基于蒙特卡罗树搜索和ε约束算法的配电网源网荷储协同运行优化方法.pdfVIP

第44卷第2期 电力与能源 2023年4月 179 DOI:10.11973/dlyny202302017 基于蒙特卡罗树搜索和ε约束算法的配电网 源网荷储协同运行优化方法 林君豪，严 浩，王俊翔 （国网上海市电力公司市北供电公司，上海200072） 摘要：配电网负荷侧和需求侧出现的大量灵活资源给网络运行带来了较大不确定性。为了充分调动各种灵 活资源、改善网络的运行状态，提出了计及源网荷储的协同运行优化模型。该模型以降低系统运行成本及提 高电源侧和负荷侧的曲线相关性为目标，调控资源包括配电网中的分布式电源、可响应负荷、储能系统以及网 架结构。针对多时段优化问题的复杂性，提出了基于蒙特卡罗树搜索和约束算法的联合求解方法，该方法 同时还解决了传统配电网动态重构方法对重构时段划分不够客观的局限性。最后，通过对PG69标准配电网 系统进行仿真分析，验证所提方法的有效性。 关键词：源网荷储协同运行优化；需求响应；配电网重构；蒙特卡罗树搜索 作者简介：林君豪（1992一），男，博士，主要从事人工智能在配电网中的应用。 中图分类号：TM715 文献标志码：A 文章编号：2095-1256（2023)02-0179-08 Source-network-load-storage Cooperative Operation Optimization Method of Distribution Network Based on Monte Carlo Tree Search and e-constraint Algorithm LIN Junhao, YAN Hao, WANG Junxiang (State Grid Shibei Power Supply Company, SMEPC, Shanghai 200072, China) Abstract: A large number of flexible resources on load side and demand side of distribution network bring a lot of uncertainty to network operation. In order to fully mobilize various flexible resources and improve the running state of the network, a source-network-load-storage collaborative operation optimization model is proposed. The model aims at reduce the operating cost of the system and improve the curve correlation between source power and load. The controlled resources include distributed power, responsive load, energy storage system and the grid structure in the distribution network. In view of the complexity of multi-period optimization problem, a joint solu- tion method based on Monte Carlo tree search and e-constraint algorithm is proposed in this paper. The method also so