FUZZY LOGIC BASED ELECTROTHERMAL LIFE MODEL FOR INDUCTION MOTOR INSULATION UNDER NON-SINUSOIDAL VOLTAGE AND CURRENT WAVEFORMS

Abstract

Power electronic voltage regulators used for speed control of induction motors give rise to non-sinusoidal voltages and currents. This leads to increase in voltage and thermal stresses which result into accelerated insulation aging and premature failure of the motors. With increase in number of stresses the life models that help in predicting the capability of insulation become complex and ambiguous. In this paper an electrothermal life model is derived using fuzzy logic to investigate the synergic effects of voltage and thermal stresses on intrinsic aging of induction motor insulation. Three parameters, voltage stress factor, waveform slope stress factor and thermal stress factor are proposed to describe the insulation stresses due to non-sinusoidal waveforms. They are computed from the experimental results and used in fuzzy logic based life estimation algorithms. Accelerated aging test is also performed on solid insulating material samples with the same non-sinusoidal voltages. Insulation life computed by the fuzzy expert system is in close agreement with the results of the accelerated aging test. An electrothermal life model is derived from the fuzzy logic results which can be directly used for the life estimation of any single phase induction motor insulation under non-sinusoidal voltage and current waveforms.