To enable conflict avoidance operation planning for support vehicles in complex environments, a dynamic priority classification method has been proposed. This method prioritizes conflict avoidance based on two levels of operation efficiency and risk. Additionally, a conflict avoidance operation planning model has been established, along with a two-stage algorithm for global path optimization. A low time complexity conflict detection method has also been proposed to ensure accuracy while solving large-scale complex problems quickly. Experimental results show that the proposed model can reasonably plan vehicle conflict avoidance paths, with the two-stage algorithm based on dynamic priority strategy proving to be more effective in both large and small scale scenarios. Compared to the fixed-priority-based two-stage algorithm, the proposed approach improves conflict point reduction by 7.6% and reduces the vehicle-aircraft conflict ratio by 7.5%.