Mobile Robot Localization with Map Building and Obstacle Avoidance for Indoor Navigation

Abstract

This paper describes mobile robot localization using laser-based maps for indoor navigation. It is shown that correlations arise between errors in the robot and map estimates, identified as fundamental to the solution of mobile robot localization. Data association method is used to update the location of existing features and to identify new features. fuzzy logic controller, with Mamdani based fuzzy inference engine, is used for obstacle avoidance. Experiments are carried out on a Pioneer 3-DX mobile robot equipped with sonar and laser range finders.