Abonnement à la biblothèque: Guest
Ninth International Symposium on Turbulence and Shear Flow Phenomena
June 30 - July 3, 2015, University of Melbourne, Australia

DOI: 10.1615/TSFP9

DYNAMIC ESTIMATION OF VORTEX SHEDDING

pages 529-534
DOI: 10.1615/TSFP9.890
Get accessGet access

RÉSUMÉ

This paper considers the estimation of a flow field using only information from a single time-resolved sensor. We call this the estimation problem, and it is closely related to the feedback control problem since, if one wishes to control a flow with limited measurements available, it is important that one can estimate what is happening in other parts of the flow. Tackling this problem is important not only for feedback control purposes, but also for any situation where one wishes to estimate a flow field using limited measurements. The estimation is performed using a Kalman filter, an estimation tool which is widely used in guidance, navigation and control, as well as in signal processing. The Kalman filter uses dynamic (i.e. time-resolved) measurements, coupled with a model of a system, to estimate a system's full state. There are two parts to the study. The first part considers the estimation problem for the flow past a cylinder at low Reynolds numbers. We look at the accuracy of the estimate when the single sensor measurement is i) a single velocity sensor in the wake (whose location can vary); and ii) the lift force on the cylinder. The second part is concerned with how well the estimation problem can be performed. To do this we consider the Ginzburg-Landau equation, a well-known model system which displays many of the phenomena seen in fluid systems. This allows us to quantify the efficacy of the estimator as the measurement type and the disturbance characteristics vary.

Portail numérique Bibliothèque numérique eBooks Revues Références et comptes rendus Collections Prix et politiques d'abonnement Begell House Contactez-nous Language English 中文 Русский Português German French Spain