[Purposes] To improve the passing efficiency of emergency vehicles and reduce their negative impact on social vehicles, this paper proposes an emergency vehicle priority control model considering the dedicated lanes of connected and automated vehicles（CAV dedicated lanes）. [Methods] Firstly, the study focuses on the basic section of a multi-lane urban expressway equipped with a CAV dedicated lane. The CAV dedicated lane was temporarily opened as an emergency lane, and the concept of a dynamic avoidance zone for emergency vehicles was introduced. Secondly, the avoidance priority of connected and autonomous vehicles（CAV）within the avoidance zone was considered. At the same time, the objective was to minimize emergency vehicle delays and reduce the negative impact on social vehicles. Specifically, the longitudinal speed and lateral lane-changing of CAVs within the dynamic avoidance zone were used as decision variables. A mixed integer nonlinear programming model based on multi-objective optimization was constructed, incorporating constraints such as vehicle dynamics, car-following, and lane-changing safety distances. Finally, using SUMO (Simulation of Urban Mobility), a joint simulation experiment is conducted on a 1.2 km one-way, three-lane urban expressway section to verify the advancement and effectiveness of the proposed model. [Findings] The simulation results show that, compared to no optimization scheme, the optimization model can reduce the affected time of emergency vehicles by an average of 96.05%, reduce emergency vehicle delay by an average of 93.12%, and increase emergency vehicle speed by an average of 19.73%. Meanwhile, the model has no negative impact on the passing efficiency of social vehicles. Additionally, during the process of assisting one emergency vehicle with priority passage, the average number of vehicles changing lanes to yield on the dedicated lane is 6.31. [Conclusions]The proposed control model significantly improves the efficiency of emergency vehicles passing under different traffic volume and CAV penetration rate scenarios, with no impact on the passing of social vehicles. It provides a new solution for emergency vehicle priority passing in mixed traffic flow scenarios.