Automatic control of navigation system for robot

• Main Author: Zhang, Xue
• Other Authors: Wang Han
• Format: Final Year Project
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/60592
• Institution: Nanyang Technological University

This paper will introduce the principle of autonomous navigation system. The main parts are scan matching, gmapping, frontier exploration, avoiding obstacles. The method and algorithm of scan matching specific are introduced in this paper. The mathematical model applied in field of scan matching can be four kinds. They are senor model, probabilistic model, original map model, ICP (Iterated Closest Point) method. And the SLAM (Simultaneous Localization and Mapping) is main algorithm applied in gmapping. I provide the detailed information on the building of partial map and whole map. And frontier exploration is a process of extending map. Terrain classification with different color is crux application in the frontier exploration. At last the part of avoiding obstacles is the key link in the autonomous navigation system. Kinematics equation applied in the avoiding obstacles is the crux method. And the protection distance always can effective avoid the hitting obstacles. I refer to the basic principle and method of the predecessor and turn theories into the practical. The experimental result achieves the goal of autonomous navigation and building current map.