In order to solve the path planning problem of quad-rotor UAV navigation system,an improved A∗ search algorithm is proposed.The algorithm integrates the idea of skipping point search to prune the map to reduce the search space;changes the search direction judgment mechanism to determine the size of the direction angle and the set threshold,and limits the search direction to reduce search nodes;introduces a collision detection mechanism to calculate the minimum distance from obstacles to the drone's flight path to improve its autonomous obstacle avoidance reliability.The feasibility of the improved A∗ algorithm is verified by comparing the improved A∗ algorithm with the traditional A∗ algorithm through data simulation.A test flight platform using a T265 binocular camera and a D435i depth camera as vision sensors was built.Experimental verification is carried out based on the constructed obstacle map,and the path planning results of the improved A∗ algorithm and the traditional A∗ algorithm are compared and analyzed.
万方数据