Abstract: Due to the complexity of wind power station reactive power optimization and the randomness of wind power generation, the reactive power optimization of wind power station is no longer a question with certain solution in conventional sense, and it is difficult to obtain economical and more reliable solution by traditional method. Considering the input and output power and reactive power as random variables, a reactive power optimization mathematic model for wind power station had been built based on the chanceconstrained programming.In this model the relevant constraint conditions were described in probability mode, the constraint conditions were considered which include the terminal voltage limit of wind power generator, the limit of output active and reactive power of generator, the line tidal flow equation, the output power limit of load nodes and others, and the hybrid intelligent algorithm was used to find the solution of the question. It is proved by an example of a system with three nodes that the model and the algorithm are effective.

Key words: wind power generation, chance-constrained programming, reactive power optimization, hybrid intelligent algorithm