Development of Floor Mapping Mobile Robot Algorithm Using Enhanced Artificial Neuro-Based SLAM (ANBS)

Abstract

A complex and expensive system in floor mapping mobile robot platforms are the challenges in this age of technology revolution. Sensors that are equipped with the robot could be different, the complexity of the algorithm and the robot performance itself are not adequate. In this paper, we present an efficient way with an economically cost-saving mobile robot floor mapping system based on simultaneous localization and mapping (SLAM). The paper will highlight implementing a Rplidar sensor with a floor mapping mobile robot platform with the enhanced error corrections based on the Artificial Neuro-Based SLAM (ANBS) algorithm. The proposed system runs on Robot Operating Systems (ROS) and Tensor Flow programming. The experimental results showed how the different controllers can be improved by adding the ANBS algorithm which intelligently filtering the unnecessary error and produce the precise output on the map. The different controllers also can be used with this algorithm. For this research, the ANBS are tested on Hector SLAM and Gmapping SLAM where the output produced by each SLAM method is fed into the ANBS algorithm. At the end of the experiment, the ANBS improves the output result by 14.67% for Hector SLAM and 17.36% for the Gmapping SLAM and produces a precise map than ever before. In the future, there will be more SLAM method can be embedded with this ANBS algorithm.