Generation of optimum pulse width modulation based on artificial neural network

Abstract

This paper presents a novel technique for generating an optimum pulse width modulation (OPWM) based on artificial neural network (ANN) modeling. The nonlinear relationships between the switching patterns and the per-unite fundamental voltage component are modeled using ANN. An accurate modeling of the nonlinear relationships is achieved which allows for accurate results during on-line implementation. The paper outlines the method for generating the required OPWM switching patterns for controlling the switching devices of a three-phase inverter.