Receding Horizon Control for MAVs with Vision-Based State and Obstacle Estimation
Richard J. Prazenica, Andrew J. Kurdila, Robert Sharpley · AIAA Guidance, Navigation, and Control Conference and Exhibit · 2007
This paper discusses a vision-based receding horizon control algorithm for enabling a micro air vehicle to fly autonomously through an urban environment. A vision-based geometry estimation algorithm is used to provide obstacle avoidance constraints on permissible MAV trajectories. A feature point tracker is used to track points in the images collected by an onboard camera. The three-dimensional positions of these points (i.e., the structure of the scene) are calculated using epipolar geometry. A mathematical learning algorithm is used to generate an adaptive obstacle map from the collection of measured points in the environment. State estimation is an important element of the closed-loop system since it is needed for the calculation of the scene structure and for the receding horizon control. Vision-based state estimation is briefly discussed in this paper. Simulation results are presented for a MAV flight through a virtual environment under the assumption of perfect state estimation. In these simulations, the MAV is able to detect obstacles in its path and fly safely to a goal location. The incorporation of vision-based state estimation into the closed-loop system will be considered in a future paper.