EFFICIENCY ESTIMATION OF INDUCTION MOTORS AT DIFFERENT SIZES WITH ARTIFICIAL NEURAL NETWORKS AND LINEAR ESTIMATION USING CATALOG VALUES

Abstract

Induction motors are the most preferred engines in industry because of their simple but robust structure. The efficiency of the preferred motor is crucial for the limitations of the loads to be pulled by the locomotive and the locomotive’s suitability for the geographic conditions. For this reason, determining the energy efficiency and operating conditions of induction motors is crucial. It is often not possible to realize the efficiency of induction motors experimentally because it is necessary for the motor to be stopped during the experiment. This prevents the analysis of the efffects of the experiment on the energy efficiency of the motor. The efficiency estimation of induction motors provides a significant contribution to operation and energy efficiency. There is a variety of studies in literature related to efficiency estimation. However, the future of this study is the realization of efficiency estimations of induction motors at 17 different power types with artificial neural networks and linear estimation by looking at the speed values, current and moment, listed in the manufacture’s catalog in full load. Before obtaining the estimations, the statistical analysis of the correlations between efficiency and moment, efficiency and speed, efficiency and current of the motor were applied.