Paper Title
Application of Cepstrum and Neural Network For Induction Motor Fault Diagnosis

Abstract
Induction motors are vulnerable to many faults which results in becoming catastrophic and cause production shutdown, personal injuries and wastage of raw materials. Thus it is important to prevent the faulty conditions at the initial stages so as to avoid any type of failure in the system. This paper is dealing with the rotor bar fault of the induction motor. The possibility of occurrence of rotor bar faults is about 10 % of all total induction motor faults and is caused by the rotor winding. Condition monitoring and fault diagnosis of an induction motor is important in the production line. It can reduce the cost of maintenance and risk of unexpected failures by allowing the early detection of failures. This work documents experimental results for broken rotor bar fault detection in induction motors using cepstrum analysis and artificial neural network based approach. . It has found that a combination of cepstrum plus neural network analysis is very useful tool for fault diagnosis of induction motor. A feedforward neural network was used for rotor bar fault based on fault features extracted using cepstrum analysis. Keywords— Cepstrum Analysis, Artificial Neural Network, Induction Motor, Rotor Bar Fault.