OBSTACLE AVOIDANCE PROCEDURE FOR MOBILE ROBOTS

摘要

Abstract. The common set of procedures provided by most of indoor mobile robots consists of Mapping Building, Autolocalization, and Navigation (Path Planning and Obstacle Avoidance). Considering that real indoor environments are commonly dynamic where known obstacles (e.g. chairs, tables, sofas, etc.) and unknown obstacles (persons walking close to the robot, unmapped environment characteristics, etc.) are constantly changing their positions, it is necessary to use sensor data to provide the robot a detailed real-time environmental view. Most of these obstacles may be located on a priori planned path and the robot controller must execute a skilful manoeuvre that avoids a collision and directs the robot to its goal position. Undoubtedly, it is essential that both navigation procedures (path planning and obstacle avoidance) work together and harmoniously, trying to solve possible conflicts between their two different behaviours (go to the goal position as soon as possible and avoid potential collisions during the path). In this article we describe the development of a new obstacle avoidance procedure based on the previous Obstacle Velocity approach. This procedure was denominated as BOA and was tested in a mobile robot in wide and narrow environments. The robot has been equipped with an intelligent control and navigation system and two SICK two-dimensional (2-D) laser range finder LMS 200.