The top-oil temperature of transformer windings serves as a crucial health monitoring indicator

effectively characterizing the internal state and potential defects.Leveraging the advantages of ultrasonic sensing technology

known for its strong penetrative power and sensitivity to temperature changes

a novel inversion method combining ultrasonic sensing technology with whale optimization algorithm-optimized support vector machine(WOA-SVM)is proposed.This study focuses on a 35 kV transformer

where internal temperature rise is induced through short-circuit testing

and ultrasonic signals carrying temperature information are collected.An in-depth exploration is conducted on the characteristics of ultrasonic wave signals inside the transformer under different temperature conditions ranging from 20 ℃ to 50 ℃.By extracting and processing key acoustic signal parameters

such as peak amplitude

peak-to-peak value

root mean square

waveform factor

crest factor

and variance

accurate inversion of the top-oil temperature was successfully achieved.The research results demonstrate that this method exhibits high accuracy(97.95%)in the inversion of transformer top-oil temperature

providing a new technical avenue for transformer health monitoring.