Vehicle automatic driving simulation test system

Abstract

The invention relates to a simulation test system for automatic driving of a vehicle, which belongs to the technical field of simulation test and comprises a scene simulation component, an automatic driving algorithm module and a middleware module which are deployed in a first operating system environment, the vehicle dynamics component and the automatic evaluation component are deployed in a second operating system environment; the automatic driving algorithm module is used for outputting a control instruction and internal algorithm parameters; the middleware module is used for communicating the communication between the first operating system environment and the second operating system environment; the vehicle dynamics assembly is used for carrying out vehicle dynamics response based on the control instruction and calculating dynamics parameters of the target vehicle; and the automatic evaluation assembly is used for testing the automatic driving algorithm of the vehicle according to the internal algorithm parameters and the kinetic parameters of the vehicle. The technical problems that evaluation parameters of an automatic driving simulation test system are not comprehensive, and a vehicle dynamics model is too simple, so that the reliability of the model is insufficient are solved.

Description

Technical Field

[0001] The present invention relates to the field of simulation testing technology, and in particular to a vehicle automatic driving simulation testing system. Background Art

[0002] Autonomous driving SIL simulation testing is an efficient simulation testing method. It is particularly suitable for testing autonomous driving planning and control algorithms. The simulation system can simulate different traffic flow scenarios and test and verify the functions and performance of the algorithm. The mainstream SIL simulation system is mainly composed of traffic scenario components, simple vehicle dynamics models and planning algorithms. It mainly performs functional level testing on planning algorithms.

[0003] A complete SIL simulation system needs to have the characteristics of high credibility, comprehensive evaluation, flexible adaptation and scalability. The existing SIL test system has good testing capabilities in the early stage of software development, and can perform functional verification of planning algorithms based on a variety of traffic flow scenarios. However, the current SIL simulation system still has the following problems: 1) Because the vehicle dynamics model is too simple, the simulation credibility is insufficient, and the overall performance of the autonomous driving planning and control algorithm cannot be effectively evaluated; 2) The evaluation parameters are not comprehensive. In addition to the target parameters (lane lines, obstacle distances, etc.) and the vehicle's dynamic parameters (acceleration, etc.), the complete evaluation parameters also require algorithm internal parameters (fusion target distance, behavioral decisions, planning speed, etc.). The existing simulation system does not take into account the internal parameters of the algorithm, resulting in incomplete evaluation parameters. Summary of the invention

[0004] In view of this, it is necessary to provide a vehicle autonomous driving simulation test system to solve the technical problems in the prior art that the evaluation parameters of the autonomous driving simulation test system are incomplete and the vehicle dynamics model is too simple, resulting in insufficient model credibility.

[0005] In order to solve the above problems, the present invention provides a vehicle automatic driving simulation test system, comprising: a middleware module, a scenario simulation component and an automatic driving algorithm module deployed in a first operating system environment, and a vehicle dynamics component and an automated evaluation component deployed in a second operating system environment, wherein the middleware module is used to connect the communication between the components deployed in the first operating system environment and the second operating system environment; Scenario simulation component, used to simulate traffic scenarios and output target vehicle parameters; An autonomous driving algorithm module, which is used to output control instructions and internal algorithm parameters based on target vehicle parameters; A vehicle dynamics component, used to perform vehicle dynamics response based on control commands and calculate dynamic parameters of the target vehicle; An automated evaluation component used to test the vehicle's autonomous driving algorithm based on internal algorithm parameters and the vehicle's dynamic parameters.

[0006] In a possible implementation, the vehicle dynamics component is determined by parameter configuration based on vehicle information acquired from an actual vehicle in a real driving environment.

[0007] In a possible implementation, the process of configuring parameters of the vehicle dynamics component by acquiring vehicle information of an actual vehicle in a real driving environment includes: Optimizing the driving performance of vehicle dynamics components based on the average torque error percentage; Optimize the braking performance of vehicle dynamics components based on braking distance error; The steering function of the vehicle dynamics component is optimized based on the steering wheel angle lag time and the average angle error.

[0008] In one possible implementation, the vehicle dynamics component includes a first communication channel and a second communication channel; The first communication channel is connected to the vehicle dynamics component and the scene simulation component, and is used to transmit the dynamic parameters of the vehicle dynamics component to the scene simulation component; The second communication channel connects the middleware module and the vehicle dynamics component, and is used to transmit the control instructions sent by the automatic driving algorithm module to the vehicle dynamics component through the middleware.

[0009] In a possible implementation, both the first communication channel and the second communication channel use the TCP communication protocol to implement communication interaction.

[0010] In a possible implementation, the autonomous driving algorithm module includes a data acquisition unit, a path planning unit, and a data transmission unit; The acquisition unit is used to acquire the dynamic parameters sent by the vehicle dynamics component and the target vehicle parameters sent by the scene simulation component; and send the dynamic parameters and the target vehicle parameters to the path planning unit; The path planning unit is used to perform driving strategy and path planning for the target vehicle according to the dynamic parameters, the target vehicle parameters and the internal parameters of the algorithm, convert the driving strategy and path planning into control instructions for the target vehicle, and send the control instructions to the data transmission unit; The data transmission unit is used to send the control instruction to the middleware module.

[0011] In a possible implementation, the middleware module and the data transmission unit communicate with each other through the ROS communication protocol; and the middleware module and the scene simulation component communicate with each other through the TCP communication protocol.

[0012] In a possible implementation, a communication component is further included, and the communication component is used to receive the internal algorithm parameters sent by the middleware module and forward the internal algorithm parameters to the automatic evaluation component.

[0013] In a possible implementation, the automation evaluation component is connected to the vehicle dynamics component and the scenario simulation component by host IP and port adaptation; and is connected to the autonomous driving algorithm module and the middleware module by remote SSH protocol; The automated evaluation component includes a first start-up unit and a second start-up unit; The first starting unit is used to start the autonomous driving algorithm module and the middleware module in sequence based on the SSH protocol; The second startup unit is used to start the scene simulation component and the vehicle dynamics component in sequence based on the host IP and port.

[0014] In a possible implementation, the automated evaluation component performs parameter configuration based on a full set of evaluation parameters.

[0015] The beneficial effects of the present invention are as follows: the scene simulation component simulates the traffic scene and outputs the target vehicle parameters, and the target vehicle parameters and the control instructions of the automatic driving algorithm module to be tested are synchronously transmitted to the dedicated vehicle dynamics component through the middleware module to simulate and test the vehicle performance, wherein the dedicated vehicle dynamics component performs parameter configuration according to the actual vehicle data, and can achieve the accuracy of the function and performance test of the automatic driving planning and control algorithm. On the other hand, the communication between the various components in the first operating system environment and the second operating system environment is carried out through the middleware module, which ensures the synchronization of vehicle signals and sensor signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of the structure of an embodiment of a vehicle automatic driving simulation test system provided by the present invention; Figure 2 A schematic structural diagram of an embodiment of an autonomous driving algorithm module in the vehicle autonomous driving simulation test system provided by the present invention. DETAILED DESCRIPTION

[0017] The preferred embodiments of the present invention are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not used to limit the scope of the present invention.

[0018] A specific embodiment of the present invention discloses a vehicle automatic driving simulation test system, see Figure 1 , including: a middleware module, a scenario simulation component and an autonomous driving algorithm module deployed in a first operating system environment, a vehicle dynamics component and an automated evaluation component deployed in a second operating system environment, wherein the middleware module is used to connect the communications between the components deployed in the first operating system environment and the second operating system environment; Scenario simulation component, used to simulate traffic scenarios and output target vehicle parameters; An autonomous driving algorithm module, which is used to output control instructions and internal algorithm parameters based on target vehicle parameters; A vehicle dynamics component, used to perform vehicle dynamics response based on control commands and calculate dynamic parameters of the target vehicle; An automated evaluation component used to test the vehicle's autonomous driving algorithm based on internal algorithm parameters and the vehicle's dynamic parameters.

[0019] In this embodiment, the scene simulation component simulates the traffic scene and outputs the target vehicle parameters, and transmits the target vehicle parameters and the control instructions of the automatic driving algorithm module to be tested to the dedicated vehicle dynamics component through the middleware module to simulate and test the vehicle performance, wherein the dedicated vehicle dynamics component performs parameter configuration according to the actual vehicle data, and can achieve the accuracy of the function and performance test of the automatic driving planning and control algorithm. On the other hand, the communication between the various components in the first operating system environment and the second operating system environment is carried out through the middleware module, which ensures the synchronization of vehicle signals and sensor signals.

[0020] It should be noted that, in the implementation mode provided by the present invention, the first operating system environment can be a Linux operating system environment, and the second operating system environment can be a Windows operating system environment. The scene simulation component can be based on PreScan, can be based on Carsim, or can be based on Simulink. PreScan and Carsim are relatively mature commercial simulation software. If the simulation environment generation module is based on PreScan, the data conversion interface includes a PreScan-Matlab interface and a Matlab-Ros interface. The PreScan-Matlab interface is an interface for interaction between PreScan and Matlab, and the Matlab-Ros interface is an interface for interaction between Matlab and Ros. The interface is an application programming interface. Information cannot be directly transmitted between the PreScan system and the Ros system, but both the PreScan system and the Ros system can be docked with Matlab, so it can be considered to use Matlab as an information transfer station between them, that is, the PreScan-Matlab interface and the Matlab-Ros interface can be used to transmit information between PreScan and Ros. The PreScan-Matlab interface and the Matlab-Ros interface can be implemented using methods known to those skilled in the art (eg, programming methods), which will not be elaborated herein.

[0021] In some embodiments, the vehicle dynamics component includes a first communication channel and a second communication channel; The first communication channel is connected to the vehicle dynamics component and the scene simulation component, and is used to transmit the dynamic parameters of the vehicle dynamics component to the scene simulation component; The second communication channel connects the middleware module and the vehicle dynamics component, and is used to transmit the control instructions sent by the automatic driving algorithm module to the vehicle dynamics component through the middleware.

[0022] In this embodiment, the vehicle dynamics parameters include dynamic information of the vehicle, such as speed, acceleration, direction, etc.

[0023] In a specific embodiment, the first communication channel and the second communication channel both use the TCP communication protocol to implement signal interaction. Through TCP, the scenario simulation component and the vehicle dynamics component can exchange data between different computers or processes to ensure accurate information transmission. And through the network protocol, the scenario simulation and vehicle dynamics modules can be distributed on different hardware to improve the flexibility and scalability of the system.

[0024] It should be noted that the scene simulation component and the vehicle dynamics component also transmit information such as vehicle posture and tire electric shock to each other through TCP communication. By continuously transmitting information such as vehicle posture and tire electric shock, it is ensured that the simulation system can accurately reflect the performance of the vehicle in the real environment and provide more realistic test results.

[0025] In order to ensure that the vehicle dynamics component can have more realistic and richer vehicle information and improve the credibility of the simulation model, in this embodiment, parameter configuration is performed by obtaining vehicle information of the actual vehicle in a real driving environment.

[0026] Specifically, the vehicle dynamics model parameters are set through the dynamics software Trucksim. The vehicle dynamics model includes 9 modules: sprung mass, structural dimensions, aerodynamics, tires, powertrain, tractor and trailer load ratio, suspension system, braking system, and steering system. The dynamics model is configured in detail according to the configuration data of the actual vehicle and optimized through the following methods to achieve a highly reliable vehicle dynamics model. After the setting is completed, the dynamics model is converted into a simulink model.

[0027] The specific method is as follows: First, the driving performance is optimized based on the average torque error percentage. The test data of the targetless vehicle cruising (flat road and uphill) in the automatic driving mode is collected through simulation. The test time is at least 1 minute (torque greater than 0 is valid data), of which the uphill ratio is not less than 30% (can be obtained in segments). Extract the engine torque T. Judge by comparing the required torque T0 (planned output) and the actual output torque T1. There are a total of N sampling points, the i-th torque error percentage is Pi=(T0i-T1i) / T0i, and the average torque error percentage is:

[0028] When △P<10%, the driving performance meets the requirements. If it is greater than 10%, the powertrain MAP and transmission ratio are checked for optimization and modification.

[0029] Then, the braking performance is optimized based on the braking distance error: test data in the autonomous driving mode is collected through simulation and real vehicle (the scene is a stationary target 200m ahead), and it is performed 3 times respectively. The final average stopping distance Dr and Ds of the real vehicle and simulation test are obtained. The two distance differences △D are compared. When △D is less than 0.5m, it meets the requirements. Otherwise, the optimization and modification are performed by modifying the braking curve MAP diagram and the pipeline transmission hysteresis parameters.

[0030] Finally, the steering function is optimized based on the steering wheel angle lag time and the average angle error: the test data of the S-curve (curve radius 500m~800m) in the autonomous driving mode is collected through simulation, and the test time is at least 1 minute. Extract the steering wheel angle S. Make a judgment by comparing the required steering wheel angle S0 (planning output) and the actual steering wheel angle S1 curve. Compare the peaks or troughs of the two curves to obtain the lag time t. Through N sampling points, calculate the average steering wheel angle error as

[0031] When t<0.2s and △S<5︒, the steering performance meets the requirements. Otherwise, the steering gear ratio, kingpin inclination angle, caster angle, and lateral offset at the axis center are optimized and modified.

[0032] In some embodiments, see Figure 2 , the autonomous driving algorithm module data acquisition unit, path planning unit and data transmission unit; The acquisition unit is used to acquire the dynamic parameters sent by the vehicle dynamics component and the target vehicle parameters sent by the scene simulation component; and send the dynamic parameters and the target vehicle parameters to the path planning unit; The path planning unit is used to perform driving strategy and path planning for the target vehicle according to the dynamic parameters, the target vehicle parameters and the internal parameters of the algorithm, convert the driving strategy and path planning into control instructions for the target vehicle, and send the control instructions to the data transmission unit; The data transmission unit is used to send the control instruction to the middleware module.

[0033] In this embodiment, the autonomous driving algorithm (test object) receives target object parameters and vehicle dynamics parameters (issued by the vehicle dynamics component), and outputs throttle, steering, and braking instructions to the vehicle dynamics component after calculation by the fusion, planning, and control modules. Based on the acquisition of full parameters by the autonomous driving algorithm and the vehicle dynamics component, a more comprehensive evaluation capability is achieved with the middleware-communication component-automated evaluation component as the transmission link.

[0034] In some embodiments, the middleware module is a bridge between the various components of the SIL system, and mainly implements functions such as the reception and transmission of input and output data between the scene simulation component and the autonomous driving algorithm, data analysis and construction. Specifically, the middleware module communicates with the autonomous driving algorithm module via ROS. The middleware module also analyzes and classifies the control instructions fed back by the autonomous driving algorithm module and transmits them to the vehicle dynamics component. By independently designing the middleware and connecting the dual system components, the synchronous connection of each signal data stream and the connection of the evaluation data stream are realized. Automated simulation testing is realized.

[0035] Specifically, the middleware module and the autonomous driving algorithm module implement signal interaction through the ROS communication protocol; the middleware module and the scene simulation component implement signal interaction through the TCP communication protocol.

[0036] In some embodiments, a communication component is further included, wherein the communication component is used to receive the algorithm internal parameters sent by the middleware module and forward the algorithm internal parameters to the automatic evaluation component.

[0037] In this embodiment, the communication component and the middleware module use the TCP communication protocol to implement signal interaction.

[0038] In some embodiments, the automated evaluation component is used to send instructions to start various components in the first operating system environment in a preset order, including: The automated evaluation component is connected to the vehicle dynamics component and the scenario simulation component in communication via host IP and port adaptation; and Communicate with the autonomous driving algorithm module and middleware module through the remote SSH protocol; The automated evaluation component includes a first start-up unit and a second start-up unit; The first starting unit is used to start the autonomous driving algorithm module and the middleware module in sequence based on the SSH protocol; The second startup unit is used to start the scene simulation component and the vehicle dynamics component in sequence based on the host IP and port.

[0039] In this embodiment, the automated evaluation component runs in a Windows environment and connects to different local and remote components (including vehicle dynamics, communication, and scenario simulation) through a network. It also uses the SSH protocol to remotely manage and control the operation of algorithms and middleware modules related to autonomous driving. The purpose of this is to test and evaluate the performance of the autonomous driving system in various environments.

[0040] It should be noted that when configuring the parameters of the automation evaluation component, the method of full evaluation parameters is adopted. Specifically, the full evaluation parameters include environmental parameters (target objects, lane lines, etc.), vehicle parameters (vehicle speed, acceleration, yaw rate, etc.) and algorithm internal parameters (following state, lane change state, etc.). Environmental parameters and algorithm internal parameters are included in the autonomous driving algorithm, and vehicle parameters are included in the vehicle dynamics model component. Therefore, the specific configuration method of the evaluation parameters is as follows: Configure the network port and first configure the middleware to read the target objects, lane lines and other environmental parameters and algorithm internal parameters in the autonomous driving algorithm through the ROS protocol.

[0041] Open another network port and configure the middleware to send the above information to the communication component based on the Simulink environment through the TCP protocol.

[0042] The communication component obtains vehicle parameters from the vehicle dynamics component through simulink. In this way, the communication component contains all the parameters to be evaluated. The automatic evaluation component directly reads the communication component parameters to obtain the full evaluation data. And the threshold value of the parameters is judged by the performance indicators.

[0043] Configure automated evaluation components: The automated evaluation software ECUtest runs in a Windows environment. By configuring the host IP address and port, it connects to the vehicle dynamics component of Windows and the scenario simulation component in the remote Linux environment, and controls the start and stop of the autonomous driving algorithm and middleware module through the remote SSH protocol. The startup sequence of the automated evaluation components is: autonomous driving algorithm -> middleware -> scenario simulation component -> vehicle dynamics component.

[0044] After the configuration is complete, run the automated evaluation component to complete the SIL simulation test, implement the function and performance test of the autonomous driving algorithm, and conduct a comprehensive evaluation.

[0045] Furthermore, by building a simulation scenario through the scenario software VTD, and configuring sensor models such as lane line sensors, front radars, and corner radars, the lane line information (the deviation distance between the vehicle and the lane line, the heading angle between the vehicle and the lane line, the lane curvature, the lane curvature change rate, etc.) and the information of surrounding target vehicles (relative distance, speed, acceleration, etc.) are sent through the port.

[0046] Based on this, the vehicle automatic driving simulation test system provided by the embodiment of the present invention has the following characteristics: The included proprietary vehicle dynamics components ensure the credibility of the simulation system; the vehicle dynamics components configure the parameters of the dynamics model in detail based on the real vehicle data, so that it can truly reflect the characteristics and driving status of the real vehicle. The credibility of the simulation model is ensured, and the function and performance testing of the autonomous driving planning and control algorithm can be realized.

[0047] The configuration methods adopted are the driving performance configuration method based on the average torque error percentage, the braking performance configuration method based on the average braking distance error, and the steering performance configuration method based on the steering wheel angle lag time and the average angle error.

[0048] The TCP-based dual-system communication mechanism ensures the synchronization of simulation signals (vehicle driving signals and obstacle and lane line signals): the vehicle driving signals of the vehicle dynamics component of the present invention (required by the control module) are first sent to the scene simulation component via TCP, and then sent to the middleware via the scene simulation component, thereby ensuring the synchronization of vehicle signals and sensor signals, and further ensuring the credibility of the SIL simulation system.

[0049] The simulation system has comprehensive evaluation capabilities: it obtains full parameters based on the autonomous driving algorithm and vehicle dynamics components, and achieves more comprehensive evaluation capabilities with middleware-communication components-automated evaluation components as the transmission link.

[0050] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by any technician familiar with the technical field within the technical scope disclosed by the present invention should be covered within the protection scope of the present invention.

Claims

1. A vehicle automatic driving simulation test system, characterized in that: include: A middleware module, a scenario simulation component, and an autonomous driving algorithm module deployed in a first operating system environment, and a vehicle dynamics component and an automated evaluation component deployed in a second operating system environment, wherein the middleware module is used to connect the communications between the components deployed in the first operating system environment and the second operating system environment; Scenario simulation component, used to simulate traffic scenarios and output target vehicle parameters; An autonomous driving algorithm module, which is used to output control instructions and internal algorithm parameters based on target vehicle parameters; A vehicle dynamics component, used to perform vehicle dynamics response based on control commands and calculate dynamic parameters of the target vehicle; An automated evaluation component used to test the vehicle's autonomous driving algorithm based on internal algorithm parameters and the vehicle's dynamic parameters.

2. The vehicle automatic driving simulation test system according to claim 1, characterized in that: The vehicle dynamics component is determined by parameter configuration through vehicle information acquired from an actual vehicle in a real driving environment.

3. The vehicle automatic driving simulation test system according to claim 2, characterized in that: The process of configuring parameters of the vehicle dynamics component by acquiring vehicle information of an actual vehicle in a real driving environment includes: Optimizing the driving performance of vehicle dynamics components based on the average torque error percentage; Optimize the braking performance of vehicle dynamics components based on braking distance error; The steering function of the vehicle dynamics component is optimized based on the steering wheel angle lag time and the average angle error.

4. The vehicle automatic driving simulation test system according to claim 1, characterized in that: The vehicle dynamics component includes a first communication channel and a second communication channel; The first communication channel is connected to the vehicle dynamics component and the scene simulation component, and is used to transmit the dynamic parameters of the vehicle dynamics component to the scene simulation component; The second communication channel connects the middleware module and the vehicle dynamics component, and is used to transmit the control instructions sent by the automatic driving algorithm module to the vehicle dynamics component through the middleware.

5. The vehicle automatic driving simulation test system according to claim 4, characterized in that: The first communication channel and the second communication channel both use the TCP communication protocol to achieve communication interaction.

6. The vehicle automatic driving simulation test system according to claim 1, characterized in that: The autonomous driving algorithm module includes a data acquisition unit, a path planning unit and a data transmission unit; The acquisition unit is used to acquire the dynamic parameters sent by the vehicle dynamics component and to acquire the target vehicle parameters sent by the scene simulation component; and sending the dynamic parameters and target vehicle parameters to the path planning unit; The path planning unit is used to perform driving strategy and path planning for the target vehicle according to the dynamic parameters, the target vehicle parameters and the internal parameters of the algorithm, convert the driving strategy and path planning into control instructions for the target vehicle, and send the control instructions to the data transmission unit; The data transmission unit is used to send the control instruction to the middleware module.

7. The vehicle automatic driving simulation test system according to claim 6, characterized in that: The middleware module and the data transmission unit achieve communication interaction through the ROS communication protocol; the middleware module and the scene simulation component achieve communication interaction through the TCP communication protocol.

8. The vehicle automatic driving simulation test system according to claim 1, characterized in that: It also includes a communication component, which is used to receive the algorithm internal parameters sent by the middleware module and forward the algorithm internal parameters to the automatic evaluation component.

9. The vehicle automatic driving simulation test system according to claim 8, characterized in that: The automated evaluation component is connected to the vehicle dynamics component and the scenario simulation component by host IP and port adaptation; and is connected to the automated driving algorithm module and the middleware module by remote SSH protocol; The automated evaluation component includes a first start-up unit and a second start-up unit; The first starting unit is used to start the autonomous driving algorithm module and the middleware module in sequence based on the SSH protocol; The second startup unit is used to start the scene simulation component and the vehicle dynamics component in sequence based on the host IP and port.

10. The vehicle automatic driving simulation test system according to claim 1, characterized in that: The automated evaluation component performs parameter configuration based on a full set of evaluation parameters.

Images