Design optimisation of electric motors by genetic algorithms

A new approach to the design optimisation of three-phase electric motors is presented. The novelty lies in combining a motor analysis procedure with a genetic algorithm to optimise an objective function such as the motor torque, the efficiency, the material cost or some other motor figures or a combination of these. Two approaches to the motor analysis are presented: the first is based on a fully analytical procedure, while the second utilises the finite element method. Some optimised design examples, with reference to surface-mounted permanent-magnet three-phase motors, are included, and a comparison between the genetic algorithm optimisation technique and the classical hill-climbing direct-search method is made. A discussion on the limitations and benefits of the proposed design approach concludes the paper.