MLMVNN FOR PARAMETERS FAULTS DETECTION IN A DC-DC BOOST CONVERTER

I. Baldanzi, M. Catelani , L. Ciani , M. K. Kazimierczuk , A. Luchetta , S. Manetti , A. Reatti

Abstract

This paper aims to propose an effective approach in the fault diagnosis based on neural networks. In particular, a MultiLayer based on MultiValued Neuron artificial Neural Network (MLMVNN) with a complex QR-decomposition is used to identify parameters values changing (i.e. faults detection) on a Boost converter starting from voltages and currents in steady state measurements.

Keywords

neural networks, boost converter, faults detection, diagnostics

Download

IMEKO-WC-2015-TC10-232.pdf

DOI

-

IMEKO TC

TC10 - Measurement for Diagnostics, Optimization and Control

Back to the proceedings