An intelligent scheme for categorising fault events in compensated power network using K-nearest neighbour technique
by Sunil Kumar Singh; D.N. Vishwakarma; R.K. Saket
International Journal of Power and Energy Conversion (IJPEC), Vol. 11, No. 4, 2020

Abstract: This paper describes a machine learning-based intelligent scheme for categorisation of fault events in series compensated power network using K-nearest neighbour technique. The wavelet decomposition mechanism is applied on 3-phase post fault current signal for digging out the significant features of the shunt fault events in this proposed scheme. Afterward, K-nearest neighbour-based classifier model has been utilised for ascertaining the specific category of the fault events in the network. The category of the shunt events in the network has been ascertained on the basis of the estimated entropy of the wavelet coefficients. The feasibility and strength of proposed scheme for categorising the fault events in the compensated power network are assessed for different fault events with varying network situations in simulated compensated power network. Finally, it is observed that the proposed discrete wavelet transform and K-nearest neighbour-based scheme is very effectual in identifying the specific fault events in the compensated power network and is unaffected by changing system circumstances.

Online publication date: Thu, 01-Oct-2020

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Power and Energy Conversion (IJPEC):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com