能源管理
模型预测控制
汽车工程
能源消耗
燃料效率
高效能源利用
工程类
交叉口(航空)
软件部署
控制(管理)
能量(信号处理)
计算机科学
实时计算
模拟
运输工程
人工智能
统计
软件工程
电气工程
数学
作者
Mei Yan,Guotong Li,Menglin Li,Haibo He,Hongyang Xu,Haoran Liu
标识
DOI:10.1016/j.enconman.2022.115397
摘要
This research aims to answer the question of how to manage the energy flow of fuel cell buses (FCBs) more efficiently and intelligently with the deployment of Internet of vehicles (IoV) technology. Energy management strategies on the IoV environment need to comprehensively utilize vehicles state information and traffic state information. Given that, this paper proposes a hierarchical predictive energy management strategy (HPEMS) for FCBs with launch control integrating traffic information to reduce bus lines' energy consumption and improve the powertrain's energy efficiency. In the upper level, the launch control based on deep reinforcement learning (DRL) selects the appropriate start time based on the traffic states and the vehicle states, reducing the energy consumption and traveling time caused by frequent starting and stopping through the traffic light intersection. In the lower level, model predictive control (MPC) based predictive energy management is performed to achieve efficient and reasonable power splitting of batteries and fuel cells. The results show a significant improvement for FCB in HPEMS with launch control. The average travel time, idle time, waiting time for traffic lights, and the number of bus launches are reduced by 7.12%, 7.32%, 42.29%, and 14.74%, respectively. Based on the launch control, the equivalent hydrogen consumption per 100 km of predictive energy management is reduced by 4.87%.
科研通智能强力驱动
Strongly Powered by AbleSci AI