Investigation of Machine Learning Classifiers for Diagnostics of Induction Motors
Georgi Enchev, Nikolay Djagarov, Dimitar Tsvetanov, Julia Djagarova, Josep M. Guerrero
Abstract
Modern shipboard SCADA systems monitoring many thousands of parameters of shipboard equipment. These systems arrange parameters, monitor ranges and dynamics of their changes. The collected information provides control and diagnostics of ship systems and equipment. Induction electric drives and induction motors are the largest part of ships electrical equipment. The article proposes to use a machine learning classifier for diagnosing induction motors. Using the proposed universal model of an induction motor, it is possible to simulate any motor malfunctions: in the stator, in the rotor, in the mechanical part. A part of the diagnostics of an induction motor obtained as a result of simulations and machine learning is presented.