Monte Carlo uncertainty maps-based for mobile robot autonomous SLAM navigation

Abstract

This paper presents an uncertainty maps construction method of an environment by a mobile robot when executing a SLAM (Simultaneous Localization and Mapping) algorithm. The SLAM algorithm is implemented on a Extended Kalman Filter (EKF) and extracts corners (convex and concave) and lines (associated with walls) from the surrounding environment. A navigation approach directs the robot motion to the regions of the environment with the higher uncertainty. The uncertainty of a region is specified by a probability characterization computed at the corresponding representative points. These points are obtained by a Monte Carlo experiment and their probability is estimated by the sum of Gaussians method, avoiding the time-consuming map-gridding procedure. The mobile robot has a contour-following controller implemented on it to drive the robot to the uncertainty points. SLAM real time experiments within real environments are also included in this work.