Trajectory Planning for Autonomous Mine Inspection Using UAV

In this paper, we propose a trajectory planning algorithm for autonomous mine inspection using unmanned aerial vehicles (UAV). Mine inspection presents a set of serious challenges such as narrow roadways and sharp corners. To overcome the disadvantages, the safe flight corridors are calculated to guarantee the safety of the UAV and a multi-objective function is used for optimization to enable the comprehensive performance of the trajectory. A manual inspection with a laser is taken as the input to generate the initial path and point cloud map of the mine environment using the Simultaneous Localization and Mapping (SLAM) algorithm. The initial path can be transformed into a set of waypoints after b-spline smoothing and discrete sampling. The safe flight corridors composed of convex polyhedrons are built based on the waypoints. A multi-objective optimization strategy considering smoothness and time cost is used to generate the piecewise polynomial inspection trajectory and safety is ensured by limiting the trajectory inside the safe flight corridors. Experimental results demonstrate that the proposed approach can perform well in autonomous mine inspection.