Adaptive multi-objective model predictive trajectory tracking control of 4WID-4WIS AEVs

Abstract

Distributed-drive autonomous electric vehicles have significant potential to enhance vehicle safety and efficiency through their superior control flexibility. However, their over-actuation inevitably increases control complexity, making it challenging to balance high-precision tracking, handling stability, and energy efficiency in critical scenarios. To address this issue, a novel multi-goal control architecture utilizing Adaptive Energy-Management Model Predictive Control (AEMPC) is presented. Furthermore, a fuzzy logic controller is integrated to adaptively adjust the prediction horizon based on the current vehicle motion state, thereby enhancing real-time performance. Simulation results indicate that, compared with nominal MPC, this method effectively eliminates torque fluctuations during full-throttle acceleration and reduces energy consumption by 9.03%. During the double lane-change (DLC) manoeuvre, the path-following deviation in the lateral direction is reduced by 34.67%, while a higher stability margin is achieved. The results demonstrate that the proposed approach exhibits outstanding multi-objective control performance and adaptability across challenging scenarios.