This paper proposes a new regulator of the DC-link voltage of a variable speed wind energy conversion system based on adaptive critic designs (ACDs). A zero-order Takagi-Sugeno fuzzy rule base constitutes the core of the controller. A heuristic dynamic programming based approach is used to train the controller and enable it to provide nonlinear near-optimal control at different operating conditions. Based on the connectionist systems theory, the parameters of the neurofuzzy controller, including the membership functions, undergo training. Experimental results compare the performance of the neurofuzzy controller with a conventional PI controller. Results indicate that the neurofuzzy controller can noticeably improve the performance.

wind energy conversion systems, doubly-fed induction generators, adaptive critic networks, fuzzy logic control.