原文信息：
Imitation reinforcement learning energy management for electric vehicles with hybrid energy storage system
原文链接：
https://www.sciencedirect.com/science/article/pii/S0306261924022153
Highlights
•提出了用于功率分配的对抗模仿强化学习方法。
•通过离线优化建立专家知识。
•将智能体动态地从专家指导过渡到自我探索。
•减少无效的探索，加快训练，提高奖励。
•在不同的初始SoCs和驾驶工况下，对该方法进行了验证。
Abstract
Deep reinforcement learning has become a promising method for the energy management of electric vehicles. However, deep reinforcement learning relies on a large amount of trial-and-error training to acquire near-optimal performance. An adversarial imitation reinforcement learning energy management strategy is proposed for electric vehicles with a hybrid energy storage system to minimize the cost of battery capacity loss. Firstly, the reinforcement learning exploration is guided by expert knowledge, which is generated by dynamic programming under various standard driving conditions. The expert knowledge is represented as the optimal power allocation mapping. Secondly, at the early imitation stage, the action of the reinforcement learning agent approaches the optimal power allocation mapping rapidly by using adversarial networks. Thirdly, a dynamic imitation weight is developed according to the Discriminator of adversarial networks, making the agent transit to self-explore the near-optimal power allocation under online driving conditions. Results demonstrate that the proposed strategy can accelerate the training by 42.60% while enhancing the reward by 15.79% compared with traditional reinforcement learning. Under different test driving cycles, the proposed method can further reduce the battery capacity loss cost by 5.1%-12.4%.
Keywords
Imitation learning；
Hybrid energy storage system；
Deep reinforcement learning；
Battery degradation；
Generative adversarial imitation learning;
Graphics
Fig. 1. Schematic diagram of the electric vehicle with semi-active hybrid energy storage system.
Fig. 2. The framework of the energy management strategy based on the adversarial imitation reinforcement learning.
Fig. 8. Comparison of energy management results for the proposed method, DP, DDPG, the GAIL-based method under the training driving cycle: WLTP.
Fig. 9. Comparison of robustness for the proposed method, DP, DDPG, and the GAIL-based method under WLTP with different initial values of SoCsc
Fig. 10. Comparison of generalization capacity for the proposed method, DP, DDPG, and the GAIL-based method under three test driving cycles: US06, NEDC, and UDDS.
团队简介
本研究由中南大学的研究人员完成。
通信作者简介：
武悦，中南大学电子信息学院讲师，从事电动汽车混合储能系统能量管理和热管理方面的研究，在Applied Energy、Energy、Energy Conversion and Management、Solar Energy、IEEE TITS、IEEE TVT等能源电力领域SCI期刊发表论文多篇。
第一作者简介：
刘伟荣，中南大学计算机学院教授，从事电动汽车、储能系统管理与控制、人工智能等方面的研究，在Applied Energy、Energy、Applied Soft Computing、IEEE TNNLS、IEEE TITS、IEEE TVT、IEEE TPDS等SCI期刊发表论文多篇。
姚鹏飞，中南大学计算机学院硕士研究生，从事电动汽车混合储能系统能量管理和热管理方面的研究，发表Applied Energy期刊论文一篇、IEEE International Conference on High Performance Computing and Communications 会议论文一篇。
关于Applied Energy
本期小编：陈媛；审核人：武龙星
《Applied Energy》是世界能源领域著名学术期刊，在全球出版巨头爱思唯尔 (Elsevier) 旗下，1975年创刊，影响因子10.1，CiteScore 21.2，本刊旨在为清洁能源转换技术、能源过程和系统优化、能源效率、智慧能源、环境污染物及温室气体减排、能源与其他学科交叉融合、以及能源可持续发展等领域提供交流分享和合作的平台。开源（Open Access）姊妹新刊《Advances in Applied Energy》影响因子13.0，CiteScore 23.9。全部论文可以免费下载。在《Applied Energy》的成功经验基础上，致力于发表应用能源领域顶尖科研成果，并为广大科研人员提供一个快速权威的学术交流和发表平台，欢迎关注！
公众号团队小编招募长期开放，欢迎发送自我简介（含教育背景、研究方向等内容）至wechat@applied-energy.org
点击“阅读原文”
喜欢我们的内容？
点个“赞”或者“再看”支持下吧！
