Localization of a robot using particle filter with range and bearing information

Abstract

This paper reports a localization method based on particle filter with range and bearing information for fixed landmarks. The method consists of motion model which predicts pose of a robot, sensor model which evaluates the predicted pose, and resampling which modifies the evaluated pose. The proposed particle filter method utilizes bearing information as well as range information. The results of a simulation show trajectories for estimated robot location. Also, there is a result for comparison of performances using the proposed method and extended Kalman filter based method.