考虑决策相关随机规划的光热电站容量配置  

Capacity Optimization of Concentrating Solar Power Plant Based on Decision-dependent Stochastic Programming

作　　者：史昭娣 王伟胜 黄越辉 侯云鹤[2] 董凌 SHI Zhaodi;WANG Weisheng;HUANG Yuehui;HOU Yunhe;DONG Ling(State Key Laboratory of Operation and Control of Renewable Energy&Storage Systems(China Electric Power Research Institute),Haidian District,Beijing 100192,China;Department of Electrical Engineering,University of Hong Kong,Hongkong 999077,China;State Grid Qinghai Electric Power Company,Xining 810000,Qinghai Province,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 [2]香港大学电气工程系,中国香港特别行政区999077 [3]国网青海省电力公司,青海省西宁市810000

出　　处：《中国电机工程学报》2020年第23期7511-7521,共11页Proceedings of the CSEE

基　　金：国家重点研发计划项目(2017YFB0902200);国家电网公司科技项目(5228001700CW);青海省科技计划重大科技专项(2018-GX-A6)。

摘　　要：可再生能源电力系统具有很强的不确定性。光热电站作为一种新的可再生能源利用技术,可为可再生能源电力系统提供灵活性。传统随机规划方法处理容量规划问题中的不确定性通常预先固定未来不确定性概率分布,然而现有研究表明现阶段投资决策也会影响未来不确定性。基于此,该文提出一种光热电站容量规划的决策相关随机规划模型,其中电价结果的概率分布依赖于投资决策变量,这表明未来不确定性不仅影响当前决策,也受到当前决策影响。为了解决模型中的双线性项,采用准精确法将非线性模型重构为混合整数线性规划模型以便于求解。为了充分考虑可再生能源不确定性,进一步依据历史出力数据生成多组新能源出力序列,验证所提模型可适用于不同新能源出力场景。以100%可再生能源渗透系统为测试系统,验证所提决策相关随机规划模型的合理性和有效性。Renewable energy(RE) power systems have strong uncertainties. Concentrating solar power(CSP) plant, as a new renewable energy utilization technology, can provide flexibility for RE power systems. The traditional stochastic programming method generally fixes the future uncertainty probability distribution in the capacity planning problem. However, the existing research indicates that the current decision also affects the future uncertainty. Based on this problem, this paper proposed a decision-dependent stochastic programming model for capacity planning of CSP plants. The probability distribution of electricity price output depends on decision variables, which indicates that future uncertainty is not only affecting but also affected by current decisions. In order to solve the bilinear term in the model, the quasi-precision method was used to reconstruct the nonlinear model into a mixed integer linear programming(MILP) model. The test system is a 100% renewable energy penetration system. The calculation results verify the rationality and effectiveness of the proposed decision-dependent stochastic programming model.

关 键 词：光热发电 随机规划 决策相关 内生不确定性 可再生能源 

分 类 号：TM73[电气工程—电力系统及自动化]