引用本文:张丽琴,谢俊,张秋艳,符登辉.基于Shapley值抽样估计法的风-光-水互补发电增益分配方法[J].电力自动化设备,2021,41(9):
ZHANG Liqin,XIE Jun,ZHANG Qiuyan,FU Denghui.Synergistic benefit allocation method for wind-solar-hydro complementary generation with sampling-based Shapley value estimation method[J].Electric Power Automation Equipment,2021,41(9):
【打印本页】   【HTML】   【下载PDF全文】   查看/发表评论  【EndNote】   【RefMan】   【BibTex】
←前一篇|后一篇→ 过刊浏览    高级检索
本文已被:浏览 3210次   下载 1351  
基于Shapley值抽样估计法的风-光-水互补发电增益分配方法
张丽琴, 谢俊, 张秋艳, 符登辉
河海大学 能源与电气学院,江苏 南京 211100
摘要:
在多利益主体市场环境下,建立公平、合理、高效的互补发电增益分配方法是保障风-光-水互补发电优化调度能够实施的关键。针对风-光-水互补发电的特点,在建立考虑随机性、波动性和互补效益的风-光-水互补发电优化调度模型的基础上,提出风-光-水互补发电增益量化方法;在经典Shapley值(SV)法的基础上,提出基于Shapley值抽样估计(SSVE)法的风-光-水互补发电增益分配方法,并提出基于强化学习(RL)的样本量分配法,以提高SSVE法的精确性和计算效率。算例结果验证了采用RL样本量分配法的SSVE法的有效性,且SSVE法能够有效减少计算量以及解决经典SV法的组合爆炸问题。
关键词:  风-光-水互补发电  增益分配  Shapley值抽样估计法  样本量分配方法
DOI:10.16081/j.epae.202108024
分类号:TM73
基金项目:国家自然科学基金资助项目(U1965104, U1766203);国家重点研发计划项目(2019YFE0105200)
Synergistic benefit allocation method for wind-solar-hydro complementary generation with sampling-based Shapley value estimation method
ZHANG Liqin, XIE Jun, ZHANG Qiuyan, FU Denghui
College of Energy and Electrical Engineering, Hohai University, Nanjing 211100, China
Abstract:
Under the market environment of multiple stakeholders, it is critical to build a fair, reasonable and efficient complementary generation synergistic benefit allocation method for guaranteeing implementation of wind-solar-hydro complementary generation optimal dispatching. Aiming at the characteristic of wind-solar-hydro complementary generation, an optimal dispatching model of wind-solar-hydro complementary generation is built by considering randomness, volatility and complementary benefit, on this basis, a synergistic benefit quantification method for wind-solar-hydro complementary generation is proposed. On the basis of traditional SV(Shapley Value) method, a synergistic benefit allocation method for wind-solar-hydro complementary generation is proposed based on SSVE(Sampling-based Shapley Value Estimation) method, and a sample distribution method based on RL(Reinforcement Learning) is proposed to improve the accuracy and computational efficiency of SSVE method. Case results verify the effectiveness of SSVE method with RL sample distribution method, and SSVE method can reduce computational burden effectively and solve the combinatorial explosion problem of traditional SV method.
Key words:  wind-solar-hydro complementary generation  synergistic benefit allocation  SSVE method  sample distribution method

用微信扫一扫

用微信扫一扫