电工技术学报  2022, Vol. 37 Issue (zk1): 225-234    DOI: 10.19595/j.cnki.1000-6753.tces.L90399
电力电子 |
基于强化学习的双主动半桥直流变换器的效率优化方案
胡广1, 胡维昊1, 唐远鸿1, 黄琦1, 陈哲2
1.电子科技大学机械与电气工程学院 电力系统广域测量与控制四川省重点实验室 成都 611731;
2.奥尔堡大学能源系 奥尔堡 DK-9110
Efficiency Optimization Scheme of Dual-Active-Half-Bridge DC-DC Converter Based on Reinforcement Learning
Hu Guang1, Hu Weihao1, Tang Yuanhong1, Huang Qi1, Chen Zhe2
1. University of Electronic Science and Technology of China School of Mechanical and Electrical Engineering Power System Wide Area Measurement and Control of the Key Laboratory of Sichuan Province Chengdu 611731 China;
2. Department of Energy Technology Aalborg University Aalborg DK-9110 Denmark
全文: PDF (2260 KB)   HTML
输出: BibTeX | EndNote (RIS)      
摘要 为减小双主动半桥直流变换器(DAHB)在传统双重移相(DPS)调制下轻载区较大的功率损耗,该文提出一种基于强化学习(RL)的DPS调制效率优化方案(QDPS)。该方案使用Q-learning算法作为一种RL的典型算法,通过对智能体进行离线训练,得到优化后的调制策略QDPS。该策略能够为DAHB直流变换器提供最优的移相角,有效减小变换器在轻载区的功率损耗。与现有的单移相控制策略和双重移相控制策略相比,QDPS移相控制策略能够有效提高变换器的运行效率,改善变换器的性能。最后,通过Matlab/Simulink仿真验证了所提出优化方案的有效性。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
E-mail Alert
RSS
作者相关文章
胡广
胡维昊
唐远鸿
黄琦
陈哲
关键词 双重移相控制双主动半桥直流变换器强化学习Q-learning效率    
Abstract:Aiming to optimize the modulation efficiency of the dual-active-half-bridge (DAHB) DC-DC converter under traditional dual-phase-shift (DPS) modulation in the light load region with large power loss, a DPS modulation efficiency optimization scheme (QDPS) based on reinforcement learning (RL) is proposed in this paper. Using the Q-learning algorithm as a typical RL algorithm, the scheme obtains the optimized modulation strategy QDPS through offline training of the agent, which can provide the optimal phase shift angle for the DAHB DC-DC converter and effectively reduce the power loss of the converter in the light load region. Compared with the existing single-phase-shift and dual-phase-shift control strategy, the QDPS phase shift control strategy can effectively increase the operating efficiency and improve the performance of the converter. Finally, Matlab/Simulink simulation verifies the effectiveness of the proposed optimization scheme.
Key wordsDual-phase-shift control    dual-active-half-bridge DC-DC converter    reinforcement learning    Q-learning    efficiency   
收稿日期: 2020-07-11     
PACS: TM46  
基金资助:四川省科技计划(杰出青年科技人才)资助项目(2020JDJQ0037)
通讯作者: 胡维昊 男,1982年生,教授,博士生导师, 研究方向为人工智能在电力系统中的应用、可再生能源发电技术。E-mail:whu@uestc.edu.cn   
作者简介: 胡 广 男,1995年生,硕士,研究方向为双有源桥DC-DC变换器。E-mail:1006150682@qq.com
引用本文:   
胡广, 胡维昊, 唐远鸿, 黄琦, 陈哲. 基于强化学习的双主动半桥直流变换器的效率优化方案[J]. 电工技术学报, 2022, 37(zk1): 225-234. Hu Guang, Hu Weihao, Tang Yuanhong, Huang Qi, Chen Zhe. Efficiency Optimization Scheme of Dual-Active-Half-Bridge DC-DC Converter Based on Reinforcement Learning. Transactions of China Electrotechnical Society, 2022, 37(zk1): 225-234.
链接本文:  
https://dgjsxb.ces-transaction.com/CN/10.19595/j.cnki.1000-6753.tces.L90399          https://dgjsxb.ces-transaction.com/CN/Y2022/V37/Izk1/225