The mobile robot control for obstacle avoidance with an artificial neural network application

Abstract

The article presents the results of the research, the purpose of which was to test the possibility of avoiding obstacles using the artificial neural network (ANN). The ANN functioning algorithm includes receiving data from ultrasonic sensors and control signal generation (direction vector), which goes to the Arduino UNO microcontroller responsible for mobile robot motors control. The software implementation of the algorithm was performed on the Iskra Neo microcontroller. The ANN learning mechanism is based on the Rumelhart-Hinton-Williams algorithm (back propagation).