Detection of Early Fault in Power Electronic Converters through Machine Learning and Data Mining Techniques

Abstract

The Power electronic system plays a significant role in versatile applications. The power electronic converters are largely used in energy conversion mechanisms. A fault is defined as the abnormal condition of the system that results in various consequences. The important constraints in the modelling of power electronic systems involve losses, Electromagnetic Interference (EMI) and harmonics. This includes the fault detection in the power electronic converters that includes three phase rectifier, d-dc converter and single-phase inverter. These parameter affects the overall efficiency and quality of the system. To overcome the fault in the power electronic converters, the machine learning with data mining techniques is adopted. This helps to predict the early fault and helps to increase the efficiency of the system.