A Sensor Based Navigation Algorithm for a Mobile Robot using the DVFF Approach

Often autonomous mobile robots operate in environment for which prior maps are incomplete orinaccurate. They require the safe execution for a collision free motion to a goal position. This paper addresses acomplete navigation method for a mobile robot that moves in unknown environment. Thus, a novel method calledDVFF combining the Virtual Force Field (VFF) obstacle avoidance approach and global path planning based onD* algorithm is proposed. While D* generates global path information towards a goal position, the VFF localcontroller generates the admissible trajectories that ensure safe robot motion. Results and analysis from a batteryof experiments with this new method implemented on a ATRV2 mobile robot are shown