Abstract: The uncertainties in power output of distributed generation (DG) and the randomness of charging power of electric vehicle (EV) bring certain influence on the operation benefit of distribution networks. Based on the uncertainties of distributed generation and the charging power of electric vehicle, the state probability models of the power output for distributed generation and the charging power of electric vehicles are built, based on which the multiple state operation space of system is formed, and an active power coordination model for distributed generation in distribution networks with DE and EV. In addition, the minimization of expected operation cost and expected cost of the power loss are taken as the optimization objectives. By using second\| order cone programming, the nonlinear objective function and constraint function are transformed into a linear funtions by rotated cone function. Then, the relationship between new variables is expressed as cone sets with special structure instead, which simplify the optimization model. An IEEE14 distribution system is used as example to verify the adaptability of the proposed model and the feasibility of the proposed method.