Abstract: The collaborative optimization of the electric vehicle (EV) cluster and other distributed energy resources has become a great potential to reduce the operating cost of the system. Considering the demand preferences of the EV users, the grid-connected EV is first divided into three demand modes: the rated power charging, the adjustable charging and the flexible charging-discharging, and the control models are constructed respectively. Based on the random arrival of the EVs and the dynamic switching characteristics of the demand modes during the process of the EV connection, the concept of temporal flexibility of the EV cluster is proposed and then a mathematical model is established. Based on the temporal flexibility model of the EV cluster, a two-stage day-ahead and intra-day schedule strategy is adopted to realize collaborative optimization of the EV cluster and the distributed energy: the day-ahead optimization is carried out with the goal of minimizing the system operating cost based on the wind-solar output prediction and the EV demand prediction; the day-ahead scheduling is modified in the intre-day stage with the minimum cost through rolling time domain optimization in combination with the actual output of wind and solar and the actual grid-connected demand of the EV cluster. The simulation results demonstrate that the proposed strategy can minimize the operating cost of the system and that the demand differences of the EV users are taken into account simultaneously.