Abstract: By analyzing the charging characteristics of lithium batteries, it is found that there will be significant power attenuation in the later stage of electric vehicle charging. Ignoring this phenomenon will lead to incorrect estimation of the charging time required for the electric vehicles. In order to reduce the negative impacts caused by power attenuation, this paper builds the optimal charging curve for the electric vehicles based on the optimal charging current theory. According to the charging characteristics of different electric vehicles, a two-stage optimal scheduling strategy based on multivariate hybrid optimization algorithm is proposed under the condition of power attenuation. First, based on the correlation between the electric vehicle charging currents, voltages, and powers, an electric vehicle user model is constructed, and a load aggregator model is also built based on the electric vehicle charging characteristics in different queues. Then, the optimization algorithm based on the highest response ratio is used to solve the user's charging urgency and fairness in the first stage of scheduling; and the particle swarm optimization algorithm is used to solves the minimization of the amount of photovoltaic energy consumption deviation in the second stage, in order to achieve the two-stage optimized scheduling of electric vehicles considering power attenuation characteristics. Finally, in the simulation experiment, by comparing the planned and actual charging behaviors of electric vehicles without considering the power attenuation, the impacts of charging power attenuation on the electric vehicle users and the load aggregators are analyzed. The car charging behaviors are compared to verify that the strategy in this paper has obvious effects in reducing the deviation of photovoltaic consumption, improving the users' satisfaction, and increasing the economic benefits of load aggregators.