In recent years, the Chinese government is making great efforts to develop electric vehicle (EV) technologies to reduce carbon emissions and to protect the environment. A large number of EVs and facilities, such as charging stations (CSs), battery swapping stations (BSSs), and central charging stations (CCSs), have been deployed. However, a huge number of EVs connecting into the power grid may cause problems, such as voltage fluctuation etc. Meanwhile, the peak demand for charging will also lead to an increase in the cost of deployment of charging infrastructure. A price-based demand response (DR) program can be used to achieve goals, such as cost savings and reduction of the peak demand for charging. But to maximize the charging service capacity while minimizing the total cost is a challenging task as it is usually a tripartite game between the government, EV owners, and EV battery service providers. We considered it as a multi-objective optimization problem. First, a battery-swapping cost model is designed to describe the objective problem. It includes a battery demand model, a DR-based subsidy cost model and a charging cost model. Then, we leverage a covariance matrix adaption evolution strategy (CMAES)-based algorithm to deal with the multi-objective high dimension optimization problem and achieve the goal of maximizing the charging service capacity while minimizing the total cost. The experimental results confirm that the peak demand for battery swapping service can be reduced effectively and the total cost can be reduced by nearly 12% with our optimized subsidy strategy. Furthermore, the standard deviation evaluation results confirm the stability and the effectiveness of the proposed algorithm.INDEX TERMS Electric vehicles, costing model of battery swapping mode, demand-side management, genetic algorithm, covariance matrix adaption evolution strategy.