The invention discloses a stochastic robust optimization-based energy storage allocation method of a wind field system. A scenario method, stochastic programming, interval programming and an artificial intelligence method are widely used for energy storage planning of a renewable energy-containing system at present. These methods are based on accurate renewable energy output or accurate probability distribution and have certain limitations. The method comprises the steps of describing a wind power prediction error as a random variable meeting given second-order moment and fluctuation quantity; describing an energy storage allocation problem containing the wind field system by a probability distribution function set meeting the second-order moment characteristics of the random variable by adopting a probability distribution robust joint opportunity constraint optimization model; converting the probability distribution robust joint opportunity constraint optimization model into definitive linear matrix inequalities; and finally solving the definitive linear matrix inequalities by adopting a convex optimization algorithm. According to the stochastic robust optimization-based energy storage allocation method, an economical and optimal energy storage allocation scheme meeting the safe operation requirements of a power system under the worst distribution of wind power can be obtained.