Implementation of a Neural-network-based Space-vector Pulse-width Modulation for a Three-phase Neutral-point Clamped High-power Factor Converter

Abstract

In this article, an artificial neural-network-based implementation of space-vector pulse-width modulation of a three-phase neutral-point clamped bidirectional converter with improved power quality is proposed. The neural-network-based controller offers the advantage of very fast implementation of the space-vector pulse-width modulation algorithm. This makes it possible to use an application-specific integrated circuit chip in place of a digital signal processor. The proposed scheme employs a three-layer feed-forward neural network, which receives the command voltage and angle information at the input and generates symmetrical pulse-width modulation waves for three phases of the converter with the help of a single timer and some simple logic circuits. The neural-network-based modulator distributes the switching states in such a way so as to balance the neutral-point voltage. The data to be used to train the network by a back-propagation algorithm are generated by simulating the conventional space-vector modulation-based converter for simulation and by experimentally running the space-vector modulation-based converter using a digital signal processor for experimentation. The performance of a neutral-point clamped bidirectional rectifier has been evaluated with the artificial neural-network-based modulator. The simulation results obtained are validated experimentally using a digital signal processor (DS1104) of dSPACE (dSpace, Germany). The results obtained show an excellent performance of the neural-network-based modulator.

Keywords:

• multilevel converters
• space-vector modulation
• artificial neural network
• digital signal processor
• dSPACE