Induction Machines: A Novel, Model based Non-invasive Fault Detection and Diagnosis Technique

Model based fault detection and diagnosis in induction motor is gaining importance as it can take care of model and measurement uncertainties with the help of variants of Kalman Filters. A study of such a methodology and the potential to apply the same online is discussed. Mainly soft faults are considered for this work and MATLAB simulation results are presented. The data generation, filter convergence issues, hypothesis testing, generalized likelihood estimates etc. are addressed. A SIMLINK model is used for data generation and various types of faults are introduced. An extended Kalman filter using MATLAB is run to detect the changes.