The Characteristics Identification of Demand Side Load Mode in Power Grid Based on The Combinated Recognition Model
It is essential to know the demand side load characteristics and its group effect to power grid. Currently, there are numerous research methodologies of the demand-side characteristics, but the depth of the analysis of its characteristics is not enough. More effective information has not been excavated. Different from the traditional classification of load analysis by industry classes, this paper uses math tools such as Self- Organizing Maps, K-means, FCM, ID3 decision tree and so on for constructing one kind of combinated recognition model based on clustering, classification and decision tree analysis. Superiority of these algorithms are fully used in the comprehensive analysis for the demand load of power grid such as calculating of characteristics indicators, category judgment and mining, clustering appraisal, interpretation of classification knowledge. Daily load data set of 357 customers from 14 industries is used for a numerical examples analysis. In this case study, the characteristics of pattern such as power demand side load type, industry distribution and key index could be thoroughly excavated. At the same time, the validity and applicability of project of recognition model can also be verified.
Combinated recognition Data mining Load characteristics Load profile Power grid SOM
Weiting Xu Junyong Liu Youbo Liu Lei Li Yan Zhao Yong Tang
School of Electrical Engineering and Information, Sichuan University, Chengdu, China Shanghai Power Grid Company, Shanghai, China
国际会议
The International Conference on Electrical Engineering 2009(2009 电机工程国际会议)
沈阳
英文
1-5
2009-07-05（万方平台首次上网日期，不代表论文的发表时间）