bus signals to the vehicle intelligent driving domain controller to assist decision-making.

[0055] The intelligent driving domain controller is used to formulate a driving strategy based on the injected video stream and the corresponding control instructions, and input it to the control module to execute the corresponding driving strategy in the virtual simulation scene. Specifically, after receiving the video stream and the control instructions, the intelligent driving domain controller performs calculations and formulates a driving strategy.

[0056] In one embodiment, the system further comprises a longitudinal loading module 105, which is used to simulate the longitudinal road load resistance of the vehicle in real time based on the received control instruction;

[0057] The steering loading device 106 is used to simulate the lateral resistance in real time based on the received control command.

[0058] The longitudinal loading module includes: an electric dynamometer, which is used to complete independent loading of four wheels according to the torque command sent by the control system to simulate the longitudinal road load resistance in real time;

[0059] The torque measurement unit is used to measure the torque and speed of each wheel in real time and output it to the control module for dynamic model solution.

[0060] The steering loading device is also used to collect the lateral loading force in real time and feed it back to the control module for real-time calculation of the dynamic model.

[0061] Specifically, the longitudinal loading module consists of four sets of loading modules, each of which contains an electric dynamometer and a torque measurement unit. The electric dynamometer completes the independent loading of the four wheels according to the torque command sent by the control module, thereby simulating the longitudinal road load resistance in real time. The torque measurement unit measures the torque and speed of each wheel in real time and outputs them to the control module for dynamic model solution and test bench control module closed loop.

[0062] Steering loading device: Since the object under test is a smart car, the intelligent driving control system will drive the vehicle to run longitudinally and laterally, so the lateral resistance simulation needs to be realized through the steering loading device. The steering loading device responds to the lateral loading command sent by the control module to complete the lateral resistance simulation of the test vehicle; at the same time, the lateral loading force is collected in real time and fed back to the test bench control module for real-time calculation of the dynamic model and the closed loop of the test bench control module.

[0063] The test intelligent vehicle that needs explanation: It has advanced assisted driving functions, using cameras and radars as perception devices. The camera module simulation is realized by connecting to the video injection device, and the radar sensor simulation is realized by residual bus simulation, thereby activating the test vehicle's intelligent driving function, driving the wheels to output power and steer, and realizing closed-loop operation through real-time loading of the longitudinal loading module and steering loading equipment.

[0064] The video injection equipment used in the traditional intelligent driving domain controller solution is integrated with the traditional vehicle dynamometer system. By replacing most vehicle models with the whole vehicle, the need to model the whole vehicle in the traditional intelligent driving domain controller solution is avoided, which avoids the direct impact on the verification accuracy of the regulation and control algorithm due to modeling accuracy issues.

[0065] Moreover, in the field of traditional vehicle testing, only the power loading of the vehicle is considered to verify the power performance of the vehicle. However, today's smart cars have certain auxiliary and automatic driving capabilities. Due to the lack of perception simulation and lateral simulation on the traditional test bench, the functions cannot be activated. Through this solution, video injection is introduced into the actual vehicle test to complete the perception simulation, and the steering loading equipment is added at the same time to activate the smart driving function of the smart car and complete the whole vehicle function test.

[0066] In one embodiment, the vehicle's sensors are simulated, including camera simulation and radar simulation, wherein the camera simulation uses a video injection system, and interface simulation and distortion simulation are performed according to the actual vehicle camera configuration. According to the actual vehicle parameters, the loading system is adjusted, and the vehicle is installed on the loading system, including the longitudinal loading system and the lateral loading system. The test bench system is configured according to the vehicle parameters. The system workflow is as follows: the driver / assisted driving system controls the vehicle through the accelerator pedal, brake pedal, and steering wheel, the wheel-side torque sensor collects the actual wheel-side torque, the driving dynamics calculates the speed, and the speed command is input to the wheel-side dynamometer system for loading; the lateral loading system collects the angle of the tie rod, and the load force is calculated by dynamics to act on the lateral loading system; the dynamic displacement change drives the scene simulation software to move the main vehicle, and the scene information is refreshed in real time. The scene is fed back to the intelligent driving controller through the video injection system, and the controller receives the perception signal, refreshes the control command, and completes the closed loop.

[0067] In one embodiment, video injection is used as a camera simulation device to customize the development of different camera module forms. The CMOS in the module is simulated through FPGA (field programmable gate array), and the resolution and color space are simulated. The serial chip used in the module is customized and developed and I2C configuration is performed; for fisheye, wide-angle and other distorted lenses, simulation is performed by adding real-time post-processing to the scene simulation software according to the physical parameters of the lens. Only the car body is modeled in the dynamic model software (using the real powertrain and chassis of the real car), and power loading is performed through a four-wheel dynamometer and a steering simulation device to achieve a closed-loop test of the system.

[0068] Reference Figure 2 , Figure 2 FIG. 1 is a flow chart of a vehicle-in-the-loop testing method based on video injection provided by the present invention, as shown in FIG. Figure 2 As shown, the method comprises the steps of:

[0069] Step 201: Output video streams based on multiple virtual scenes run by a scene simulation module.

[0070] Step 202: The video injection module injects the video stream into the intelligent driving domain controller in a set data format according to the received video stream.

[0071] The data formats set include: FPD LINK format and GMSL format.

[0072] In addition, the core chip of the video injection simulation circuit is implemented using FPGA, and the video stream reception and image processing algorithm are completed by the FPGA processor.

[0073] Step 203: Based on the posture information and scene data transmitted by the scene simulation module and received by the control module, the vehicle dynamics are solved and corresponding control instructions are sent.

[0074] Among them, according to the longitudinal loading instructions and lateral loading instructions calculated by the dynamic model, the longitudinal control instructions are sent to the four dynamometers for longitudinal loading, and the lateral control instructions are sent to the steering loading device for lateral loading.

[0075] Step 204: Use the injected video stream and the corresponding control instructions as input to the intelligent driving domain controller, formulate a driving strategy through the intelligent driving domain controller, and input it to the control module to execute the corresponding driving strategy in the virtual simulation scene.

[0076] Furthermore, in a possible implementation manner, the video injection module is also used to set the format to include: FPDLINK format and GMSL format.

[0077] Furthermore, in a possible implementation, the intelligent driving domain controller is also used to process the received video stream based on the FPGA processor to complete the injection of the video stream into the intelligent driving domain controller.

[0078] like Figure 3 As shown, the present invention also provides a vehicle 300, the vehicle 300 includes Figure 3 A vehicle-in-the-loop test system 200 is shown.

[0079] The video injection equipment used in the traditional intelligent driving domain controller solution is integrated with the traditional vehicle dynamometer system. By replacing most vehicle models with the whole vehicle, the need to model the whole vehicle in the traditional intelligent driving domain controller solution is avoided, which directly affects the verification accuracy of the regulation and control algorithm due to the modeling accuracy problem. In the traditional whole vehicle testing field, only the power loading of the vehicle is considered to verify the vehicle's power performance. Today's smart cars have certain auxiliary and automatic driving capabilities. Due to the lack of perception simulation and lateral simulation on the traditional test bench, the function cannot be activated. Through this solution, video injection is introduced into the actual vehicle test to complete the perception simulation. At the same time, steering loading equipment is added to activate the smart driving function of the smart car and complete the whole vehicle function test.

[0080] The above description is only a specific implementation of the present application, so that those skilled in the art can understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest range consistent with the principles and novel features applied for herein.

Claims

1. A vehicle-in-the-loop test system based on video injection, characterized in that: include: Scene simulation module, video injection module, control module, intelligent driving domain controller; The scene simulation module is used to run multiple virtual scenes to output video streams to the video injection module; The video injection module is used to receive the video stream and inject the video stream into the intelligent driving domain controller in a set data format; A control module, which is used to solve the vehicle dynamics based on the posture information and scene data transmitted by the scene simulation module, and send corresponding control instructions; The intelligent driving domain controller is used to formulate a driving strategy based on the injected video stream and the corresponding control instructions, and input it to the control module to execute the corresponding driving strategy in the virtual simulation scene.

2. The system according to claim 1, characterized in that Also includes: A longitudinal loading module, which is used to simulate the longitudinal road load resistance of the vehicle in real time based on the received control command; The steering loading device is used to simulate the lateral resistance in real time based on the received control commands.

3. The system according to claim 2, characterized in that The longitudinal loading module comprises: An electric dynamometer, which is used to independently load the four wheels according to the torque command sent by the control system to simulate the longitudinal road load resistance in real time; The torque measurement unit is used to measure the torque and speed of each wheel in real time and output it to the control module for dynamic model solution.

4. The system according to claim 2, characterized in that: The steering loading device is also used to collect the lateral loading force in real time and feed it back to the control module for real-time calculation of the dynamic model.

5. The system according to claim 1, characterized in that The control module includes: Signal acquisition unit, dynamics model and test bench control interface; The signal acquisition unit is used to collect the torque and speed of the four wheels and output them to the dynamic model; The dynamic model is used to solve the torque and speed of the four wheels collected in real time, calculate the speed, acceleration and posture information of the vehicle, and output it to the scene software to drive the vehicle in the scene to complete the real-time refresh of the scene; The bench control interface is used to send the longitudinal control instructions calculated by the dynamic model to the four dynamometers for longitudinal loading, and send the lateral control instructions calculated by the dynamic model to the steering loading device for lateral loading.

6. The system according to claim 5, characterized in that The control module also includes: The residual bus simulation unit is used to receive bus signals simulated by multiple sensors installed on the vehicle and transmit them to the intelligent driving domain controller for decision making.

7. A vehicle-in-the-loop testing method based on video injection, characterized in that: include: Based on multiple virtual scenes running in the scene simulation module, the video stream is output; The video injection module injects the video stream into the intelligent driving domain controller in a set data format according to the received video stream; Based on the posture information and scene data transmitted by the scene simulation module and received by the control module, the vehicle dynamics are solved and corresponding control instructions are sent; The injected video stream and the corresponding control instructions are used as inputs of the intelligent driving domain controller, and a driving strategy is formulated by the intelligent driving domain controller and input into the control module to execute the corresponding driving strategy in a virtual simulation scene.

8. The method according to claim 7, characterized in that include: The setting formats include: FPDLINK format and GMSL format.

9. The method according to claim 7, characterized in that: include: The received video stream is processed based on the FPGA processor to complete the injection of the video stream into the intelligent driving domain controller.

10. A vehicle, characterized in that: A vehicle-in-the-loop test system comprising any one of claims 1-6.

Images