Simulation of nonlinear filter based localization for indoor mobile robot

Abstract: Robot localization is a position estimation issue. Gaussian or non parametric executions ought to be considered to estimate position of robot. In this paper we use nonlinear filter approach to tackle the robot position estimation issue. The aim of this paper is to acquaint with the kalman filter and the extended kalman filter algorithm for determining the position of a mobile robot. The performance of EKF and the quality of position estimation depends on the exact a priori information of process and measurement noise covariance matrices. Taking this problem into a account and to solve mistakes show in estimations the simulation results can be drawn with better accuracy of the filters in estimating the robot's localization.

Keywords: Localization; Extended kalman filter; Mobile robot, Estimation.