武晓朦, 时政, 付子义, 等.基于密度峰值快速搜索聚类的多场景分布式电源规划[J].河南理工大学学报（自然科学版）,2022,41(2):117-123.

WU X M, SHI Z, FU Z Y,et al. Multi-scene distributed generation planning based on clustering by fast search and find of density peak[J].Journal of Henan Polytechnic University(Natural Science) ,2022,41(2):117-123.

基于密度峰值快速搜索聚类的多场景分布式电源规划

武晓朦1, 时政1, 付子义2, 刘欣雨3, 党建1, 李飞1

1.西安石油大学 陕西省油气井测控技术重点实验室,陕西 西安 710065;2.河南理工大学 电气工程与自动化学院,河南 焦作 454000;3.中国石油集团安全环保技术研究院大连分院,辽宁 大连 116031

摘要:针对间歇性分布式电源出力的随机性、负荷需求的不确定性以及分布式电源与负荷之间存在相关性的问题,采用拉丁超立方抽样并结合Spearman秩相关系数的Cholesky分解,得到分布式电源具有相关性的出力与负荷需求样本。通过密度峰值快速搜索聚类算法对相关性样本进行有效削减得到典型场景,以分布式电源投资运行费用和配电网向上级电网购电费用最小为优化目标,建立分布式电源多目标规划模型。最后通过二阶锥松弛将规划模型转化为混合整数二阶锥规划问题,并调用Cplex求解器对规划模型求解。IEEE 33节点算例结果验证了所提模型的合理性。

关键词:分布式电源规划;Spearman秩相关系数;密度峰值快速搜索聚类;二阶锥规划

doi:10.16186/j.cnki.1673-9787.2021020017

基金项目:国家自然科学基金资助项目（U20B2029）；陕西省科技计划基础研究项目（2021JM404）；陕西省重点实验室科研计划项 目（18JS094）；陕西省自然科学基础研究计划资助项目（2020JM-542）；西安石油大学研究生创新与实践能力培养项目 （YCS19141005）

收稿日期:2021/02/04

修回日期:2021/03/24

出版日期:2022/03/15

Multi-scene distributed generation planning based on clustering by fast search and find of density peak

WU Xiaomeng1, SHI Zheng1, FU Ziyi2, LIU Xinyu3, DANG Jian1, LI Fei1

1.Key Laboratory of Measurement and Control Technique of OH and Gas Well of Shaanxi Province, Xi ’ an Shiyou University, Xi ’ an  710065 , Shaanxi, China;2.School of Electrical Engineering and Automation, Henan Polytechnic University, Jiaozuo  454000 , Henan, China;3.Dalian Branch , China National Petroleum Corporation Research Institute of Safety & Environment Technology, Dalian  116031 , Liaoning, China

Abstract: Aiming at the problems of the randomness of intermittent distributed generation output,the uncertainty of load demand,and the correlation between distributed generation and load,Latin hypercube sampling and Cholesky decomposition combined with Spearman rank correlation coefficient were used to obtain an output and demand load relevant sample of distributed generation.The relevant sample was reduced by the algorithm of clustering by fast search and find of density peak(CFSFDP) to obtain typical scenes.The multi-objective distributed generation planning model was established with the optimization goal to minimize the investment and operation cost of the distributed generation and the power purchase cost of the superior grid.The programming model was transformed into a mixed-integer second-order cone programming( SOCP) problem through secondorder cone relaxation,and the Cplex solver was called to solve the programming model.The results of IEEE 33-bus distribution network were applied to verify the rationality of the proposed model and method.

Key words:distributed generation planning model;Spearman rank correlation coefficient;clustering by fast search and find of density peak;second-order cone programming

基于密度峰值快速搜索聚类的多场景分布式电源规划_武晓朦.pdf