Optimal Online Path Planning with Multi-Obstacles Avoidance for Autonomous Fixed-Wing Unmanned Aerial Vehicles

Abstract

Fixed-Wing Unmanned Aerial Vehicles (UAVs) are popular in military, commercial and civilian fields featuring its higher flight speed and lower consumption of energy compared with quadrotors. Apart from the speed feature, flexibility and intelligence play important roles as well. In order to obtain a safe and short flight path for a given task, an intelligent flight path planning algorithm is required. In this paper, a fast optimal flight path planning algorithm for fixed-wing UAVs, based on a refined Dubins Path geometry, is proposed. To ensure the safety feature, this algorithm consists the ability of collision avoidance for known obstacles in a limited sight, with given safe radius or curvature. The avoidance ability is guaranteed by a fast collision points generation method, which allows the UAVs to avoid collisions by flying a Dubins Path with the given safe radius or curvature. The input contains longitude, latitude, heading angles, turning radius and rotation direction. Among the input, the rotation direction is a binary/integer variable. The nonlinear constraints are implemented to evaluate the Euclidean distance violation between path and obstacles. Thus, a single objective, mixed-binary, nonlinear optimization problem is posed so as to find out a short and safe flight path among a plenty of possibilities.