This paper deals with the power generation control in variable speed wind turbine. In this context, the wind energy conversion system (WECS) is equipped with a doubly fed induction generator (DFIG) and back-to-back five-level neutral-point-clamped converters in the rotor circuit. The modeling and the control of the five-level converter is presented. A vector control of the rotor side converter allows independent control of the stator active and reactive power and optimal speed tracking for maximum power capture from the wind. An adaptive neuro-fuzzy inference system is proposed as alternative of Mamdani type fuzzy controller to improve the robustness and reject any disturbance in the system. Three neuro-fuzzy controllers (NFCs) are used to control the rotational speed, and the stator active and reactive power. Another fuzzy logic system is proposed as a PI gain tuner in the DC-link voltage control loop to improve the dynamic response and robustness of the DC-link voltage control. In purpose to prove the performances of the global system, simulation was carried out in Matlab-Simulink software environment with 1.5MW DFIG-WECS.