A Decentralized Cooperative Control Framework for Active Steering and Active Suspension: Multi-Agent Approach
Jinhao Liang, Yanbo Lu, Dawei Pi, Guodong Yin, Weichao Zhuang, Faan Wang, Jiwei Feng, Chaobin Zhou
Abstract
A modular electric vehicle (EV) platform enables ability to integrate different control systems in a flexible way. In this article, a decentralized cooperative control framework is proposed to achieve the integration of active front steering (AFS) system and active-suspension system (ASS) by applying a multi-constrained distributed model predictive control (MDMPC) approach, which aims to improve the vehicle lateral stability, ride comfort, and roll safety during path tracking. First, a partly decoupled 6-degree-of-freedom (DoF) half-vehicle dynamics model is constructed. Then, a multi-agent-system (MAS)-based framework is introduced to coordinate AFS and ASS, which allows for the multi-constraints within the information exchange between agents. Through minimizing linear convex combination of objective functions, the cooperative control strategy is ultimately solved by the Pareto-optimality theory. Moreover, vehicle lateral and roll motion stability region described by the phase plane is employed to bound the controllable limits and achieve the dynamic cooperation between AFS and ASS. The simulation and hardware-in-the-loop (HIL) test results show that the proposed framework is effective for coordinating AFS and ASS, thereby enhancing the vehicle lateral and vertical stability during path tracking.