The Induction motor fault can lead to financial loss in any place where it is used. Hence, an early detection and diagnosis of fault and its classification becomes most important for the smooth working of the system. In this paper the study proposes an algorithm for fault classification. The second order blind identification method (SOBI) is used for calculating the estimates and these estimates are given to the ANN for fault classification. Result indicates the input requirements are reduced for the classification of faults when 14 statistical parameters of 2^nd and 3^rd estimates of SOBI are given to the ANN

Keywords: SOBI, BSS, ANN, Fault Classification
View Full Paper