Development of an autonomous mobile robot outdoor navigation system.

摘要

Navigation is a fundamental capability that mobile robots must possess. Outdoor navigation is especially challenging due to numerous uncertainties and dynamic changes in the environment such as varying lightning conditions. A robust outdoor navigation system utilizing navigational sensors such as GPS, vision, laser, and sonar improves autonomy of robots and their task handling capabilities. This thesis focuses on the development and testing of an outdoor navigation system for mobile robots that improves autonomy. This navigation system enables a mobile robot equipped with GPS, laser and sonar range finders, and vision system to navigate from a known initial position to a given target location in an outdoor environment. Waypoints that the robot will track are created in XML format by waypoints generator web interface and are given to robot prior to starting movement. The real-time tracking web interface continuously displays the position of the robot on the map by utilizing the GPS readings. An artificial neural network was designed and developed for detecting roads. Real-time computer vision algorithms were developed for online path planning and obstacle avoidance. The developed system was also tested for road detection success rate and eighty seven (87%) percent of successful road center detection rate was accomplished. The system has been successfully implemented and tested using a Pioneer 3-AT mobile robot in outdoor environments.