Application of Monte Carlo for Localizing a Mobile Robot using Odometry and Laser scanner Data

Most robotic missions would be impossible unless the position of the robot at every moment is known. Accurate localization plays an important role in robotic motion control. Research has shown that probability-based methods suggest most reliable solutions in this regard. Localization based on particle-filtering is among most efficient methods that provides a model by using a set of weighted samples and based on Monte Carlo’s method. The presented work describes how a ROS-based control system is used with a Pioneer 3-AT robot for indoor mapping and localization. Main purpose of this paper is to demonstrate accuracy of domestication and an increase in accuracy of localization. For experiments, one real environment has been tested. Some implementation was done in C and Python. This paper has also compared Monte Carlo’s output results to odometry demonstrating that more accurate results are provided the purpose is reduction of positioning errors by Monte Carlo’s method.

Keywords:

monte carlo , odometry , localization