Fusion of Odometry with Magnetic Sensors Using Kalman Filters and Augmented System Models for Mobile Robot Navigation

Abstract

Abstract: This paper presents a comparative study of two data fusion methods for high precision mobile robot’s pose estimation. Odometric data, provided by wheels encoders, are fused with data from magnetic markers detection. One of the methods uses an extended Kalman filter and the other uses a linear Kalman filter whose application is made possible by using an augmented state system model. The measurement system is composed by wheel encoders and two magnetic sensing rulers, one on the front and the other on the rear of the mobile robot, for magnetic markers detection. Simulation results with a very realistic approach are presented.1 I.