3D PATH PLANNING FOR UNMANNED AERIAL VEHICLES USING VISIBILITY LINE BASED METHOD

摘要

In path planning, visibility graph (or visibility line (VL)) method is capable of producing shortest path from a starting point to a target point in an environment with polygonal obstacles. However, the run time increases exponentially as the number of obstacles grows, causing this method ineffective for real-time path planning. A 2D path planning framework based on VL has recently been introduced to find a 2D path in an obstacle-rich environment with low run time. In this paper we propose 3D path planning algorithms based on the 2D framework. Several steps are used in the algorithms to find a 3D path. First, a local plane is generated from a local starting point to a target point. The plane is then rotated at several pre-defined angles. At each rotation, a shortest path is calculated using 2D algorithms. After rotations at all angles have been done, the shortest one is selected. Simulation results show that the proposed 3D algorithms are capable in finding paths in 3D environments and computationally efficient, thus suitable for real time application.