Efficiency optimization of transverse flux permanent magnet machine using genetic algorithm

Particular interest is presently focused on electric propulsion system for the reduction of air pollution and noise. Transverse flux permanent magnet machine (TFPM) is one of the most suitable designs in such application so far as its high power production potential. In this paper a modified genetic algorithm (GA) is developed to find the independent geometrical parameters of rotor and stator with the aim of reducing the TFPM's power losses, which occur in the iron and the winding. The evolutionary algorithm GA is proved to be a simple and efficient stochastic optimization method for solving optimal design problem in industry. It is found that analytical analysis of machine combined with GA offers an effective way to improve the motor's efficiency. The calculation result validates the proposed design can achieve a good prediction of TFPM performance.