Knowledge representation in electrical insulation diagnosis

Author

Ning, Gao ; Li, Yang ; Zhang, Yan ; Deheng, Zhu

Author_Institution

Dept. of Electr. Eng., Tsinghua Univ., Beijing, China

Volume

1

fYear

1998

fDate

18-21 Aug 1998

Firstpage

96

Abstract

An artificial neural network (ANN) representation method is presented and discussed in this paper. ANN representation of symbolic and numeric knowledge is realized by the improved backpropagation (BP) algorithm. The method of management and organization of ANN representation is presented also. The problem that numeric knowledge is difficult to be represented is effectively solved by the network. After learning, the threshold values and weights of ANN are ensured and the diagnostic knowledge is learned. The practical examples are used to check the diagnostic results. It is shown that the ANN representation method is effective

Keywords

automatic test software; backpropagation; electric breakdown; fault diagnosis; insulation testing; knowledge representation; neural nets; artificial neural network; electrical insulation diagnosis; improved backpropagation algorithm; insulation testing automation; knowledge representation; learning; numeric knowledge; symbolic knowledge; threshold values; weights; Accidents; Artificial neural networks; Dielectrics and electrical insulation; Fault diagnosis; Intelligent networks; Knowledge management; Knowledge representation; Power system faults; Power system reliability; Power system security;

fLanguage

English

Publisher

ieee

Conference_Titel

Power System Technology, 1998. Proceedings. POWERCON '98. 1998 International Conference on

Conference_Location

Beijing

Print_ISBN

0-7803-4754-4

Type

conf

DOI

10.1109/ICPST.1998.728932

Filename

728932