Performance optimization of switched reluctance generator in wind systems an approach based on design of computer experiment

This paper presents a proposal to optimize the performance of switched reluctance generator in a variable wind energy conversion system. An approach based on design of computational experiment is applied to determine the optimal firing angles and the optimal DC link voltage that guarantee the best system behavior for each rotor speed. Third order response surface model based on space-filling designs are applied to build a multiobjective function considering efficiency improvement and torque ripple reduction. Interior point method is applied to find the minimum of the surface, optimizing the process. Direct power control based on hysteresis current control is used to obtain the maximum power as function of the rotor speed, employing the optimal parameters. Simulation results are presented to verify the proposed methodology.