Improved RRT-Connect Algorithm for Urban low-altitude UAV Route Planning

Abstract

Path length and computation time in path planning seriously affect the safety and efficiency of urban low-altitude UAV flight. This paper improves RRT-connect Algorithm for path planning, the search step length, parent node selection and branch orientation selection are optimized to reduce the path length and algorithm time. Starting from the shortest parent node of the two trees, track search is carried out in the Middle Direction, the search step length is changed according to the field of view, the search step length is cut down quickly through the open field, and the search step length in narrow space is reduced, at the same time, the growth direction of the main branch of the search tree is controlled by the angle to reduce the probability of the UAV turning sharply. Simulation experiments show that the proposed algorithm has better time and space advantages and convergence than the current popular algorithm, the comparison results show that the optimized Algorithm RRT-connect is superior to the current flow Algorithms RTT-extend and Lazy-RTT in terms of the time consumption of short-time fast search and the length of path generation.