Adaptive model predictive control of distributed drive electric vehicles based on improved sum of square programming estimation of stability domain

Abstract

For the stability control demands of distributed-drive electric vehicles under different operating conditions, this paper proposes a hierarchical control architecture based on improved sum-of-squares-programming (SOSP). Firstly, a nonlinear vehicle mathematical model based rational function tire model is established, providing an application foundation for the stability domain estimation and stability control. Then, the hierarchical control architecture is analyzed in detail. In the upper-level control, considering the challenges such as insufficient accuracy and high computational complexity of the traditional stability domain estimation methods, this paper presents a novel stability domain construction method based on improved sum-of-squares programming (ISOSP), which achieves precise quantification of the stability domain. Based on the stability domain estimation results, an adaptive model predictive control (AMPC) strategy is designed to achieve stability control at the lower-level control. Finally, the effectiveness of the proposed hierarchical control architecture is validated in the co-simulation platform of Matlab/Simulink and CarSim. The simulation results show that compared to SOSP estimation method for stability domain, ISOSP reduces the computation time by 51.44%–83.31% without compromising conservatism, and AMPC controller designed based on ISOSP ensures that the yaw rate error remains within 0.04 rad/s, and the side slip angle does not exceed 0.017 rad under extreme operating conditions.