The reactive power optimization with wind power generator in multi-scenario is discussed. Through probability statistics method, the uncertainty problem of active power output is solved and reactive power output scope is obtained based on rotor side maximum current limitation. By integrating traditional capacitor reactive power compensation, the wind power generator takes part in the reactive power optimization as a continuous adjustable reactive source. Combining differential evolution algorithm and crossover-mutation operator, the bacterial colony chemotaxis algorithm is improved and the global optimization capability of algorithm is enhanced. The reactive power optimization model is built with the objective function of the minimum loss and out-of-limit voltage punishment. The result shows that reactive power optimization with wind power generator in different scenario can reduce the system loss, improve the voltage level, and enhance the adaptability of the power distribution system under the impact of wind speed.