Abstract: Multi-rotor aircraft has a good application prospect in coal mine production inspection because of its advantages of simple structure, hovering and multi-directional flight. However, multi-rotor aircraft moves at a high speed, and the aircraft is easily influenced by various external factors during flying. It is difficult to establish a precise mathematical model. The design of a flight control algorithm is complicated. The existing synchronous positioning and map construction method based on laser radar is difficult to meet the real-time requirement of rapid flying of the multi-rotor aircraft. In view of the above problems, an obstacle avoidance control method of multi-rotor aircraft in coal mine by using remote virtual control technology is studied. The virtual remote control system of the multi-rotor aircraft in coal mine underground roadway is constructed. The virtual roadway model and the global navigation map are established in the virtual remote control system according to the initial information of the coal mine roadway. The known static obstacle information in the moving process of the aircraft is obtained, and the known static environment model is established. The task quantity of environment perception modeling in the moving process of the multi-rotor aircraft is reduced, and the operation efficiency of virtual remote control can be improved. In the inspection process, the multi-rotor aircraft detects dynamic obstacle information in the moving direction through sensing equipment carried by the multi-rotor aircraft. The remote control system reconstructs the dynamic obstacle information in an initial virtual roadway model in real-time according to obstacle data. The virtual environment state is updated in real-time to provide a reliable environment basis for local obstacle avoidance control of the aircraft. The remote control system uses the compound virtual force field (CVFF) obstacle avoidance control algorithm to plan the obstacle avoidance path by reading the positioning data and moving speed information of obstacles and aircraft. If the obstacle in front is detected to pose a great threat to the movement of the aircraft, the remote controller can implement remote intervention on the aircraft according to the planned obstacle avoidance path. The system realizes autonomous obstacle avoidance flight and human remote intervention control. In order to improve the perception efficiency and accuracy of aircraft to dynamic obstacles, a CVFF obstacle avoidance control algorithm is studied based on virtual force field (VFF) algorithm by introducing the influence of relative velocity between aircraft and obstacles and target points. The CVFF obstacle avoidance control algorithm is verified by simulation from two aspects of static and dynamic obstacle avoidance paths. The results show that under static conditions, compared with the VFF algorithm, the CVFF algorithm reduces the number of iterations and also shortens the trajectory length of the aircraft. Under dynamic conditions, the aircraft successfully avoids the two dynamic obstacles set in advance and successfully reaches the set target point. The effectiveness of the obstacle avoidance control method of multi-rotor aircraft using the CVFF algorithm is verified.