Including wind power electric power system dispatching stroke has certain deviation between actual output and its planned output. Plan high output is called reserve capacity of power system, so as to produce overestimate the cost of deviation; Planning output is too low to abandon the wind will appear, which produce underestimate the cost of deviation; The planned output of wind power costs and overestimate underestimate the cost included in the objective function of the deviation, With the output of the thermal power unit, the wind power project as the control variable, the actual output of the wind power is a random variable, conditional Value-at-Risk (CVaR) theory which deals with the random issues was introduced to construct a economic dispatch model of power systems which the objective function contains the risk value of wind power output deviation conditions. The analytic expression of density function risk is difficult to obtain, so the analytic method and Monte Carlo simulation are combined to resolve the economic dispatch model and make it easy to calculate. The simulation result of IEEE30-bus system with wind power generation show that the method can effectively solve the problem of economic dispatch of wind power system; Compared with the traditional scheduling model, the total cost is lower.