Power distribution abnormal data identification and processing system based on correlation classification. Issue 1 (1st June 2022)

• Record Type: Journal Article
• Authors: Bin, Dongmei; Yu, Tong; Ling, Ying; Chen, Wendi; Lu, Liyu; Chen, Haiguang
• Is Part Of: Journal of physics. Volume 2290:Issue 1(2022)
• Publication Date: 2022-06-01
• Subjects: Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/2290/1/012027
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: when the conventional system classifies the power distribution data, the data category is not fine enough, resulting in low accuracy of abnormal data identification, high false detection rate and large correction error. A power distribution abnormal data identification and processing system based on correlation classification is proposed. In terms of hardware, multi-layer framework is adopted, and the overall hardware structure of the system is composed of communication modules, so as to optimize the power distribution data and signal acquisition circuit and suppress signal noise; In terms of software, set the data classification convergence conditions, iteratively increase the number of distribution data clustering centers until the convergence conditions are met, realize the fine correlation classification of data, draw the classified data curve, obtain the normal data feasible region, judge that there are abnormalities in the data beyond the feasible region, and use the vertical comparison method for correction. The results show that the designed system improves the accuracy of abnormal data identification, reduces the false detection rate and correction error, and the accuracy of abnormal data identification and processing is better than that of the conventional system.