无模型坐标补偿积分滑模约束的自动驾驶汽车轨迹跟踪控制

Trajectory Tracking Control Based on Model-Free Coordinate Compensation Integral Sliding Mode Constraints

为了解决四轮自动驾驶汽车轨迹跟踪的复杂控制问题,提出一种新的全格式无模型自适应坐标补偿积分滑模约束控制方案。控制方案只需要自动驾驶车辆轨迹跟踪的I/O数据,不涉及车辆模型信息,即前轮转角输入数据和车辆横摆角输出数据。因此,该方案对于不同车型均能实现轨迹跟踪控制。在轨迹跟踪过程中只针对车辆横摆角进行控制易造成跟踪轨迹偏差,本文在全格式无模型自适应控制的基础上加入坐标补偿算法;为了提高系统在运行过程中的鲁棒性,加入积分滑模控制;为了应对在滑模控制过程中系统运行在饱和区域,设计了动态补偿,使轨迹跟踪系统运行在滑模控制的线性区域。最后,对无模型自适应积分滑模约束控制方案、原型无模型自适应控制算法和PID算法进行仿真比较。仿真结果表明,所提算法比传统控制方法具有更好的控制性能,跟踪波动误差在0.09%以内,稳定时间在0.5 s以内,跟踪曲线平滑。

In order to solve the complex control problem of the trajectory tracking of four-wheel autonomous vehicles,a new full model-free adaptive coordinate compensation integral sliding mode constraint control scheme is proposed.The control scheme only needs the I/O data of the trajectory tracking of the autonomous driving vehicle,and does not involve the vehicle model information,that is,the input data of the front wheel angle and the output data of the body angle.Therefore,this new scheme can realize trajectory tracking control for the trajectory tracking systems of different vehicle models.In the process of trajectory tracking,only controlling the body angle is easy to cause the deviation of the tracking trajectory.In this paper,firstly,a coordinate compensation algorithm is added on the basis of the full-format model-free adaptive control.Secondly,in order to improve the robustness of the system during operation,an integral sliding mode control is added.Thirdly,to deal with the system running in the saturation region in the process of sliding mode control,dynamic compensation is designed to make the trajectory tracking system run in the linear region of sliding mode control.Finally,a simulation comparison among the model-free adaptive integral sliding mode constraint control scheme,the prototype model-free adaptive control algorithm and the PID algorithm is carried out.The results show that the proposed algorithm has better control performance than traditional control methods.The tracking fluctuation error is within 0.09%.The stabilization time is within 0.5 s,and the tracking curve is smooth.