摘要
讨论和完善在市场化条件下,电力交易中心对中长期交易的组织和管理工作的思路,提出利用负荷聚类等数据挖掘技术,获取并分析用户负荷特性,进而为中长期交易安排提供参考的设想,以提升其安全性和科学性,减少制度性成本;并利用k-means聚类方法分析陕西省负荷特性数据,将交易的组织和与相应的聚类中心联系起来,结合聚类结果对中长期交易的组织和管理提出建议。
To improve the provincial electricity trading platform's management and arrangement of medium and long-term electricity transactions in the context of electrical marketization,this paper proposes to use data mining technology such as loadinng clustering method,acquiring and analyzing the user load characteristics and then providing the reference plan for medium and long-term trading arrangements so as to enhance safety and scientificalness and reduce the institutional cost. In addition,the K-means clustering method is used to analyze the load characteristic data of Shaanxi province,linking the organization of transaction and the corresponding clustering centers,and finally suggestions are put forward for the organization and management of medium and long term transactions based on the clustering result.
引文
[1]刘思捷,蔡秋娜,杨韵,等.适应月度直接交易的电量安全校核模型及方法[J].广东电力,2017,30(3):15-20.LIU Sijie,CAI Qiuna,YANG Yun,et al.Electricity quantity checking model and method for monthly direct transaction[J].Guangdong Electric Power,2017,30(3):15-20.
[2]谷峰.电力市场顶层设计与交易现实的博弈妥协——解读电力中长期交易基本规则[J].中国电力企业管理,2017(1):32-35.GU Feng.The top of the electricity market design and trading reality of the game and compromise—Interpreting the basic rules of medium and long term transactions[J].China Power Enterprise Management,2017(1):32-35.
[3]张忠华.电力系统负荷分类研究[D].杭州:浙江大学,2017.
[4]肖琪.基于优化K-means算法的电力负荷分类研究[D].大连:大连理工大学,2015.
[5]程祥.基于负荷量测数据的电力负荷聚类方法研究[D].杭州:浙江大学,2017.
[6]陈凡,刘海涛,黄正,等.基于改进k-均值聚类的负荷概率模型[J].电力系统保护与控制,2013(22):128-133.CHEN Fan,LIU Haitao,HUANG Zheng,et al.Probabilistic load model based on improved k-means clustering algorithm[J].Power System Protection and Control,2013(22):128-133.
[7]毛国君.数据挖掘原理与算法[M].北京:清华大学出版社,2005.
[8]VISALAKSHI N K,THANGAVEL K.Impact of normalization in distributed K-Means clustering[J].International Journal of Soft Computing,2009(4):168-172.