Induction machine winding faults identification using Bacterial Foraging Optimization technique

The performance of a stochastic search algorithm, Bacterial Foraging Optimization (BFO), when used for fault identification of induction machine stator and rotor winding faults, is investigated in this paper. The proposed condition monitoring technique uses time domain terminal data in conjunction with the optimization algorithm and an induction machine model to indicate the presence of a fault and provide information about its nature and location. The proposed technique is evaluated using experimental data obtained from a 1.5 kW wound rotor three-phase induction machine. BFO is shown to be effective in identifying the type and location of the fault without the need for prior knowledge of various fault signatures.