Optimisation bi-objectif explicite d'un multiplicateur magnétique pour l'éolien

This article deals with a bi-objective mass-efficiency optimization approach of a magnetic gear for high power wind turbine applications (a few MW). This optimization evaluates the magnetic performances of the magnetic gear from a magnetostatic analytical model based on Laplace's and Poisson's equation resolutions. This model permits to determine the magnetic torque, eddy current losses in magnets, iron losses in yokes and also iron losses in pole pieces by coupling this model with a permeance network. Finally, a particle swarm optimization algorithm (PSO) is implemented to optimize mass of active parts and efficiency of a 6 MW magnetic gear.