Co-optimization of vehicle queue speed planning and energy management with reinforcement learning
In recent years,with the rapid development of intelligent transportation system,including vehicle-vehicle communica-tion,vehicle-road communication and other short-range real-time wireless communication information,as well as road traffic infor-mation and other long-distance traffic information,so that the vehicle is able to obtain real-time knowledge of the surrounding ve-hicle movement as well as the traffic environment in front of it,which is conducive to improving the vehicle's perception of the surrounding traffic environment in order to achieve reasonable travel arrangements and driving control,thus improving the vehicle performance.In order to realize the energy-saving driving of fleet in multi-signal light scenarios,a co-optimization method based on reinforcement learning for vehicle queuq speed planning and energy management was proposed.Through the SUMO platform,a scenario including five vehicles for a fleet of vehicles passing through multiple signals was established.The results show that the proposed method outperforms the traditional driving model Intelligent Driver Model(IDM)in terms of comfort,economy and effi-ciency.
万方数据
维普
