Two-stage trajectory planning for stable image acquisition of a fixed wing uav

Abstract

We propose a two-stage trajectory planning (TSTP) method using monocular vision for keeping continuous surveillance or reconnaissance of a target, and for relaying communication between a ground vehicle and an unmanned aerial vehicle (UAV). The cost function for optimal trajectory includes the variation of the altitude and speed of the UAV to prevent sudden maneuvering. The position of the target in the image is also included in a cost function for continuous observation. A virtual image can be calculated from the relationship between the coordinates of the UAV and those of the camera. The proposed cost function can be affected to a greater degree than in generic optimization problems because it has the position of the target in the image. Therefore, TSTP is introduced to mitigate the problem induced by the image included in the cost function. TSTP consists of two stages of optimization. The stability of the UAV, and the relative distance between the target and the UAV on the horizontal plane, are optimized in the first stage. The results are used as initial guesses for the second stage to improve optimality. The second stage optimizes the stability of the UAV and the position of the target in the image. Proportional-integral-derivative (PID) controllers are employed to follow the optimized trajectory in both the lateral and longitudinal directions. Our simulation results show stable optimized trajectories, with a stable image and target tracking performance.