Evaluation Method for the Health Status of Power Transformers Based on the IPSO-SVM

To accurately evaluate the health status of power transformers and address limitations in traditional methods such as high dependence on expert judgment and difficulties in processing multi-source heterogeneous data, this study proposes a transformer condition evaluation method based on an improved particle swarm optimization support vector machine (IPSO-SVM). First, a state evaluation index system for six component subsystems is established, incorporating time decay factors, variable weight mechanisms, and expert scoring to standardize state quantity processing. Subsequently, subsystem health indices are predicted using improved particle swarm optimized support vector regression (IPSO-SVR), and an SVM classification model enhanced by particle swarm optimization is developed for comprehensive state evaluation. Experimental results demonstrate that the optimized model achieves superior accuracy compared to traditional approaches, as measured by mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination ($\mathbf{R}^{\mathbf{2}}$). In conclusion, the proposed model provides a more precise characterization of power transformer health status and offers robust technical support for ensuring the safe and stable operation of power transformers.