计及深度信念网络场景生成的风/光/储协同优化规划方法  

Collaborative Optimization Planning for Wind/PV/storage Based on Scenario Generated by Deep Belief Network

作　　者：史昭娣 黄越辉 李湃 王伟胜 SHI Zhaodi;HUANG Yuehui;LI Pai;WANG Weisheng(State Key Laboratory of Operation and Control of Renewable Energy&Storage Systems(China Electric Power Research Institute),Haidian District,Beijing 100192,China)

机构地区：[1]新能源与储能运行控制国家重点实验室(中国电力科学研究院有限公司),北京市海淀区100192

出　　处：《电网技术》2023年第4期1331-1339,共9页Power System Technology

基　　金：国家电网有限公司总部管理科技项目(1300-202155460A-0-0-00):“碳达峰碳中和背景下我国能源电力低碳发展模型方法研究及关键技术评估”。

摘　　要：新能源出力具有随机波动性,考虑具有随机性的新能源出力场景对新能源与储能规划具有重要意义。现有基于随机抽样的场景生成方法难以考虑新能源波动性和出力时序性。基于此,首先提出一种基于深度信念网络的新能源出力时间序列场景生成方法,以充分挖掘新能源出力特性。其次,综合考虑系统经济性和环保性,建立了基于中长期时序电力平衡的风/光/储协同规划模型。为充分计及新能源出力随机性,采用多组风/光出力模拟序列作为规划模型输入,以所有场景期望结果作为最终规划结果,并对规划结果进行可靠性校核。最后,以我国某省级电网为例进行算例分析,研究不同随机场景和参数灵敏度下的风/光/储容量规划方案,为考虑新能源出力随机性的电源规划问题提供解决方案。As renewable energy output has stochastic and fluctuating characteristics,it is of great significance to have the renewable energy output stochastic scenarios in solving the problems as renewable generation expansion and energy storage planning.However,the existing methods of scenario generation based on the statistical models hardly consider the stochastic and fluctuating characteristics of renewable energy output.Based on this,this paper first proposes a time sequence scenario generation based on the deep belief network to fully explore the characteristics of renewable energy output.Secondly,comprehensively considering the system economy and the environmental protection costs,a optimization planning model for wind/photovoltaic(PV)/storage based on the mid-to-long-term timing power balance simulation is established.In order to fully take into account the stochastics output,this paper adopts several annual wind/PV output sequences as the input of the planning model.Taking the expected results of all the scenarios as the final optimization result,the reliability of the planning results is checked.Finally,a provincial power grid in China is taken as an example to analyze the wind/PV/storage capacity planning schemes under different stochastic scenarios and parameter sensitivity.Therefore,a solution to the renewable generation planning problem considering the stochastic output of the wind/PV is provided.

关 键 词：新能源 出力随机性 电源规划 深度信念网络 场景生成 

分 类 号：TM721[电气工程—电力系统及自动化]