A Modified Method for UAV Obstacle Avoidance Pathfinding Algorithm in Power Inspection Scenario
Bin Lan, Minfeng Xing, Yuanyuan Yang, Haitao Lyu, Jiang Qian, Tang Hao · 2024
Enhanced sensor technology and advanced control algorithms have expanded the use of autonomous Unmanned Aerial Vehicles (UAVs) in critical sectors such as power inspection. However, the limitations in positioning accuracy and onboard computational power constrain the robustness and versatility of UAV motion planning algorithms in outdoor environments. Therefore, we enhanced the existing generalized UAV obstacle avoidance pathfinding methods for application in electric power inspection, ensuring effective performance despite limited GPS signal quality and computational power. Initially, we delineate impassable areas by marking the non-collision space beneath obstacles at a specific height based on actual requirements. Next, environmental factors are integrated into the assessment of local target points, mitigating risks of UAVs encountering obstacles during trajectory planning. Finally, by discretizing the output B-spline trajectory with node fitness, we tightly couple the UAV's real-time position with the initiation of trajectory replanning. Experimental results confirm the method's robustness and efficiency.