The Distributed Generation (DG) is of flexibility. Its capability in supplying power is greatly affected by natural environment. A large scale of Electric Vehicles (EV) accessed to distribution network cause serious decline of power quality, therefore it is urgent to absorb the added load timely, which raises a higher demand on the reasonable generation dispatch for distributed generation. This paper establishes a multi-objective optimization model for the generation output of DG considering the access of EV, which takes the minimum fluctuation of voltage and the minimum added power network loss as the objectives to dynamically adjust DG daily output. The result shows that the model smooths the voltage fluctuation, reduces the risk of voltage violation, and meanwhile improves the economic efficiency. The standard node system is simulated as an example, and the effects of three models of considering economic optimization, single-objective optimization and no optimization are compared and analyzed.