Design optimization of a BLDC motor: a comparative analysis

The paper presents an application of the Matlab optimization algorithms to a slotless brushless DC (BLDC) motor design, with 7 design parameters chosen to be free. The tested algorithms are: classical (gradient-based), direct-search and genetic. As a conclusion, the direct-search algorithm is not suitable for this application, as it is highly dependent on the initial values. The classical or genetic algorithms are suitable to apply, but within loops. The best results are produced by the sequence genetic-classical algorithms.