User-Centric Multi-Objective Predictive Control for Mixed Vehicular Platoon

Le résumé fourni par la source

The penetration rate of automated vehicles (AVs) will remain unsaturated for a long period, leading to the coexistence of AVs and human-driven vehicles (HDVs), namely, mixed vehicular platoon (MVP). This paper proposes a novel user-centric multi-objective predictive control (UMPC) strategy to address dynamics uncertainty and multi-objective conflict for MVP. A data-driven model is established with subspace identification to alleviate the adverse effects of the non-ideal driving behavior and complex powertrain structure of electric vehicles. To provide a personalized driving experience, soft constraints and a user-centric multi-objective cost function are formulated. Based on this, a multi-objective optimization scheme in terms of data-driven model predictive sequence is designed. It aims to mitigate conflicts among multiple optimization objectives involving driving safety, driving comfort and energy economy. Then, a grey wolf optimizer (GWO) is devised to navigate the search process, striving for globally optimal trade-offs among conflicting objectives. With the above preparations, a UMPC strategy is suggested. Then, a hardware-in-the-loop experiment platform with CarMaker software and driving simulator is constructed, and twenty drivers participate in the experiment. The experimental results demonstrate the effectiveness of the proposed UMPC strategy.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.