An Enhancement of Simultaneous Localization and Mapping Model Using Artificial Neural Networks

Abstract

This paper presents a model of Environment Representation Architecture for Intelligent Robot. The model consists of a vehicle, an environment, and landmarks. The proposed method is based on using SLAM based on ANN, back propagation algorithm, to be trained on predefined datasets on some environment. In this paper dun Using Artificial Neural on the Simultaneous Localization and Mapping enhances the obtained maps of the robot. Limitation of this paper to test the proposed system, different maps with different datasets are required, however, these datasets, need expensive sensors, vehicles and GPS receivers to be built. Use this information to build their decisions. The systematic error was solved by the proposed approach using ANN, depends only on the initial values that were used during the training phase, it considers previous landmarks in order to build the next route, but on the other hand, it does not accumulate the previous error.