无人驾驶越野车辆纵向速度跟踪控制试验

Experiment on Longitudinal Speed Tracking Control for Unmanned Off-road Vehicles

以无人驾驶轻型战术轮式越野车辆为平台,开展模型预测纵向速度跟踪控制实车试验研究。针对平台控制特性设计合适的下位控制器,使用Matlab/Simulink与包含气压制动系统的TruckSim车辆联合仿真初步测试系统可行性,并在沥青路和土路分别进行实车试验。试验结果表明:模型预测速度跟踪控制系统能够克服气压制动延时长、整车质量重、越野路况行驶阻力波动大等模型误差和不确定干扰,自适应调节期望加速度大小,实现不同行驶工况高精度速度跟踪。试验过程驱动/制动切换平稳、无振荡,且能够像熟练驾驶员一样充分利用发动机辅助制动,必要时既不施加电控制动,也不请求发动机输出转矩。系统使用现代车辆易于获得的车辆状态参数,便于向其他车辆移植,可作为无人车辆车体控制得力技术加以推广。

An experimental research on model predictive longitudinal speed tracking control is implemented based on a test platform of unmanned light tactical wheeled off-road vehicle. A lower level controller is designed according to the control characteristics of the platform. The feasibility of the system is first verified by using co-simulation of Matlab/Simulink and TruckSim vehicle with air brake system, and then vehicle tests are carried out on the asphalt road and dirt road, respectively. The results show that the system robustly model error and uncertain disturbances, such as long time delay of air brake system, the weight of the off-road vehicle, the fluctuation of resistance on the off-road condition, and adjust the desired acceleration adaptive. High-precision speed tracking control for various driving conditions is achieved. The drive control and brake control switch smoothly without oscillation during the test process, and the system can make full use of engine braking as a skilled driver, neither the electronic control brake nor the engine output torque is requested when necessary. The system uses states and parameters that can be easily acquired in modern vehicles, so the approach can be easily applied to other vehicles, and used as a powerful tool for unmanned ground vehicles control.