In recent years, how to more effectively extract wind energy form wind power system has become a significantly research issue. The thesis proposes modulated perturb and observe algorithm, this algorithm can improve the disadvantages, such as power loss and longer tracking time. Only the output voltage and current of wind turbine are required in proposed modulated perturb and observe algorithm. By sensing those conditions to adjust the perturb step and perturb time, so that it not only can reduce the oscillation around the maximum power point, but also increase the system tracking speed. Finally, the system is controlled with a digital signal processor, dsPIC3F011, to realize and prove the proposed maximum power point tracking.