IGBT Fault Diagnosis Method Based on Information Gain and Least Squares Support Vector Machine

This study proposes an IGBT fault diagnosis method based on Information Gain and Least Squares Support Vector Machine (LSSVM). Considering the significance of IGBTs in power systems and the severe consequences that their faults can cause, this paper employs Information Gain technology for feature selection to effectively extract fault characteristics. Moreover, an improved Chimpanzee Optimization Algorithm (ICOA) is introduced to optimize the parameters of the LSSVM model, aiming to enhance the accuracy and efficiency of fault diagnosis. Experimental results demonstrate that this method achieves a high accuracy rate and good generalization ability in IGBT fault diagnosis, verifying its practical value in enhancing the stability and reliability of power systems.