The paper deals with a Doubly Fed Induction Motor (DFIM) supplied by two PWM voltages inverters. The aims of this paper are sensorless adaptive Fuzzy-PI speed control decoupled by a vector control applied to a Doubly Fed Induction Motor using full order Extended Kalman Filter. The application of the adaptive Fuzzy-PI controller for speed control brings a very interesting solution to the problems of robustness and dynamics. In order to reduce the number of sensors used, and thus the cost of installation, Extended Kalman Filter Observer is used to estimate the rotor speed, rotor fluxes and stator currents, this observer is a unique observer which offers best possible filtering of the noise in measurement and of the system if the noise covariances are known. Simulation results of the proposed scheme show good performances.