RESEARCH ON SELF-PROPELLED ROBOTS CONTROL APPLICATION FOR INTELLIGENT NAVIGATION BASED ON Q-LEARNING ALGORITHM

Abstract

This paper presents a study on controlling automotive robots applied in industry, civil, etc. for intelligent navigation in unknown environment on the basis of Q-Learning algorithm. The programming tool is the operating system for the robot ROS (Robot Operating System) and performs automatic intelligent navigation for the robot with the process of locating the robot in a flat environment and mapping (called SLAM-Simultaneous Localization and Mapping). Research results using ROS programming tool, in Gazebo environment. The information is updated from the map, operating environment, control position of the robot, and obstacles to calculate the trajectory for the robot in the automatic navigation system. The goal is to safely avoid the obstacles without encountering any obstacles along the way.