Design of Dry-type Transformer Online Monitoring and Fault Diagnosis System

Abstract

Dry-type transformer is prone to failures during operation, so online monitoring of dry-type transformers is of great significance. This paper proposes a dry-type transformer fault diagnosis method based on radial basis function (RBF) neural network, designs and develops a dry-type transformer on-line monitoring and fault diagnosis system. The system can perform real-time monitoring and over-threshold warning on the characteristics of dry-type transformer such as temperature, partial discharge, vibration, and can diagnose faults through supervised learning based on RBF neural network. The results show that this method can effectively identify the fault type of dry-type transformer and provide a valuable reference to ensure the normal operation of dry-type transformers.