无人驾驶汽车横向运动LQR控制器设计及工况测试

Design and test of LQR controller for lateral movement of unmanned vehicle

为提高无人驾驶汽车横向运动效果,发明了一种能够自主调节加权矩阵的预瞄策略与线性最优控制(LQR)优化调节器,加入前馈控制方法来调节智能汽车横向运动参数。通过非预瞄方式构建得到路径跟踪误差的力学仿真模型,构建了横向前馈以及LQR反馈控制器,并采用仿真方法比较带前馈控制状态下LQR优化控制器。双移线工况测试结果表明:包含前馈控制的LQR优化控制器可以对0.3 m以内的路径进行跟踪距离偏差,并使航向偏差不超过0.1 rad,且仅需要调节较少次数。连续换道工况测试结果表明:设置前馈控制调节的LQR控制器可以获得较小的距离偏差与航向偏差,并更加精确跟踪目标路径,达到了理想控制状态。该研究显著提高了控制精度,且增强控制器自适应能力,为提高汽车运行稳定性提供了良好理论支撑。

In order to improve the lateral motion effect of unmanned vehicle,a linear quadratic regulator(LQR)optimization regulator which can adjust the weighting matrix autonomously was developed,and a feedforward control method was added to adjust the lateral motion parameters of intelligent vehicle.The mechanical simulation model of path tracking error was constructed by nonpreviewing mode,and the lateral feedforward and LQR feedback controllers were constructed.The SIMULATION method was used to compare the LQR optimization controllers with feedforward control.The results show that the LQR optimization controller with feedforward control can track the distance deviation within 0.3 m,and make the heading deviation less than 0.1 rad,and need less adjustment.The test results of continuous lane change conditions show that the optimized LQR controller with feedforward control adjustment can obtain lower range and heading deviation,and track the target path more accurately,and achieve the ideal control state.This research can significantly improve the control accuracy and enhance the adaptive ability of the controller,which has a good theoretical support value for improving the stability of vehicle operation.