电力需求侧管理2024，Vol.26Issue(4)：1-8,8.DOI:10.3969/j.issn.1009-1831.2024.04.001
基于双延迟深度确定性策略梯度算法的微电网能源优化分配策略研究
Research on energy optimization allocation strategy for microgrids based on double delay deep deterministic strategy gradient algorithm
摘要
Abstract
In island mode,microgrids need to operate independently from traditional power systems,efficiently coordinating internal ener-gy to ensure the continuity and efficiency of energy supply.The twin delayed deep deterministic policy gradient algorithm significantly im-proves the processing efficiency and accuracy of complex continuous control tasks through policy delay updates and the introduction of du-al Q networks.Based on this way,an energy optimization allocation strategy is designed for microgrids embedded with fuel cells based on the TD3 algorithm,to improve the stable power supply capacity and quality of the microgrid system,reduce energy consumption and opera-tion costs,and enhance the system's economy and reliability.Through comprehensive analysis,the comprehensive performance of the de-signed energy optimization allocation strategy in different scenarios is comprehensively evaluated.The results show that by optimizing the charging and discharging modes and ratios of fuel cell systems,the energy optimization allocation strategy designed based on TD3 algo-rithm performs better than traditional algorithms in improving energy allocation efficiency,shortening response time,and reducing operat-ing costs.The research results have verified the efficient adaptability of TD3 algorithm in dealing with fluctuations in renewable energy generation power output and changes in load demand,and it has wide applicability in practical energy management scenarios.关键词
微电网/燃料电池/能源优化分配/TD3算法/深度强化学习Key words
microgrids/fuel cells/optimal energy allocation/TD3 algorithm/deep reinforcement learning分类
信息技术与安全科学引用本文复制引用
杨家令,陈涛,高赐威..基于双延迟深度确定性策略梯度算法的微电网能源优化分配策略研究[J].电力需求侧管理,2024,26(4):1-8,8.基金项目
国家自然科学基金《基于深度强化学习技术的可交易能源系统智能决策问题研究》(52107079) （52107079）
