To solve the problem of low accuracy of transformer fault diagnosis

a transformer fault diagnosis method based on VSRP β-GWO-SVM was proposed. First

the very sparse random projection（VSRP） was used to reduce the dimension of the original data and to eliminate redundant features between variable information. Then

the β-chaotic map enabled grey wolf optimizer（β-GWO） was used to dynamically optimize the penalty factor and kernel parameters of support vector machine（SVM）

and a comprehensive transformer fault diagnosis model combining VSRP and β-GWO-SVM was obtained. Finally

the actual transformer fault data were input into the diagnosis model

the results were comparatively verified by those of the traditional grey wolf optimization support vector machine model（GWO-SVM） and particle swarm optimization support vector machine model（PSO-SVM）

and the diagnostic accuracy was 91.87%

82.13% and 75.10%

respectively. The results show that the method proposed in this paper can be employed to effectively improve the diagnosis performance of transformer.