Design of induction motor for optimum efficiency and power factor using soft computing techniques

Abstract

Induction motors are widely used in industries because of their newlinerugged construction and simple operation. Owing to their relatively low cost, newlinereliability and efficiency, 80% of the electrical motors are the Three-Phase newlineSquirrel Cage Induction Motors (SCIMs). In most industries, they are the newlinemain energy consuming devices, contributing to more than 80% of electromechanical newlineenergy consumption. Large sized SCIMs operate with low newlineefficiency and poor power factor, which is the most important cause of poor newlinepower factor in industrial installations. The performance of induction motor is mainly affected by its poor newlineefficiency and power factor. The objective of the present research is to newlineprovide a design procedure using soft computing algorithms such as Genetic newlineAlgorithm (GA), Particle Swarm Optimization (PSO) and Extreme Learning newlineMachine Algorithm (ELMA) for single winding and multiple winding newlineinduction motor that will improve the efficiency and power factor. The flux in the stator core is almost constant irrespective of the newlineshaft load. Hence, the core loss is constant from no-load to full-load in a three newlinephase induction motor. An attempt is made in the present research, to control newlinethe flux level, by splitting the stator coils into two groups and arranging them into different combinations like Delta-parallel, Star-parallel, Delta-Series, newlineStar-Delta and Star-Series depending on the shaft load. In order to validate the objective of research, different stator newlinewinding combinations have been attempted on a 2.2 kW and 7.5 kW three newlinephase Squirrel Cage Induction Motor (SCIM). Using three soft computing newlinetechniques, the maximum efficiency and power factor are computed for newlinedifferent stator winding combinations. The optimal value of the power factor newlineand efficiency for each winding configuration are determined by using the newlineabove soft computing techniques. It is aimed at obtaining a flat efficiency by switching the stator newlinewinding in various combinations for different load currents.