Induction Motors Prediction Maintenance and Lifetime Estimation

Abstract

In this paper it was analyzed modern ways for induction motors predictive maintenance basing on electrical signals analysis. It was proposed to provide IM monitoring system, which collects motor electrical signals and recalculates motor remaining lifetime basing on its current condition. To get more accurate information, monitoring system also implements IM diagnostic algorithms described in previous works of the authors. Another main result is possibility to improve IM performance introducing modified control algorithm for faulty IM. This algorithm based on introduction developed neural network to FOC IM algorithm, and aims to compensate variable component of electrical signals which appears due to fault progress. Such approach allows one to ensure safe and reliable IM operation, however with decreased possible load level comparing to its rated value. Finally, it was developed specialized software for IM monitoring system, which allows one to control current value of the motor remaining lifetime, depending on its real conditions, and provides recommendation for motor further exploitation or maintenance basing on its real condition.