Development of genetic algorithm embedded Kohonen neural network for dynamic security assessment

Author

El-Sharkawi, M.A. ; Huang, S.J.

Author_Institution

Dept. of Electr. Eng., Washington Univ., Seattle, WA, USA

fYear

1996

Firstpage

44

Lastpage

49

Abstract

A Kohonen self-organizing neural network embedded with a genetic algorithm is proposed in this paper. The genetic algorithm is embedded to initiate the Kohonen classifiers. By the proposed approach, the neural network learning performance and accuracy are greatly enhanced. In addition, the genetic algorithm can successfully avoid the neural network from being trapped in a local minimum. The proposed method is developed and tested on an electric utility system to access its dynamic security

Keywords

genetic algorithms; learning (artificial intelligence); power system analysis computing; power system security; self-organising feature maps; Kohonen classifiers initiation; Kohonen neural network; dynamic security assessment; electric utility system; embedded genetic algorithm; neural network learning performance; Clustering algorithms; Genetic algorithms; Industrial training; Neural networks; Power system dynamics; Power system security; Power systems; Software algorithms; Software testing; System testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Systems Applications to Power Systems, 1996. Proceedings, ISAP '96., International Conference on

Conference_Location

Orlando, FL

Print_ISBN

0-7803-3115-X

Type

conf

DOI

10.1109/ISAP.1996.501042

Filename

501042