Methodology of Searching the Best AE Sensor Array to localize PD sources in Power Transformer

Abstract

The partial discharge (PD) is the main cause of the insulation failure in power transformers. Localization of the partial discharge source in the transformer effectively and precisely has a very important meaning in the insulation maintenance. The acoustic emission(AE) method has attracted researchers' attention because AE sensors have excellent ability to resist electromagnetic interference and they are convenient to be installed. Many researchers use an array of multiple AE sensors to collect PD signals and use the time difference between the signals to localize the PD source. However, if the sensor arrangement is unreasonable, it will inevitably introduce a huge error in time difference estimation. In this paper, the propagation characteristics of AE signals and the basic principles of time-difference-of-arrival (TDOA) methods are briefly introduced. After that, this paper describes how to set up a transformer physical model proposes a method of optimizing the arrangement of ultrasonic sensor arrays to better localize the partial discharge power in transformers. The algorithm is used to optimize the sensor array on two planes of a 35kV transformer.