Abstract: In order to achieve accurate diagnosis of large transformer faults, this paper proposes a transformer fault diagnosis method based on improved particle swarm optimization(IPSO) to optimize the deep belief network(DBN). This method utilizes the IPSO algorithm to optimize the network parameters of the DBN network, with the dissolved gas content in oil as the input and the transformer fault state as the output, to construct an IPSO-DBN model. The IPSO-DBN model was used for fault diagnosis of transformers, and the results showed that the diagnostic accuracy of the IPSO-DBN model was as high as 98.57%, which was higher than the other three comparative models, verifying that the model can significantly improve the accuracy of transformer fault diagnosis.