In view of the complex low-altitude airspace environment and the fact that unmanned aerial vehicle（UAV） operation risks cannot be quantified， a low-altitude UAV operation risk assessment method considering the uncertainty in the operation process is proposed. According to the force model during the descent of the UAV， the area affected by the sudden failure of the UAV is deduced. Considering the uncertainty in the operation of the UAV， the Monte-Carlo method is used to determine the UAV’s fall point distribution， the impact area of the fall is calculated， and the system risk value is obtained. Through case analysis， the validity and rationality of the proposed risk assessment method are verified， which can provide real-time safety assessment for low-altitude UAV air traffic management， and guarantee the UAV operates safely in the low-altitude airspace.

The lack of autonomous decision-making ability of UAV has always been the main factor restricting UAV to enter the field of air combat. In a large number of existing research， action store combined with machine learning algorithm is mainly used to realize UAV air combat maneuver decision. However， the quality of UAV training will greatly affect the effect of maneuver decision. Using heuristic algorithm to construct the attack path of UAV lateral maneuver can get rid of the dependence on training method， and has strong expansibility in the future， which is easy to optimize. Based on the A-star algorithm， a path planning algorithm for UAV lateral maneuver decision-making in medium range air combat is proposed. The algorithm can plan position and speed of UAV. Through the formulation of the two aircraft cooperation strategy， the target allocation algorithm in the two UAV is designed， which can expand the UAV single aircraft confrontation to the scenario of two aircraft confrontation. Through the simulation experiment， the feasibility of the algorithm in achieving tactical effect and decision efficiency is verified.