Multi-agent deep reinforcement learning based real-time planning approach for responsive customized bus routes

Binglin Wu, Xingquan Zuo, Gang Chen, Guanqun Ai, Xing Wan

Abstract: Highlights•This paper studies an emerging multiple routes planning problem for customized bus.•An innovative deep reinforcement learning based approach (MRL-RP) is proposed.•A new multi-agent system powered by encoder–decoder neural network is developed.•Empirical results show that MRL-RP outperforms some state-of-the-art algorithms.