Hostname: page-component-7b9c58cd5d-dlb68 Total loading time: 0 Render date: 2025-03-14T21:58:08.614Z Has data issue: false hasContentIssue false
  • English
  • Français

Closed-loop separation control using machine learning

Published online by Cambridge University Press:  10 April 2015

N. Gautier ,
J.-L. Aider ,
T. Duriez ,
B. R. Noack ,
M. Segond  and
M. Abel
N. Gautier*
Affiliation:
Laboratoire de Physique et Mécanique des Milieux Hétérogènes (PMMH), UMR 7636 CNRS, École Supérieure de Physique et Chimie Industrielles de la ville de Paris, 10 rue Vauquelin, 75005 Paris, France
J.-L. Aider
Affiliation:
Laboratoire de Physique et Mécanique des Milieux Hétérogènes (PMMH), UMR 7636 CNRS, École Supérieure de Physique et Chimie Industrielles de la ville de Paris, 10 rue Vauquelin, 75005 Paris, France
T. Duriez
Affiliation:
Institut PPRIME, CNRS – Université de Poitiers – ENSMA, UPR 3346, Département Fluides, Thermique, Combustion CEAT, 43 rue de l’Aérodrome, F-86036 Poitiers CEDEX, France Laboratorio de FluidoDinámica, CONICET/Universidad de Buenos Aires, Facultad de Ingeneria, Paseo Colon 850, Ciudad Autonoma de Buenos Aires, Argentina
B. R. Noack
Affiliation:
Institut PPRIME, CNRS – Université de Poitiers – ENSMA, UPR 3346, Département Fluides, Thermique, Combustion CEAT, 43 rue de l’Aérodrome, F-86036 Poitiers CEDEX, France
M. Segond
Affiliation:
Ambrosys GmbH, Albert-Einstein-Str. 1-5, D-14469 Potsdam, Germany
M. Abel
Affiliation:
Ambrosys GmbH, Albert-Einstein-Str. 1-5, D-14469 Potsdam, Germany
*
†Email address for correspondence: nclgautier.espci@gmail.com
Get access Rights & Permissions [Opens in a new window]

Abstract

We present the first closed-loop separation control experiment using a novel, model-free strategy based on genetic programming, which we call ‘machine learning control’. The goal is to reduce the recirculation zone of backward-facing step flow at $\mathit{Re}_{h}=1350$ manipulated by a slotted jet and optically sensed by online particle image velocimetry. The feedback control law is optimized with respect to a cost functional based on the recirculation area and a penalization of the actuation. This optimization is performed employing genetic programming. After 12 generations comprised of 500 individuals, the algorithm converges to a feedback law which reduces the recirculation zone by 80 %. This machine learning control is benchmarked against the best periodic forcing which excites Kelvin–Helmholtz vortices. The machine learning control yields a new actuation mechanism resonating with the low-frequency flapping mode instability. This feedback control performs similarly to periodic forcing at the design condition but outperforms periodic forcing when the Reynolds number is varied by a factor two. The current study indicates that machine learning control can effectively explore and optimize new feedback actuation mechanisms in numerous experimental applications.

JFM classification

Flow Control: Control theory Flow Control: Flow Control Wakes/Jets: Separated flows
Type
Papers
Information
Journal of Fluid Mechanics , Volume 770 , 10 May 2015 , pp. 442 - 457
Check for updates
Copyright
© 2015 Cambridge University Press 

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

Armaly, B. F., Durst, F., Pereira, J. C. F. & Schonung, B. 1983 Experimental and theoretical investigation of backward-facing step flow. J. Fluid Mech. 127, 473–496.Google Scholar
Beaudoin, J.-F. & Aider, J.-L. 2008 Drag and lift reduction of a 3D bluff body using flaps. Exp. Fluids 44 (4), 491–501.Google Scholar
Beaudoin, J.-F., Cadot, O., Aider, J.-L. & Wesfreid, J. E. 2004 Three-dimensional stationary flow over a backwards-facing step. Eur. J. Mech. 38, 147–155.Google Scholar
Beaudoin, J.-F., Cadot, O., Aider, J.-L. & Wesfreid, E. 2006 Drag reduction of a bluff body using adaptive control methods. Phys. Fluids 18, 085107.Google Scholar
Bergmann, M. & Cordier, L. 2008 Optimal control of the cylinder wake in the laminar regime by trust-region methods and pod reduced-order models. J. Comput. Phys. 227, 7813–7840.Google Scholar
Berman, G. & Wang, Z. 2007 Energy-minimizing kinematics in hovering insect flight. J. Fluid Mech. 582, 153–168.Google Scholar
Brandt, L., Sipp, D., Pralits, J. & Marquet, O. 2011 Effect of base flow variation in noise amplifiers: the flat-plate boundary layer. J. Fluid Mech. 687, 503–528.Google Scholar
Champagnat, F., Plyer, A., Le Besnerais, G., Leclaire, B., Davoust, S. & Le Sant, Y. 2011 Fast and accurate PIV computation using highly parallel iterative correlation maximization. Exp. Fluids 50, 1169–1182.Google Scholar
Chong, M. S., Perry, A. E. & Cantwell, B. J. 1990 A general classification of 3-dimensional flow fields. Phys. Fluids A 2, 765–777.Google Scholar
Chun, K. B. & Sung, H. J. 1996 Control of turbulent separated flow over a backward-facing step by local forcing. Exp. Fluids 21, 417–426.Google Scholar
Dahan, J. A., Morgans, A. S. & Lardeau, S. 2012 Feedback control for form-drag reduction on a bluff body with a blunt trailing edge. J. Fluid Mech. 704, 360–387.Google Scholar
Davoust, S., Jacquin, L. & Leclaire, B. 2012 Dynamics of $m=0$ and $m=1$ modes and of streamwise vortices in a turbulent axisymmetric mixing layer. J. Fluid Mech. 709, 408–444.Google Scholar
Duriez, T., Parezanovic, V., Laurentie, J.-C., Fourment, C., Delville, J., Bonnet, J.-P., Cordier, L., Noack, B. R., Segond, M., Abel, M. W., Gautier, N., Aider, J.-L., Raibaudo, C., Cuvier, C., Stanislas, M. & Brunton, S.2014 Closed-loop control of experimental shear layers using machine learning (invited). In 7th AIAA Flow Control Conference, Atlanta, Georgia, USA, pp. 1–16.Google Scholar
Ferreira, C. 2001 Gene expression programming: a new adaptive algorithm for solving problems. Complex Syst. 13, 87–129.Google Scholar
Fourrié, G., Keirsulck, L., Labraga, L. & Gillieron, P. 2010 Bluff-body drag reduction using a deflector. Exp. Fluids 50, 385–395.Google Scholar
Gardner, B. & Selig, M.2003 Airfoil design using a genetic algorithm and an inverse method. In 41st Aerospace Sciences Meeting and Exhibit, 6–9 January, Reno, Nevada, USA.Google Scholar
Gautier, N. & Aider, J.-L. 2013a Control of the separated flow downstream a backward-facing step using real-time visual feedback. Proc. R. Soc. Lond. A 469, 20130404.Google Scholar
Gautier, N. & Aider, J.-L. 2013b Effects of pulsed actuation upstream a backward-facing step. C. R. Méc. 342 (6–7), 382–388; Proceedings GDR2502 Controle Des Decollements.Google Scholar
Gautier, N. & Aider, J.-L. 2014a Feed-forward control of a perturbed backward-facing step flow. J. Fluid Mech. 756, 181–196.Google Scholar
Gautier, N. & Aider, J.-L. 2014b Real-time planar flow velocity measurements using an optical flow algorithm implemented on GPU. J. Vis.; doi:10.1007/s12650-014-0222-5.Google Scholar
Gillieron, P. & Kourta, A. 2010 Aerodynamic drag reduction by vertical splitter plates. Exp. Fluids 48, 1–16.Google Scholar
Harik, G. R.1997 Learning gene linkage to efficiently solve problems of bounded difficulty using genetic algorithms. PhD thesis, University of Michigan.Google Scholar
Henning, L. & King, R. 2007 Robust multivariable closed-loop control of a turbulent backward-facing step flow. J. Aircraft 44 (1).Google Scholar
Hervé, A., Sipp, D., Schmid, P. & Samuelides, M. 2012 A physics-based approach to flow control using system identification. J. Fluid Mech. 702, 26–58.Google Scholar
Hung, L., Parviz, M. & John, K. 1997 Direct numerical simulation of turbulent flow over a backward-facing step. J. Fluid Mech. 330, 349–374.Google Scholar
Joseph, P., Amandolese, X. & Aider, J. L. 2012 Drag reduction on the 25 degrees slant angle ahmed reference body using pulsed jets. Exp. Fluids 52 (5), 1169–1185.Google Scholar
Joseph, P., Amandolese, X., Edouard, C. & Aider, J.-L. 2013 Flow control using MEMS pulsed micro-jets on the Ahmed body. Exp. Fluids 54 (1), 1–12.Google Scholar
Kim, J. 2003 Control of turbulent boundary layers. Phys. Fluids 15 (5), 1093–1105.Google Scholar
Koza, J. R., Bennett, F. H. III & Stiffelman, O. 1999 Genetic Programming as a Darwinian Invention Machine, Lecture Notes in Computer Science, vol. 1598. Springer.Google Scholar
Le Besnerais, G. & Champagnat, F.2005 Dense optical flow by iterative local window registration. In ICIP (1), pp. 137–140.Google Scholar
Lee, C., Kim, J., Babcock, D. & Goodman, R. 1997 Application of neural networks to turbulence control for drag reduction. Phys. Fluids 9 (6), 1740–1747.Google Scholar
Luchtenburg, D. M.2010 Low-dimensional modelling and control of separated shear flows. PhD thesis, Berlin Institute of Technology.Google Scholar
Luchtenburg, D. M., Dirk, G., Gunther, M., Noack, B., Rudibert, K. & Gilead, T. 2009 A generalized mean-field model of the natural and high-frequency actuated flow around a high-lift configuration. J. Fluid Mech. 623, 283–316.Google Scholar
M’Closkey, R. T., King, J. M., Cortelezzi, L. & Karagozian, A. R. 2002 The actively controlled jet in cross-flow. J. Fluid Mech. 452, 325–335.Google Scholar
Milano, M. & Koumoutsakos, P. 2002 A clustering genetic algorithm for cylinder drag optimization. J. Comput. Phys. 175, 79–107.Google Scholar
Morimoto, K., Iwamoto, K., Suzuki, Y. & Kasagi, N.2002 Genetic algorithm-based optimization of feedback control scheme for wall turbulence. In Proceedings of the 3rd Symposium on Smart Control of Turbulence, pp. 107–113.Google Scholar
Noack, B. R., Morzyński, M. & Tadmor, G. 2011 Reduced-Order Modelling for Flow Control, CISM Courses and Lectures, vol. 528. Springer.Google Scholar
Parezanovic, V., Laurentie, J.-C., Duriez, T., Fourment, C., Delville, J., Bonnet, J.-P., Cordier, L., Noack, B. R., Segond, M., Abel, M., Shaqarin, T. & Brunton, S. 2014 Mixing layer manipulation experiment – from periodic forcing to machine learning closed-loop control. Flow Turbul. Combust.; Invited paper for the special issue of TSFP8 (in preparation).Google Scholar
Pastoor, M., Henning, L., Noack, B. R., King, R. & Tadmor, G. 2008 Feedback shear layer control for bluff body drag reduction. J. Fluid Mech. 608, 161–196.Google Scholar
Rechenberg, I. 1994 Evolution Strategy. Frommann-Holzboog.Google Scholar
Semeraro, O., Bagheri, S., Brandt, L. & Henningson, D. S. 2011 Transition delay in a boundary layer flow using active control. J. Fluid Mech. 677, 63–102.Google Scholar
Shah-Hosseini, H. 2009 The intelligent water drops algorithm: a nature-inspired swarm-based optimization algorithm. Intl J. Bio-inspired Comput. 1 (1–2).Google Scholar
Spazzini, P. G., Luso, G., Onorato, M., Zurlo, N. & Di Cicca, G. M. 2001 Unsteady behavior of a back-facing flow. Exp. Fluids 30, 551–561.Google Scholar
Tadmor, G., Lehmann, O., Noack, B. R., Cordier, L., Delville, J., Bonnet, J.-P. & Morzyński, M. 2010 Reduced order models for closed-loop wake control. Phil. Trans. R. Soc. Lond. A 369 (1940), 1513–1524.Google Scholar
Toivanen, J., Makinen, R. E., Periaux, J. & Cloud Cedex, F. 1999 Multidisciplinary shape optimization in aerodynamics and electromagnetics using genetic algorithms. Intl J. Numer. Meth. Fluids 30, 149–159.Google Scholar
Wahde, M. 2008 Biologically Inspired Optimization Methods: An Introduction. WIT Press.Google Scholar
Zhou, J., Adrian, R. J., Balachandar, S. & Kendall, T. M. 1999 Mechanisms for generating coherent packets of hairpin vortices. J. Fluid Mech. 387, 535–396.Google Scholar