Application of artificial neural network for modelling of discharge inception voltage

Author

Ghosh, Saradindu ; Kishore, N.K.

Author_Institution

Dept. of Electr. Eng., Indian Inst. of Technol., Kharagpur, India

Volume

2

fYear

1997

fDate

19-22, Oct 1997

Firstpage

508

Abstract

The present work attempts to apply artificial neural networks (ANNs) with supervised learning for modelling of discharge inception voltage and stress based on different void parameters. The void depth and gas pressure are the prime considerations of this model. The requisite training data are obtained from experimental studies, published in the literature. Detailed studies are carried out to determine the ANN parameters which give the best results. The results obtained from the ANN are found to be correct within a few % indicating its effectiveness as an efficient tool in estimation

Keywords

backpropagation; feedforward neural nets; insulation testing; partial discharges; power engineering computing; voids (solid); ANNs; HV power apparatus; artificial neural networks; backpropagation learning; function estimation; gas pressure; insulation diagnostics; multilayer feedforward network; partial discharge inception voltage modelling; stress modelling; supervised learning; training data; void depth; void parameters; Area measurement; Artificial neural networks; Convergence; Dielectrics and electrical insulation; Manufacturing; Mathematical model; Neurons; Nonlinear equations; Stress; Voltage;

fLanguage

English

Publisher

ieee

Conference_Titel

Electrical Insulation and Dielectric Phenomena, 1997. IEEE 1997 Annual Report., Conference on

Conference_Location

Minneapolis, MN

Print_ISBN

0-7803-3851-0

Type

conf

DOI

10.1109/CEIDP.1997.641122

Filename

641122