Study on Transformer’s Equivalent Acoustic Model Based on the Least Square Method

Abstract

The study on the acoustic radiation model of transformer is very important for the noise radiation prediction of substation. The equivalent source method (ESM) is an approach to substitute b ody structure noise with a series of point sound sources, which is applicable to building the acoustic model of transformer. However, there are some limitations on the calculation speed when using genetic algorithm to solve equivalent source intensities of the transformer. In this paper, transformer ’ s equivalent acoustic model is calculated by least square method , and the computing speed is compared between the genetic algorithm and the least square method. The computing speed of least square method is foun d 200 times of that of the genetic algorithm. It is proved that least square method is more suitable for engineering applications than genetic algorithm.