A fuzzy self-coupled PI control scheme is proposed for maximum power point tracking at low wind speeds in response to the nonlinearity and parameter uncertainty brought about by the randomness of wind speed to the wind energy conversion system (WECS). The self-coupled PI is used to accomplish basic speed tracking for optimal control of the wind turbine tip speed ratio, while the fuzzy controller is used to obtain the control parameters of the self-coupled PI at different operating points to improve the system's adaptability to wind speed variations. To verify the feasibility of the proposed scheme, a comparison with the conventional method is carried out in a simulation model of a 2?MW WECS built in Matlab/Simulink. Simulation results show that compared with conventional PI, fuzzy PI and GA-PI, the proposed method has better speed tracking, a smoother response curve and more power output.