Computer simulation of a neural network controlled buck converter

Abstract

This work presents the first results obtained by the simulation of a neural network (NN) controlling a DC to DC buck converter. An approach has been chosen in which the NN observes the input disturbances, the state variables and controls the output voltage. Input voltage and output current have been considered as input disturbances, and the network has been trained on typical signals, and then tested on another set of combined disturbances. The simulation has been carried out using MATLAB.<>