MICROCONTROLLER NEURAL NETWORK SOLUTION

ROTAR DAN, ANGHELUT MARIUS, ANDRIOAIA DRAGOS

UNIVERSITY OF BACĂU, ENERGETIC MECATRONIC AND COMPUTERS SCIENCE DEPARTMENT

Abstract

The paper describes a solution for the neural network implement with microcontroller. One of the difficulties of the microcontroller neural network realization is the parameter acquiring and transmission. The neural network inputs consist in analog signals from the process. A solution for input signals conversion and transmission to microcontroller is presented. This method allows the remote connection of the converted numeric signals and an economic solution for neural network implement. The experimental results are presented and discussed.

Keywords

neural network microcontroller infrared remote transmission modulation analog to digital conversion