Optimal UAV Path Planning for Tracking a Moving Ground Vehicle with a Gimbaled Camera.

摘要

This research develops a path planning algorithm that autonomously controls a UAV to provide convoy overwatch. The optimization algorithm determines the best path to y through developing a cost function that minimizes the control e ort of the UAV and the deviation from a desired slant range. A heuristic-based algorithm was developed and implemented on the autopilot to approximate the optimal solution. In ight test, the UAV successfully tracked a moving ground vehicle by continually placing the UAV's loiter point directly above the ground vehicle's current location. This method was called the follow-me' mode and provided the baseline for real-world UAV convoy overwatch. The follow-me mode resulted in a cost function value that was 113 times greater than the optimal path. Through an in-depth analysis, the heuristic-based approach reduced this ratio down to only 7.5 times greater than the optimal path. The data collected shows tremendous promise for improving autonomous UAV performance through optimal control techniques.