Comparison of Ant Colony and Differential Evolution Optimization Methods Applied to a Design of Synchronous Reluctance Machine

This paper describes the process of synchronous reluctance motor design optimization on an example of a motor with circular barriers modeled using commercial finite element software Infolytica MagNet combined with two stochastic optimization methods implemented in Matlab environment. The goal is to present a generalized approach to parametrization of motor geometry which can be used for various types of rotor geometries, to demonstrate the modular approach to automated pre-processing and post-processing of the motor model in MagNet software, and to compare the performance of two very robust and powerful stochastic optimization algorithms (Differential Evolution and Ant Colony Optimization).