KNN based fault diagnosis system for induction motor

The paper deals with an online fault diagnosis system of 3-Φ induction motor with sequence component analysis. The fault diagnosis system uses only time synchronized three phase stator voltage and current samples, from which the positive and negative sequence components have been calculated using Sample Shifting Technique (SST). With the objectives to detect the type of fault, fault severity and faulty phase using sequence components analysis. The computational technique like K-nearest neighbor algorithm has been utilized to enhance the accuracy in diagnosis of faulty phase and the severity of fault. The severity information will definitely help to make a machine maintenance schedule and accordingly plant shut down can be made minimum.