Cascading Control Based on Intelligent Algorithms for a Wind Turbine Equipped with a Doubly-Fed Induction Generator

This paper presents an intelligent cascaded nonlinear control of a Doubly-Fed Induction Generator, based on variable speed wind turbine. The whole system is presented in d-q synchronous reference frame. The main objectives of the controller defined in the partial load region, are optimizing wind energy captured, improving the quality of the power generated and minimizing mechanical stress in the drive train. The energy conversion is based on the proposed dual loop control structure using two introduced algorithms: the extreme learning machine, which is used to improve the system knowledge and the adaptive particle swarm optimization used to search the optimal gains of the conventional proportional integral controller, widely used in control of electrical part. The global controller is first tested for a velocity profile of the high wind turbulence. Secondly, it is compared to the conventional PI for showing its performances in terms of power maximization, sensitivity to perturbations and robustness against changes in parameters of the machine. The proposed control strategy is approved by simulation using software Matlab/Simulink.