Navigation control in FPGA for a differential-drive mobile robot

Abstract

FPGA(Field Programmable Gates Array) allows to embed digital circuits which occupy little space in comparison to traditional deployments that take up much physical space in an integrated circuit, also, the use of FPGA allows taking advantage of capabilities like parallel processing and therefore is possible to embed bio-inspired digital circuits to enable skills like associative memory and learning. This paper describes the general concepts and procedure to deploy of a perceptron in a FPGA. The first step was the programming of a perceptron using MATLAB; the next step was the deployment of the same perceptron in FPGA. The perceptron deployed performs the learning stage and after perceptron was trained, the information from inputs is processed simultaneously and thus the perceptron provides the outputs for which was trained. The perceptron outputs were used to drive servomotors of a prototype mobile robot that has a locomotion system of two wheels differential drive.