A Geometry Optimization of Fundamental Characteristics of a Brushless DC Motor Using FEM and Evolutionary Algorithms

This paper describes a design procedure that optimizes the three principle functions of a brushless DC (BLDC) motor. The weight, average torque and cogging torque of a 3phase, 6 slot BLDC motor with 4 surface mounted permanent magnet has been optimized by MOGA and DE. The upper and lower limits of the motor geometries are cast into the role of optimization in DE but MOGA is bounded to no constraints as far as it doesn’t conflict. Finally the paper suggests a group of geometries from MOGA that each of them has been dominated in one or some of the objective functions and compared them to DE and other references.