Open circuit Fault Diagnosis using Machine Learning Classifiers

Abstract

Power electronic devices plays major role in controlling and improving the drive system. because of the robustness, good performance and solidness of an induction motor it is highly recommended for the industrial application. In order to obtain better performance from this kind of motors we use converters to control the motor with input parameters. But when we use the power electronic elements there is high chance of failure of this elements. This failure may lead to short-circuit fault, open-circuit fault and many other which may occur in DC link in converter of drive system. Short-circuit faults in converters will make big differences in every parameter and we have our normal conventional methods to deal with it. But open-circuit faults make system run at low efficiency and these faults are unable to find immediately. Neglecting these kinds of faults may lead to damage the other elements in the system. So, in this paper diagnosis models for the open-circuit faults in inverter fed induction motor using machine learning models and multilayer perceptron classifier is presented. In this model RMS currents and RMS voltages of each phase have been considered as a feature by which models have been trained and also valid simulation test results provided.