Monitoring of Three-Phase Signals based on Singular-Value Decomposition

Abstract : n this paper, a new approach for the analysis ofthree-phase electrical signals is considered. While most of theexisting techniques are based on fixed transforms such as theClarke transform, this paper investigates the use of a data-drivenapproach called the Singular Value Decomposition (SVD). Ascompared to other transforms, this study shows that the SVDhas the distinct advantage of clearly separating the contributionsof the phasor configuration and signal instantaneous parameters.Under additive white Gaussian noise, this paper also describesseveral algorithms based on the SVD for signal monitoring. Thefirst algorithm can detect unbalanced systems and classify theminto two categories: system with an off-nominal subspace andsystems with ellipticity. The second algorithm can estimate theangular frequency based on the periodogram of the right singularvectors. As compared to other existing approaches, simulationsshow that the proposed techniques give a good compromisebetween computational complexity and statistical performance
Complete list of metadatas
Contributor : Vincent Choqueuse <>
Submitted on : Wednesday, May 22, 2019 - 10:01:12 PM
Last modification on : Thursday, October 17, 2019 - 12:36:56 PM



Vincent Choqueuse, Pierre Granjon, Adel Belouchrani, François Auger, Mohamed Benbouzid. Monitoring of Three-Phase Signals based on Singular-Value Decomposition. IEEE Transactions on Smart Grid, Institute of Electrical and Electronics Engineers, In press, pp.1-1. ⟨10.1109/TSG.2019.2898118⟩. ⟨hal-02137323⟩



Record views