A sensor based navigation algorithm for moving obstacles assuring convergence property

Masaaki Tomita, Motoji Yamamoto

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

We propose a sensor based navigation algorithm for a mobile robot that assures a convergence property when the robot is navigated from a given start position to a goal position while avoiding moving and fixed obstacles. This algorithm guides the robot toward the goal by using its sensor information in an unknown environment. Conventional sensor based navigation algorithms such as Bug algorithm and Tangent Bug algorithm do not assure the convergence property to the goal position, and they may fail because of dead-locks in the presence of moving obstacles, because they work only for fixed obstacles in the unknown environment. In the real world, a robot should reach the goal avoiding fixed and moving obstacles. Typical examples of moving obstacles include human beings in a real world workspace. We propose a new algorithm to guide a mobile robot toward the goal position that treats the navigation problem for such workspace. The basic concept of this algorithm is based on detecting a loop in the robot's path that is characteristic for the avoidance of moving obstacle. Simulation examples of sensor based navigation in the presence of moving and fixed obstacles are shown to demonstrate effectiveness of the proposed algorithm.

Original languageEnglish
Title of host publicationProceedings of the 11th IAPR Conference on Machine Vision Applications, MVA 2009
Pages295-299
Number of pages5
Publication statusPublished - Dec 1 2009
Event11th IAPR Conference on Machine Vision Applications, MVA 2009 - Yokohama, Japan
Duration: May 20 2009May 22 2009

Publication series

NameProceedings of the 11th IAPR Conference on Machine Vision Applications, MVA 2009

Other

Other11th IAPR Conference on Machine Vision Applications, MVA 2009
Country/TerritoryJapan
CityYokohama
Period5/20/095/22/09

All Science Journal Classification (ASJC) codes

  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'A sensor based navigation algorithm for moving obstacles assuring convergence property'. Together they form a unique fingerprint.

Cite this