Genetic Algorithm and Finite Element Anal- Ysis for Optimum Design of Slotted Torus Axial-flux Permanent-magnet Brushless Dc Motor

This paper presents an inside-out axial-flux permanentmagnet brushless DC motor optimized by Finite Element Analysis (FEA) and Genetic Algorithm (GA) that uses sizing equation. The double-sided slotted-stator designed TORUS motor has sinusoidal back EMF waveform and maximum power density. The GA obtained the dimensions that gave the motor its highest power density. Field analysis of the dimensions was then put through FEA, to obtain and re-optimize the motor’s characteristics. Possible design parameters were investigated via use of Commercial Vector Field Opera 14.0 software used in three-dimensional FEA simulation and of MATLAB 2010a in GA programming. Techniques such as modifying winding configuration and skewing the permanent magnets were explored to achieve the most-sinusoidal back-EMF waveform and minimized cogging torque. The desired technical specifications were matched by simulation results of the 3D FEA and the GA. The FEA and the GA simulation results comparison of the flux density in different parts of the designed motor at no-load condition agreed well.