A Fast-Computing Path Tracking Control Strategy for Autonomous Multi-Axle Electric Vehicle Considering Safety and Stability

This paper provides a fast-computing path tracking control strategy for an autonomous multi-axle electric vehicle considering safety and stability. Firstly, a high-fidelity nonlinear vehicle dynamic control model is proposed for the control prediction and state updating, which consists of a path tracking error model, a double track vehicle dynamic model as well as a nonlinear tire model. To reflect the dynamic properties accurately and improve the safety and stability comprehensively, a nonlinear model predictive control (NMPC) problem is formulated and applied to the path tracking control, in which multiple dynamic constraints are designed for various safety and stability demands, including the phase plane relationship, the wheel anti-slip, the body rollover prevention and tire wear alleviation. Meanwhile, to further guarantee the closed-loop control stability, a barrier Lyapunov function (BLF) is designed and the auxiliary control rule is determined in the form of the initial constraint. Moreover, to enhance the computation efficiency and solve the nonlinear control effectively, a multiconstrained fast NMPC algorithm is adopted, which combines the projection gradient based method and the augmented Lagrange method based on the first-order optimality conditions and mitigates the online calculation burden. Finally, the simulation and HIL test is carried out to verify the control performance, which yields that the proposed strategy can effectively track the target path, improve the safety and stability and realize fast calculation compared with the traditional method.