Search

With the development of the times, UAVs are gradually promoted and popularized in military and civil fields, and the future airspace will also face more security risks. This paper combines the graph theory network algorithm to plan the initial path of UAV inspection and completes the dynamic collision avoidance path planning in the process of UAV inspection through the established joint model of UAV inspection sensing and avoidance. At the same time, the ant colony algorithm is introduced to improve the graph theory network algorithm to solve the dynamic collision avoidance optimal path in the process of UAV inspection. On this basis, simulation experiments of path planning design for UAV collision avoidance and kinematic simulation with two-fold priority judgment are carried out, and the kinematic parameters corresponding to the collision avoidance path are selected as the analysis anchor points. The extreme values of horizontal speed, climb speed, trajectory inclination angular rate, and heading angular rate are 31.6 m/s, 3 m/s, 5.7°/s, and 18.4°/s, respectively, which are within the given constraints, and verify the reasonableness and effectiveness of the proposed optimization scheme. The average time of the proposed algorithm is 0.0458s, which is much lower than the corresponding 0.7105s of the original scheme algorithm, through data comparison of simulation experiments and experimental validation. Using a graph theory network algorithm, the proposed optimization scheme is more efficient and stable and has a higher success rate for collision avoidance.