基于数字孪生的智能驾驶汽车整车在环测试系统研究

Intelligent driving vehicle in-the-loop test system based on digital twin

数字孪生(Digital Twins,DT)是借助实时仿真技术将真实物体映射到虚拟世界的系统,应用数字孪生技术可实现虚拟道路场景下真实智能汽车的自动驾驶测试。设计了一套基于数字孪生的智能驾驶汽车整车在环测试系统,并将其应用于AEB系统的开发验证。在数字孪生技术框架下结合仿真测试工具、V2X通信设备、真实测试车辆等功能单元,形成虚拟环境与真实环境相互映射的整车在环测试系统。基于虚幻引擎(Unreal Engine)4构建真实驾驶环境的镜像仿真场景,利用V2X通信技术将实车状态数据实时传输到虚拟AirSim车辆控制器,同时向被测车辆反馈仿真场景环境感知数据。考虑测试系统信号延时的影响,通过多种工况的整车在环测试对AEB控制算法进行优化改进。通过实车试验验证了基于数字孪生的智能汽车测试方法的有效性与优越性,结果表明该测试方法既保留了实车测试的真实性,也可以有效避免实车测试的危险性。

Digital Twin(DT)is a system that maps real objects to the virtual world with the help of real-time simulation technology.The DT technology can realize the automatic driving test of real intelligent vehicles in virtual scenarios.A vehicle-in-the-loop test system for intelligent driving vehicles based on DT was designed,and applied to the development and verification of AEB system.Under the framework of DT technology,by combining functional units such as simulation tool,V2X communication equipment,and test vehicle,a vehicle-in-the-loop test system in which the virtual and real environments were mapped to each other was formed.Based on Unreal Engine 4,a mirror simulation scene of the real driving environment was constructed.Using V2X communication technology,real vehicle states were transmitted to the virtual AirSim vehicle controller in real time,and the environment perception data of the simulation scene was fed back to the real vehicle.Considering the influence of the signal delay of the test system,the AEB control algorithm was optimized and improved through vehicle-in-the-loop tests under various conditions.The validity and superiority of DT-based testing method for intelligent vehicle was verified.The results showed that the test method not only retained the authenticity of the real vehicle test,but also could effectively avoid the danger of the real vehicle test.