Classification of faults in distribution transformer using machine learning. (2022)

• Record Type: Journal Article
• Authors: Sudha, B.; Praveen, L.S.; Vadde, Anusha
• Is Part Of: Materials today. Volume 58:Part 1(2022)
• Page Start: 616
• Page End: 622
• Publication Date: 2022
• Subjects: Distribution Transformer -- Machine Learning -- Short Circuit Resistance; Materials science -- Congresses -- Periodicals; 620.1
• DOI: 10.1016/j.matpr.2022.04.514
• Languages: English
• ISSNs: 2214-7853

Abstract

Abstract: The distribution transformers are the very important power equipment's that allows the high degree of electricity flow in distribution network. Distribution transformers are also called as service transformers. It will step down the distribution level voltage to the voltage level used by the consumer. The main components of distribution transformers are windings, core, and main tank, on load tap changer. The main faults of transformers occur in windings and in on load tap changer. The windings faults can be easily predicted by using its resistance values. In the present work the machine learning algorithms are adopted for finding the common short circuit faults in distribution transformers. The test dataset of short circuit resistance for 2 KVA distribution transformer is taken for machine learning algorithm. The suitable machine learning algorithm for finding the short circuit resistance in 2 KVA distribution transformer is K-Nearest Neighbor algorithm.