Localization and Mapping in Autonomous Mobile Robots

Abstract

Control of Mobile Robots is one of the most active fields of research nowadays. A robot must know all the time where it is and where it needs to go. A well- known problem of this area is Simultaneous Localization And Mapping (SLAM) where a robot has to map its unknown environment while keeping track of its current position at the same time. While performing SLAM, limited perception capabilities of exteroceptive sensors lead to Perceptual Aliasing where two different locations can appear to be the same to a robot. In this project, we present a new Fuzzy-Logic based solution to Perceptual Aliasing during the process of robot localization. And in the second approach, we present another solution to this problem using a novel Fuzzified implementation of Scale Invariant Feature Transform (SIFT). We presented we will present a novel solution for Perceptual Aliasing using ANFIS model. ANFIS model combines Neural Networks and Fuzzy system advantages. Lastly, we have proposed to solve Perceptual Aliasing in dynamic environment using Graph based Nearest Neighbor Search Approach. Simulations will show the performance of these algorithms.