Use of evolutionary techniques for multi-objective optimisation of electromotion devices

Design optimisation of electromotion devices has assumed unprecedented importance in recent decades due to stringent demands placed on electromagnetomechanical energy conversion device design in terms of criteria such as cost, size, efficiency, noise, vibration and electromagnetic compatibility. This paper demonstrates the method of developing algorithms for the practical multi-objective design optimisation of electromotion devices using evolutionary computation techniques. The approach is exemplified for a permanent magnet machine, which is subjected to optimisation with independent design variables, which include continuous and discrete values.