On-Line Diagnosis of Faulty Insulators Based on Improved ART2 Neural Network

  • Authors:

  • Hailong Zhang;Weimin Guan;Genzhi Guan

  • Affiliations:

  • Department of Electrical Engineering, Wuhan University, Wuhan, China 430072;Department of Electrical Engineering, Wuhan University, Wuhan, China 430072;Department of Electrical Engineering, Wuhan University, Wuhan, China 430072

  • Venue:
  • ISNN '08 Proceedings of the 5th international symposium on Neural Networks: Advances in Neural Networks, Part II
  • Year:
  • 2008

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, an improved ART2 neural network is applied to on-line diagnosis of faulty insulators. Deterioration of insulator is a gradual phenomenon. Thus, the cluster centers tend to drift during the process of recognizing on-line monitor signals. The drifts of cluster centers cause wrong judgments. In order to solve this problem, an initial layer is settled in layer F2. The improved ART2 neural network divides layer F2into upper-sublayer F22and lower-sublayer F21. When insulators start working for the first time, the upper-sublayer F22store initial states and typical-malfunction patterns. Furthermore, the improved ART2 structure of the network map, and detailed algorithm flowchart are also illustrated in this paper.