Localization System for Autonomous Mobile Robots Using Machine Learning Methods and Omnidirectional Sonar

Abstract

In the past few years, the mobile robotics has presented a growth in the research field, due to its high application area. The robot localization and its displacement in an unknown environment are fundamental to its operation. Signal processing techniques, machine learning methods, and sensors assist in the environment mapping, localization, and displacement of the mobile robot in the environment. In this paper, five machine learning techniques used: Bayes, MLP, KNN, SVM and OPF. We have used the following evaluation metrics: accuracy, sensitivity, and specificity obtained through matrices of confusion. This paper has as certain objective the analysis of the machine learning methods for later use in an embedded system. The OPF method using Canberra distance obtained an accuracy of 98.13% and is, therefore, the most appropriate method to use on the proposed location system.