Kalman Filter Based Indoor Mobile Robot Navigation

Md Anam Mahmud,Md Sayedul Aman,Haowen Jiang,Ahmed Abdelgawad,Kumar Yelamarthi

DOI: https://doi.org/10.1109/iceeot.2016.7755029

2016-01-01

Abstract:This paper focuses on navigation of a mobile robot in an indoor environment. Accuracy is an important issue in robot navigation. Thus, many approaches for mobile robot navigation have been proposed to improve the navigation accuracy. This paper presents work aimed to navigate a mobile robot, which uses wheeled encoder as a sensor. While using wheel encoder the robot is commanded to move along certain paths like an octagon, rectangle and its coordinates and travelled distances are monitored. The wheel encoder gives us Cartesian coordinate and orientation of the robot. The encoder accumulates error over time. Thus, a Kalman filter is proposed to minimize this error. The Kalman filter estimates the next position based on robot's action, like velocity, acceleration etc and sensor's output. The error made by both encoder and the Kalman filter is measured. Experimental results show that Kalman filter improves the accuracy.