基于恢复力约束的分布式储能优化规划  

Distributed Energy Storage Optimization Planning Based on Resilience Constraints

作　　者：严松 李穆[2] 朱俊宏 李妍[3] YAN Song;LI Mu;ZHU Junhong;LI Yan(State Grid Hubei Electric Power Company,Wuhan 430074,Hubei,China;Wuhan NARl Limited Liability Company,State Grid Electric Power Research Institute,Wuhan 430074,Hubei,China;School of Electronic Engineering,Huazhong University of Science and Technology,Wuhan 430074,Hubei,China)

机构地区：[1]国网湖北省电力有限公司,湖北武汉430074 [2]国网电力科学研究院武汉南瑞有限责任公司,湖北武汉430074 [3]华中科技大学电气与电子工程学院,湖北武汉430074

出　　处：《电网与清洁能源》2022年第6期17-23,共7页Power System and Clean Energy

基　　金：国家重点研发计划“智能电网技术与装备”重点专项(2017YFB0902800);国网湖北省电力有限公司科技项目(20210572)。

摘　　要：提出基于恢复力约束的分布式储能优化规划方法,以保证重要用户的恢复力为前提条件,采用双层耦合规划模型。内层模型在满足电网运行的潮流约束下,灵活地控制重要用户侧分布式储能参与需求侧响应,实现用电成本与动作频次最小的目标,采用竞争深度Q网络(dueling deep Q network,DDQN)结构的深度增强学习方法进行求解,内层模型将分布式储能响应策略传递给外层模型;外层模型进一步基于重要用户的恢复力约束和投资收益校核分布式储能的配置方案,通过双层优化耦合反馈,最终实现基于恢复力约束的分布式储能优化规划。通过分时电价引导分布式储能等重要互动资源参与配电网的优化运行,保证重要用户电力供应连续性的同时给用户明显的投资收益。最后以某10 kV变电站的重要用户储能优化配置为例,分析了所提方法的有效性。In this paper,an optimal planning method for distributed energy storage based on resilience constraints is proposed.On the premise of ensuring the resilience of important users,a two-layer coupled programming model is developed.The inner layer model flexibly controls the distributed energy resources of important users to participate in the demand-side response under the constraints of power flow and operation voltage,so as to achieve the goal of minimizing the cost of electricity consumption and action frequency.The deep reinforcement learning algorithm of the dueling deep Q network(DDQN)structure is used to solve the problem.The inner model passes the important user-side energy storage scheduling strategy to the outer model.The outer model further checks the distributed energy storage configuration scheme based on resilience constraints and investment income.Through the double-layer optimization coupling feedback,the optimization planning of distributed energy storage based on resilience constraints is finally realized.The time-of-use electricity price guides important interactive resources such as distributed energy storage to participate in the optimal operation of the distribution network,which ensures the continuity of power supply for important users and giving users significant investment returns.Finally,the paper takes the optimal configuration of energy storage for important users of a 10 kV substation as an example,analyzes the effectiveness of the proposed method.

关 键 词：恢复力 分布式储能 竞争深度Q网络 配电网规划 

分 类 号：TM732[电气工程—电力系统及自动化]