Considering the possible flight issues due to multiple aircrafts performing tasks in the same airspace in the future, an evaluation index system for the optimal avoidance scheme of large UAVs in mixed airspaces was constructed by studying various scenarios of possible collisions, and a hybrid airspace strategy of multi-scale analytic hierarchy process was proposed by analyzing the pros and cons of different scale methods. First, a large-scale unmanned aerial vehicle was used as a research object to establish a hybrid airspace model, which classified various birds or aircraft that have the possibility of conflict with the UAV in the mixed airspace, and considered the factors of collision risk level and the target execution task level. Secondly, the multi-scale analytic hierarchy process was used to obtain the weights of various avoidance targets in different situations, and the consistency was determined. The membership function was further introduced, and the final result was further weighted by analyzing the obtained weights. Finally, the airspace environment that large drones may encounter in the airspace is analyzed to verify whether the drone avoidance decision plan is effective. The verification results show that the determined multi-scale analytic hierarchy process can meet the establishment of the avoidance strategy in a complex airspace environment, and has a dynamic effect to meet the needs of UAVs to construct and adjust the avoidance strategy in real time. The research result provides new ideas for drones to complete avoidance strategies and safe flight under complex airspace conditions.