| 142 | 0 | 444 |
| 下载次数 | 被引频次 | 阅读次数 |
针对主动配电网（active distribution networks,ADNs）源负荷不确定性及分布式能源普及带来的优化调度挑战，提出基于改进双延迟深度确定性策略梯度（twin delayed deep deterministic policy gradient,TD3）的分布式资源协同优化方法。首先，构建含燃气轮机、储能、可再生能源及综合需求响应的调度框架，通过卷积-长短期记忆组合网络（convolutional neural network-long short-term memory,CNN-LSTM）提取源荷时空耦合特征。然后，设计动态权重机制平衡多目标需求，并在Critic评价网络中融合评价器错误正则化（error regularized critic,ERC）与L2正则化，强化Critic网络稳定性。最后，通过制定燃气轮机启停与储能充放电基础计划，并基于实时数据动态调整策略，抑制功率波动。仿真结果表明，相较传统策略，本文所提方法可降低系统运行成本12.3%，提高新能源利用率17%，抑制主配电网互联线路功率波动37%，所提优化方法的收敛速度与时变适应性显著提升，可为高比例新能源接入的ADNs提供兼顾经济性与可靠性的低碳高效调度方案。
Abstract：In response to the challenges of bidirectional power flow scheduling caused by the uncertainty of sources and loads,as well as the widespread adoption of distributed energy in active distribution networks(ADNs),this paper proposes a multitime-scale collaborative optimization method based on an improved twin delayed deep deterministic policy gradient(TD3). A scheduling framework containing gas turbines,energy storage,renewable energy,and integrated demand response is constructed. Specifically,a hybrid convolutional neural network-long short-term memory(CNN-LSTM)network is utilized to extract the spatiotemporal coupling features of source and load. Meanwhile,a dynamic weighting mechanism is designed to balance the multi-objective demands. Furthermore,error regularized critic(ERC)and L2 regularization are integrated into the critic network to enhance its stability. By formulating the basic plans for the start-stop operations of gas turbines and the charging-discharging of energy storage,and dynamically adjusting strategies based on real-time data,power fluctuations can be suppressed. Simulation results show that compared to traditional strategies,the system operation cost is reduced,the utilization rate of new energy is increased,the power fluctuations of the main distribution network interconnection lines are suppressed,the convergence speed and time-varying adaptability are significantly improved,providing a low-carbon and efficient scheduling solution that balances economy and reliability for ADNs with a high proportion of new energy.
[1]王振华.构建新型电力系统助力实现“双碳”目标[J].中国经济报告，2021(4):133-134.Wang Zhenhua. Building a new power system to help achieve the goal of“double carbon”[J]. China Policy Review,2021(4):133-134.
[2]卓振宇，张宁，谢小荣，等.高比例可再生能源电力系统关键技术及发展挑战[J].电力系统自动化，2021,45(9):171-191.Zhuo Zhenyu,Zhang Ning,Xie Xiaorong,et al. Key technologies and developing challenges of power system with high proportion of renewable energy[J]. Automation of Electric Power Systems,2021,45(9):171-191.
[3]欧阳武，傅皆恺.主动配电网研究进展[J].江西电力，2016,40(12):6-10.Ouyang Wu,Fu Jiekai. Research progress of active distribution network[J]. Jiangxi Electric Power,2016,40(12):6-10.
[4]黄河，高松，朱磊，等.考虑多能互补的主动配电网规划与高效Benders求解方法[J].中国电力，2019,52(6):11-18.Huang He,Gao Song,Zhu Lei,et al. Active distribution network planning model and benders solution strategy considering multienergy complementation[J]. Electric Power,2019,52(6):11-18.
[5]熊珞琳，毛帅，唐漾，等.基于强化学习的综合能源系统管理综述[J].自动化学报，2021,47(10):2321-2340.Xiong Luolin,Mao Shuai,Tang Yang,et al. Reinforcement learning based integrated energy system management:a survey[J]. Acta Automatica Sinica,2021,47(10):2321-2340.
[6]张启亮，李开灿，孔维娜，等.基于分布鲁棒的主动配电网多时间尺度优化调度[J].山东电力技术，2025,52(10):34-45.Zhang Qiliang,Li Kaican,Kong Weina,et al. Multi-time scale optimal dispatch of active distribution network based on distributionally robust optimization[J]. Shandong Electric Power,2025,52(10):34-45.
[7]颜家扬.基于深度强化学习的光伏微电网调度策略研究[D].成都：电子科技大学，2025.
[8]刘丽莉.考虑EV和可再生能源接入的SST两级联动微分平坦控制[D].乌鲁木齐：新疆大学，2021.
[9]唐文虎，韦慈航，柳洲，等.考虑负荷削减公平性的配电网两阶段韧性提升策略[J].广东电力，2025,38(9):52-67.Tang Wenhu,Wei Cihang,Liu Zhou,et al. Two-stage resilience enhancement strategy for distribution networks considering fairness of load shedding[J]. Guangdong Electric Power,2025,38(9):52-67.
[10]张贵辰，田磊，周京华.基于改进灰狼优化算法的光储微电网经济优化调度[J].广东电力，2025,38(6):30-38.Zhang Guichen,Tian Lei,Zhou Jinghua. Optimal dispatch of photovoltaic and energy storage microgrid based on improved grey wolf optimization algorithm[J]. Guangdong Electric Power,2025,38(6):30-38.
[11]陈将宏，李雪莲，袁腾，等.源网荷储协同的主动配电网多时间尺度优化调度[J].重庆理工大学学报，2024,38(5):343-353.Chen Jianghong,Li Xuelian,Yuan Teng,et al. Multi-timescale optimal scheduling of active distribution network based on sourcegrid-load-storage coordination[J]. Journal of Chongqing Institute of Technology,2024,38(5):343-353.
[12]郑日红，都成刚，葛景，等.高比例新能源接入的主动配电网优化策略[J].内蒙古电力技术，2025,43(4):76-84.Zheng Rihong,Du Chenggang,Ge Jing,et al. Active distribution network optimization strategies for high proportion of new energy access[J]. Inner Mongolia Electric Power,2025,43(4):76-84.
[13]彭钢，寇启龙，方涛，等.考虑新能源消纳率的主动配电网储能优化配置策略[J].浙江电力，2025,44(1):84-94.Peng Gang,Kou Qilong,Fang Tao,et al. An optimal strategy for energy storage allocation in active distribution networks considering new energy consumption rates[J]. Zhejiang Electric Power,2025,44(1):84-94.
[14]李振坤，徐廷君.考虑“源储荷”协同的主动配电网双层优化配置[J].智慧电力，2025,53(3):77-83.Li Zhenkun,Xu Tingjun. Bi-level optimal configuration of active distribution network considering source-storage-load coordination[J]. Smart Power,2025,53(3):77-83.
[15]余莎，何光层，刘志坚，等.含碳捕集的电-气综合能源系统低碳经济调度[J].广东电力，2022,35(2):74-82.Yu Sha,He Guangceng,Liu Zhijian,et al. Low-carbon economic dispatch of integrated electricity-gas energy system with carbon capture[J]. Guangdong Electric Power,2022,35(2):74-82.
[16]息佳琦，石晓楠，杨昭，等.基于SSA-XGBoost的变压器故障诊断[J].黑龙江大学自然科学学报，2023,40(6):721-729.Xi Jiaqi,Shi Xiaonan,Yang Zhao,et al. Transformer fault diagnosis based on SSA-XGBoost[J]. Journal of Natural Science of Heilongjiang University,2023,40(6):721-729.
[17]王宇.考虑水位时序分布的水电调度降维方法研究[D].大连：大连理工大学，2022.
[18]黎家明，霍群海，尹靖元，等.考虑柔性有载调压变压器的两阶段电压无功优化协调控制方法研究[J].电网技术，2025,49(1):272-283.Li Jiaming,Huo Qunhai,Yin Jingyuan,et al. Research on twostage voltage and reactive power optimization coordinated control method considering flexible on-load tap changer[J]. Power System Technology,2025,49(1):272-283.
[19]薛霖，牛涛，方斯顿，等.计及高比例风电暂态电压安全的主从协同动态无功优化方法[J].电力系统自动化，2023,47(17):57-66.Xue Lin,Niu Tao,Fang Sidun,et al. Master-slave cooperative dynamic reactive power optimization method considering transient voltage security of high proportion of wind power[J]. Automation of Electric Power Systems,2023,47(17):57-66.
[20]高倩，杨知方，李文沅.电力系统混合整数线性规划问题的运筹决策关键技术综述与展望[J].电工技术学报，2024,39(11):3291-3307.Gao Qian,Yang Zhifang,Li Wenyuan. Prospect on operations research for mixed-integer linear programming problems in power systems[J]. Transactions of China Electrotechnical Society,2024,39(11):3291-3307.
[21]Guan J Y,Tang H,Wang J Y,et al. A GAN-based fully modelfree learning method for short-term scheduling of large power system[J]. IEEE Transactions on Power Systems,2022,37(4):2655-2665.
[22]董哲源，徐天奇，李琰，等.考虑源荷储协同的主动配电网多时间尺度优化调度[J].科学技术与工程，2024,24(24):10321-10329.Dong Zheyuan,Xu Tianqi,Li Yan,et al. Multi-time scale optimal scheduling of active distribution network considering source load storage coordination[J]. Science Technology and Engineering,2024,24(24):10321-10329.
[23]Wan H C,Fang R C,Lei H,et al. Day-day-day Cooperative scheduling method of active distribution network based on optimal economic operation region[J]. Power Grid Technology,2019,48(4):1602-1611.
[24]徐钰涵，季天瑶，李梦诗.基于深度强化学习的微电网日前日内协调优化调度[J].南方电网技术，2024,18(9):106-116.Xu Yuhan,Ji Tianyao,Li Mengshi. Day-ahead and intra-day coordinated optimal scheduling of microgrid based on deep reinforcement learning[J]. Southern Power System Technology,2024,18(9):106-116.
[25]唐杰，吕林，叶勇，等.多时间尺度下主动配电网源-储-荷协调经济调度[J].电力系统保护与控制，2021,49(20):53-64.Tang Jie,LüLin,Ye Yong,et al. Source-storage-load coordinated economic dispatch of an active distribution network under multiple time scales[J]. Power System Protection and Control,2021,49(20):53-64.
[26]LUO Chunjian,Li Yaowang,Xu Hanping,et al. Analysis of the influence of demand response uncertainty on Day-Ahead optimal scheduling[J]. Automation of Electric Power Systems,2017,41(5):22-29.
[27]罗金满，刘丽媛，刘飘，等.考虑源网荷储协调的主动配电网优化调度方法研究[J].电力系统保护与控制，2022,50(1):167-173.Luo Jinman,Liu Liyuan,Liu Piao,et al. An optimal scheduling method for active distribution network considering source network load storage coordination[J]. Power System Protection and Control,2022,50(1):167-173.
[28]Tomin N,Shakirov V,Kozlov A,et al. Design and optimal energy management of community microgrids with flexible renewable energy sources[J]. Renewable Energy,2022,183:903-921.
[29]Duan J L,Wang W X,Xiao L M,et al. Distributional soft actorcritic with three refinements[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence,2025,47(5):3935-3946.
[30]Beikmohammadi A,Magnússon S. Accelerating actor-critic-based algorithms via pseudo-labels derived from prior knowledge[J].Information Sciences,2024,661:120182.
[31]Yan L,Liu Z,Philip Chen C L,et al. Adaptive fixed-time inverse optimal consensus of multi-agent systems with limited-time interval state constraints[J]. Information Sciences,2024,661:119884.
[32]唐昊，张庆虎，方道宏，等.需求响应下基于深度强化学习的综合能源系统能量管理策略[J].控制理论与应用，2026,43(1):205-215.Tang Hao, Zhang Qinghu, Fang Daohong, et al. Energy management strategy of integrated energy system considering demand response by using deep reinforcement learning[J]. Control Theory&Applications,2026,43(1):205-215.
[33]王岩，李冠冠，邸蕴鹏，等.含电热综合能源系统的直流配电网分布式优化调度[J].山东电力技术，2025,52(4):40-48.Wang Yan,Li Guanguan,Di Yunpeng,et al. Distributed optimization scheduling of DC distribution network with electric heating integrated energy system[J]. Shandong Electric Power,2025,52(4):40-48.
[34]杨校辉，邱俊宏，张娟，等.一种基于综合能源优化调配的充电站运营管理系统研究[J].电力系统保护与控制，2021,49(5):173-179.Yang Xiaohui,Qiu Junhong,Zhang Juan,et al. Research on a charging station operation management system based on integrated energy optimization[J]. Power System Protection and Control,2021,49(5):173-179.
基本信息:
DOI：10.20097/j.cnki.issn1007-9904.2501017
中图分类号:TM73
引用信息:
[1]王开林,郭阳,唐昊,等.基于改进TD3的主动配电网分布式资源协同优化调度[J].山东电力技术().DOI:10.20097/j.cnki.issn1007-9904.2501017.
基金信息:
国家自然科学基金项目（62273130）; 国网安徽省电力有限公司蚌埠供电公司科技项目（B312L024000K）
2026-05-27
2026-05-27
2026-05-27