Multi-objective design optimization of three phase induction motor using Hooke and Jeeves method & GA

Abstract

Optimization of electrical machines is making trade-off between different objectives and is usually done for any one of the parameters, for example, cost, efficiency, torque and starting current associated with the system (i.e., single objective), or with more than one parameter (i.e., multi objective). Optimization process shall include, if any, the constraints imposed in the system. In the recent years, Genetic Algorithm (GA) has been successfully applied to the design optimization of electromagnetic devices and electrical machinery. In this paper, multi - objective design optimization of three phase Induction Motor (IM) is presented considering cost and efficiency as objectives. A design package has been developed for a three phase squirrel cage type IM having specifications: 5 hp, 4 - pole, 400 V, 50 Hz. delta connected. The motor design procedure consists of a system of non - linear equations which evaluates the various parameters associated in the design without violating the constraints imposed. Optimization has been carried out using a direct search method, Hooke and Jeeves method and also using GA. A combination of conventional optimization technique with GA (i.e., results obtained in Hooke and Jeeves method are fed as input data in GA process) has also been carried out. A comparative study and analysis of the end results obtained through computer simulations are presented in the paper.