Abstract: Grid-connected wind power will bring a series of interference to the power system. How to allocate the wind-storage capacity reasonably in the grid-connected wind-storage power generation system, so as to reduce the power fluctuation , is the key link in the design of wind -storage power generation system. The wind and storage power are modeled jointly. We adopt the adaptive genetic algorithm ,and the improved charging and discharging strategy of energy storage device. Taking the number of fans and batteries as the decision variables ,and the probability of fluctuation over limit as the objective function, considering of battery charge and discharge as constraint function, we can quickly get the capacity configuration , and a year of real-time wind power data shows that the program feasible.