Abstract: An online wind power prediction method is proposed, which is different from the traditional intelligence algorithm. By constructing the internal and external force functions and the energy function, the wind power prediction problem is transformed into an unconstrained optimization problem, which makes the quick, realistic and synchronouswind power prediction possible. Under the action of internal and external forces, the prediction converges towards the direction of carrying the physical contour and matching the fluctuation characteristics. With the prediction progress, the weight of the energy function is synchronously corrected, which guarantees the prediction accuracy. The method is simple and fast in operation, which overcomes such problems as being easy over fitting of the traditional algorithm, large edge prediction error, and the long-time training. Under the situation of increasing development of renewable energy, the feasibility and importance of this method applied in real-time and synchronous forecasting of wind power in systems are highlighted.