Design optimization of spoke-type ferrite magnet machines by combined design of experiments and differential evolution algorithms

Abstract

In this paper, a combined design optimization method utilizing Design of Experiments (DOE) and Differential Evolution (DE) algorithms was implemented to provide practical insights in the multi-objective design optimization of a 12-slot, 8-pole, spoke-type, ferrite magnet machine, with fractional-slot concentrated windings. A robust parametric model of this type of machine is provided here. In this combined design optimization approach, the Central Composite Design (CCD) method, as one of the DOE approaches, was used to perform the variables sensitivity study, from which the significant design variables were selected and the corresponding variables ranges were defined, for the DE design optimization procedure. A Computationally Efficient-Finite Element Analysis (CE-FEA) approach was utilized to estimate the performance and characteristics of 3000 candidate designs. From the study of design objectives utilizing the pareto-set and color maps, three candidate designs with minimized losses, material cost and torque ripple were compared.