Vous êtes ici : GIPSA-lab > Formation > Thèses soutenues
MURGAN Irina Alexandra

’Ultrasound signal processing contributions for instantaneous transitory flow measurements’


Directeur de thèse :     Cornel IOANA

École doctorale : Electronique, electrotechnique, automatique, traitement du signal (EEATS)

Spécialité : Signal, image, parole, télécoms

Structure de rattachement : Grenoble-INP

Établissement d'origine : INPG - ENSE3

Financement(s) : Contrat doctoral


Date d'entrée en thèse : 01/10/2014

Date de soutenance : 23/11/2017


Composition du jury :
Mme. Marie CHABERT, Professeur, INP-ENSEEIHT Toulouse, France (Rapporteur)
M. Romeo SUSAN-RESIGA, Professeur, Université Polytechnique Timisoara, Roumanie (Rapporteur)
Mme. Henda DJERIDI, Professeur, Grenoble INP, France (Membre)
M. Alexandru ȘERBĂNESCU, Professeur, Académie Technique Militaire de Bucarest, (Membre)
M. Ljubisa STANKOVIC, Professeur, Université de Monténégro (Membre)
M. Cornel IOANA, Maitre de conférences, Grenoble INP, France (Directeur de thèse)
Mme. Diana Maria BUCUR, Professeur, Université « Politehnica » de Bucarest (Membre invité)
M. Stéphane BARRE, Chargé de recherche CNRS, LEGI, France (Membre invité)
M. Gabriel VASILE, Chargé de recherche CNRS,Gipsa-Lab, France (Membre invité)


Résumé : We propose ultrasonic signal processing methods in order to improve the transitory flow velocity non-intrusive detection through pipes, in complex measurement conditions. By complex measurement conditions, we refer to high or very low flow rates and also to transitory or turbulent flows. Usually, the flow velocity can be non-intrusive estimated, using ultrasonic flow meters based on transit time estimation. Conventional transit time flowmeters are based on the alternating emission of single-frequency acoustic pulses (ie. narrow-band acoustic pulses) and the calculation of the absolute difference between flight times in the direction of flow (direct) and in the opposite direction (reverse). The fluid velocity (and the flow rate), or rather the precision of estimation of these quantities, rest mainly on the estimation of this difference. The sensitive part of this technique is the choice of the threshold (assuming that the received signal is not affected by other phenomena such as echoes, excessive attenuation or Doppler effects) determined mainly empirically: above 50% or 80% of the maximum expected value of the signal. Techniques for reducing measurement errors are quite conceivable and provide acceptable accuracy under almost ideal measurement conditions. However, apart from the case with idealized measurement conditions, there are several scenarios where current techniques are deficient: sensor misalignment, excessive flow velocity which leads to the “flow blow effect, two-phase flow and / or the presence of the Doppler effect. These facts, presented in the second chapter of the manuscript, led us to consider, within the framework of this thesis, research axes whose common objective are to provide the signal processing tools capable of lifting the operational locks. Thus, the signal processing principles considered to meet this objective are: the principle of wide-band signals which gives the signal processing system a finer resolution and better robustness to disturbances; the concept of compressing sensing in order to reconstruct the missing samples due to interference at the reception point; the principle of beamforming and the multi-sensor principle which makes it possible to evaluate the velocity profile in a pipe section.

GIPSA-lab, 11 rue des Mathématiques, Grenoble Campus BP46, F-38402 SAINT MARTIN D'HERES CEDEX - 33 (0)4 76 82 71 31