To address the issue of insufficient regulation resources and capabilities in the current active distribution networks (ADNs), a two-stage regulation method for Electric Vehicle (EV) participation in active distribution network optimization is proposed. The method considers the travel factors of EV users, and aims to fully release the regulation potential of EV on the premise of reducing the impact of regulation on the travel of owners. Focusing on the electric vehicle (EV) resources, a charging deviation index was first established to characterize the impact on EVs' charging. Then, EV aggregation considering charging deviation was carried out, followed by ADN optimal scheduling of EV aggregation, energy storage and distributed generations. Finally, the overall EV charging command is decomposed to each EV by minimizing the impact on EV users. The effectiveness of the proposed strategy was verified in the simulation of improving the IEEE33 system. Numerical experiments demonstrate that the proposed strategy effectively considers both the regulation capability of EVs and the charging demands of users. Compared to traditional methods, it reduces EV power deviation by 25% and increases consumption capacity by 47%.