Path following control for 4WID-EV based on extended state observer and sliding mode control considering yaw stability

Abstract

Automatic driving technology has developed over time and the path-following problem for four-wheel independent drive electric autonomous vehicles (4WID-EV) is a topic of increasing interest. During the path-following process, a consideration of uncertainties in vehicle modeling including parameter variation, modeling error, and external disturbances is made, which have an important effect on path-tracking performance. As so, we proposed a new path following control strategy that can be applied to robotics applications. First, it was determined to establish a 2 DOF dynamic model and path following error model, and were then converted to yaw angle tracking problems. Therefore, the desired control input of steering angle was obtained via yaw angle tracking controller. To estimate and compensate the uncertainties associated with systems, the nonlinear extended state observer (NESO) was designed while the NTSM controller was used to realize yaw angle tracking. Next, considering the system stability, a sliding mode controller with modified reaching law was used to obtain the desired additional yaw moment. Then, the optimized torque distributor was designed for allocating optimal tire forces. Finally, the simulation experiments were carried out in Matlab/Simulink and Carsim joint simulation platform. In conclusion, the proposed control scheme exhibits highly accurate path following as well as strong robustness to speed, tire-road friction coefficient and external disturbance.