An obstacle avoidance system for mobile robotics based on the virtual force field method

Abstract

A basic task an autonomous mobile vehicle must perform in any environment is moving from its current position to a desired position. In order to do so, a capacity to avoid obstacles should be present. However, sensors introduce measurement errors that increase the uncertainty regarding the robot current state. In this work, an obstacle avoidance algorithm based on the methodology of the field of virtual forces for path planning will be implemented. This algorithm is based on a heuristic optimization of the balance between two design objectives: the fast convergence to the objective position and navigation free of collisions. The obstacle avoidance system will be tested on a mobile robot, the Pioneer 3at <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">®</sup> , using its front facing sonars for sensing its environment.