Abstract: 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.

Key words: self-driving, trajectory tracking, coordinate compensation, full format data driven control, integral sliding mode control (SMC), integral windup compensation