Postgraduate Conference Abstract

Power Electronics, Drives and Machines

Zakaria, Khairul P


The application of neural networks for condition monitoring in induction motor drives.


Supervisor: Paul Acarnley

Abstract
Most motor failures will give some warning that they are about to occur. This potential failure can be detected by monitoring equipment. Therefore, the aim of this project is to investigate and to establish the use of neural networks as a tool in condition monitoring for induction motor drives.

In the first stage of the project an experimental system for the investigation of faulted motor operation is being developed. Among the possible faults are broken end rings and broken bars on the rotor. A MATLAB-based computational method is going to be used to develop models of faulted operation. Finally, Artificial Neural Networks (ANN) will be used for analysis of motor waveforms and diagnosis of faults.