Simple Global Path Planning Algorithm Using a Ray-Casting and Tracking Method

Abstract

This paper proposes a simple global path planning algorithm using a ray's feature of straight in nature with a random reflection model. The ray-casting and tracking (RCT) method is designed to solve global single-query path planning problems with fast convergence time. It is a random sampling-based algorithm that reflects rays with the maximum search length, which is a line of sight restricted only by the obstacles blocking the rays. RCT guarantees a competent path that follows an obstacle's edges like a path generated by a visibility graph (VG). We demonstrated RCT's superior performance in terms of both convergence time and path length on various environments that have their own features compared to other well-known path planning algorithms such as the A*, rapidly-exploring random trees, and VG.

http://ift.tt/2jJAcLC