A Probabilistic based Path and Perception Planning method for Unmanned Aerial Vehicles
Siyuan Xing, Bin Xian · 2024
The small quadrotor Unmanned Aerial Vehicles(UAVs) typically utilize cameras as onboard sensors for environmental perception. However, cameras come with limitations such as a limited field of view(FOV). This may result in incomplete environmental sensing and introduce issues of flight safety. Additionally, cameras are sensitive to image blur during high-speed motion and require that the quadrotors’ attitude angles and angular velocities should not exhibit large and abrupt changes. However, lots of existing trajectory planning methods do not adequately address these issues. This paper proposes a new path and perception planning strategy that takes into account the limitations of the FOV for quadrotor UAVs. The algorithm plans for the quadrotors’ continuous-time trajectory including position, linear velocity, linear acceleration, and yaw motion. Through comparisons in simulation, it has been demonstrated that the proposed algorithm can generate trajectories that are more optimal and smoother compared with other methods. Moreover, the algorithm proves effective in perceiving the environment with a limited FOV.