VTV: Real Time Obstacle Avoidance of Mobile Robots for Local Path Planning Using LMS

Abstract

In this paper, a new algorithm, called Virtual Tangential Vector (VTV), for real time mobile robot obstacle avoidance for local path planning in indoor environment is proposed and demonstrated. It is for a 2-dimensional space using a 2D laser measurement system (LMS) to detect obstacles. VTV consists of two processes. The first defines a virtual encircled obstacle from three points obtained by a LMS. The second decides movement direction combining the gravity vector to the given target position and the tangential vector obtained from the encircled obstacle. The magnitude of tangential vector is taken proportional to the square of the radius of the encircled obstacle to generate a shorter path. The VTV is tested by several cases to verify the performance of the algorithm and shown to be better when compared with a popular algorithm, the so-called virtual force field (VFF) algorithm.