LFNNR: Light Follower Neural Network Robot Conducted by Machine Learning Technique

Abstract

Humans and animals have a brain and nervous system, and they process in-formation with neural networks. Trillions of neurons (nerve cells) exchanging brief electrical pulse called action potentials form neural networks. The endeavor of this paper aims to present the development of a neural network robot that mimics these biological structures to distinguish them from the squishy things in-side of animals. This robot is divided into hardware and OS (Operating System) part. In hardware, this robot has embedded systems that include an AVR microcontroller, L293D motor driver controller, OLED display, 4 LDR/Photoresistor, and DC gear motors. The AVR microcontroller working as the brain of this robot. This robot has an OS written in C programming language. This OS has a user interface that gives 3 options to select the operation mode. To train this robot a feed-forward backpropagation network is used. Some results about testing and validating the robotic system are presented and discussed.