pith. sign in

arxiv: 1907.01664 · v1 · pith:IPJUGO3Znew · submitted 2019-07-02 · ⚛️ physics.flu-dyn

Control-oriented model reduction for minimizing transient energy growth in shear flows

Aniketh Kalur , Maziar S. Hemati This is my paper

Pith reviewed 2026-05-25 10:19 UTC · model grok-4.3

classification ⚛️ physics.flu-dyn
keywords model reductiontransient energy growthfeedback flow controlshear flowschannel flowproper orthogonal decompositionbalanced truncationlinear matrix inequalities
0
0 comments X

The pith

Reduced-order models via POD projection and balanced truncation enable feedback controllers that reduce maximum transient energy growth in channel flow.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that projecting system outputs onto proper orthogonal decomposition modes to capture energy, then applying balanced truncation, produces low-dimensional models that retain the input-output dynamics driving transient energy growth. These models make the linear matrix inequality problems for controller synthesis computationally feasible for high-dimensional linearized flows. On a channel flow with wall actuation, the resulting controllers achieve lower peak transient energy growth than standard linear quadratic designs. A reader would care because this non-modal energy growth mechanism is known to initiate subcritical transition to turbulence in shear flows.

Core claim

An output projection onto proper orthogonal decomposition modes followed by balanced truncation produces reduced-order models that preserve the input-output properties governing transient energy growth. Controllers designed on these models using linear matrix inequalities minimize the maximum transient energy growth when applied to the original linearized channel flow system, outperforming linear quadratic regulators.

What carries the argument

The control-oriented reduced-order model obtained by proper orthogonal decomposition output projection followed by balanced truncation, preserving input-output properties for LMI-based controller synthesis.

If this is right

  • Controller synthesis for high-dimensional linearized fluid systems becomes computationally tractable via convex LMI optimization.
  • The designed controllers reduce the maximum transient energy growth relative to conventional linear quadratic optimal control.
  • The method applies to a linearized channel flow with blowing and suction actuation at the walls.
  • The approach combines energy-capturing POD projection with input-output preserving balanced truncation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same reduction steps could be tested on other canonical shear flows to check whether tractable control design generalizes beyond channel flow.
  • Embedding the low-order controller in a nonlinear simulation would reveal whether reduced linear energy growth translates to delayed transition.
  • If model order is low enough, the method opens a route to experimental implementation where real-time feedback is required.

Load-bearing premise

The reduced-order model must preserve the input-output properties that govern transient energy growth sufficiently well for a controller synthesized on the reduced model to perform as intended on the full high-dimensional system.

What would settle it

Direct simulation showing that a controller designed on the reduced-order model produces no reduction or less reduction in maximum transient energy growth than LQR when applied to the full linearized channel flow system would falsify the central claim.

read the original abstract

A linear non-modal mechanism for transient amplification of perturbation energy is known to trigger sub-critical transition to turbulence in many shear flows. Feedback control strategies for minimizing this transient energy growth can be formulated as convex optimization problems based on linear matrix inequalities. Unfortunately, solving the requisite linear matrix inequality problem can be computationally prohibitive within the context of high-dimensional fluid flows. In this work, we investigate the utility of control-oriented reduced-order models to facilitate the design of feedback flow control strategies that minimize the maximum transient energy growth. An output projection onto proper orthogonal decomposition modes is used to faithfully capture the system energy. Subsequently, a balanced truncation is performed to reduce the state dimension, while preserving the system's input-output properties. The model reduction and control approaches are studied within the context of a linearized channel flow with blowing and suction actuation at the walls. Controller synthesis for this linearized channel flow system becomes tractable through the use of the proposed control-oriented reduced-order models. Further, the resulting controllers are found to reduce the maximum transient energy growth compared with more conventional linear quadratic optimal control strategies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes using POD-based output projection followed by balanced truncation to obtain control-oriented reduced-order models of a linearized channel flow, enabling tractable LMI-based synthesis of feedback controllers that minimize maximum transient energy growth. It claims these controllers outperform conventional LQR designs on the full-order system.

Significance. If the numerical results are robust and the reduction reliably transfers performance, the approach would provide a practical route to convex-optimization control design for high-dimensional non-normal fluid systems where direct LMI solution is intractable, potentially aiding transition-delay strategies.

major comments (2)
  1. [§3] §3 (model reduction): balanced truncation is invoked to preserve input-output properties, yet no a-priori bound is derived linking the reduced-model LMI cost or Hankel singular values to the full-order finite-time induced norm ||e^{At}|| that governs non-modal transient growth; the non-normality of the channel-flow operator means residual energy in the POD-orthogonal complement can still amplify before control acts.
  2. [Results / abstract] Results section / abstract claim: the statement that the synthesized controllers 'reduce the maximum transient energy growth compared with more conventional linear quadratic optimal control strategies' is presented without reported quantitative metrics (peak growth ratios, time horizons, Reynolds number, or comparison tables), error bounds on the closed-loop full-order response, or verification that the LMI solution on the ROM transfers without degradation.
minor comments (2)
  1. [§2] Notation for the output projection matrix and the precise definition of the energy norm used in the POD step should be stated explicitly with an equation reference.
  2. [§4] The LMI formulation for minimizing transient growth (e.g., the specific matrix inequalities involving the Lyapunov or peak-to-peak bounds) is referenced but not written out; adding the explicit LMI would aid reproducibility.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive feedback on our manuscript. We address each major comment below and indicate where revisions will be made to improve clarity and rigor.

read point-by-point responses

  1. Referee: [§3] §3 (model reduction): balanced truncation is invoked to preserve input-output properties, yet no a-priori bound is derived linking the reduced-model LMI cost or Hankel singular values to the full-order finite-time induced norm ||e^{At}|| that governs non-modal transient growth; the non-normality of the channel-flow operator means residual energy in the POD-orthogonal complement can still amplify before control acts.

    Authors: We acknowledge that the manuscript does not derive an a-priori bound relating the reduced-order LMI cost or Hankel singular values to the full-order finite-time induced norm. The method instead relies on the energy-capturing property of the POD output projection combined with balanced truncation's preservation of input-output behavior, followed by direct numerical verification of closed-loop performance on the full-order system. Due to the inherent non-normality, obtaining a rigorous a-priori guarantee remains an open theoretical question beyond the scope of this primarily computational study. We will revise §3 to explicitly note this limitation and emphasize the role of post-design full-order simulation for validation. revision: partial

  2. Referee: [Results / abstract] Results section / abstract claim: the statement that the synthesized controllers 'reduce the maximum transient energy growth compared with more conventional linear quadratic optimal control strategies' is presented without reported quantitative metrics (peak growth ratios, time horizons, Reynolds number, or comparison tables), error bounds on the closed-loop full-order response, or verification that the LMI solution on the ROM transfers without degradation.

    Authors: The results section of the manuscript contains numerical comparisons at a fixed Reynolds number (Re = 2000) showing reduced peak transient growth relative to LQR, with the controllers applied to the full-order system to confirm transfer. However, the abstract statement is indeed stated without accompanying quantitative values. We will revise the abstract to include specific metrics (e.g., peak growth reduction ratios and time horizons) and add a brief statement on full-order verification to make the claims self-contained. revision: yes

standing simulated objections not resolved
  • Derivation of a general a-priori bound on the approximation error for the finite-time induced norm under model reduction for non-normal operators

Circularity Check

0 steps flagged

No circularity detected; derivation relies on independent standard methods

full rationale

The paper's chain consists of applying POD output projection to capture energy, followed by balanced truncation to preserve input-output properties, then LMI-based controller synthesis on the reduced model, with numerical comparison to LQR on the channel-flow example. These steps invoke established techniques whose validity does not depend on the final performance numbers or on any self-referential definition. No equations or claims reduce a result to its own inputs by construction, no fitted parameters are relabeled as predictions, and no load-bearing uniqueness theorems or ansatzes are imported via self-citation. The approach is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. The method relies on established POD and balanced truncation without introducing new postulated entities.

pith-pipeline@v0.9.0 · 5716 in / 1069 out tokens · 38514 ms · 2026-05-25T10:19:17.267858+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

38 extracted references · 38 canonical work pages

  1. [1]

    J., and Henningson, D

    Schmid, P. J., and Henningson, D. S.,Stability and transition in shear flows, Springer, 2001. doi:10.1115/1.1470687

    work page doi:10.1115/1.1470687 2001

  2. [2]

    Some observations on skin friction and velocity profiles in fully developed pipe and channel flows,

    Patel, V. C., and Head, M. R., “Some observations on skin friction and velocity profiles in fully developed pipe and channel flows,”Journal of Fluid Mechanics, Vol. 38, No. 1, 1969, pp. 181–201. doi:10.1017/S0022112069000115

    work page doi:10.1017/s0022112069000115 1969

  3. [3]

    Energy growth in viscous channel flows,

    Reddy, S. C., and Henningson, D. S., “Energy growth in viscous channel flows,”Journal of Fluid Mechanics, Vol. 252, No. -1, 1993, p. 209. doi:10.1017/S0022112093003738

    work page doi:10.1017/s0022112093003738 1993

  4. [4]

    Hydrodynamic stability without eigenvalues

    Trefethen, L. N., Trefethen, A. E., Reddy, S. C., and Driscoll, T. A., “Hydrodynamic stability without eigenvalues.”Science (New York, N.Y.), Vol. 261, No. 5121, 1993, pp. 578–84. doi:10.1126/science.261.5121.578

    work page doi:10.1126/science.261.5121.578 1993

  5. [5]

    On the role of linear mechanisms in transition to turbulence,

    Henningson, D. S., and Reddy, S. C., “On the role of linear mechanisms in transition to turbulence,”Physics of Fluids, Vol. 6, No. 3, 1994, pp. 1396–1398. doi:10.1063/1.868251

    work page doi:10.1063/1.868251 1994

  6. [6]

    Nonmodal Stability Theory,

    Schmid, P. J., “Nonmodal Stability Theory,”Annual Review of Fluid Mechanics, Vol. 39, 2007, pp. 129–62. doi:10.1146/ annurev.fluid.38.050304.092139

    work page arXiv 2007

  7. [7]

    Subcritical transition in channel flows,

    Chapman, S. J., “Subcritical transition in channel flows,”Journal of Fluid Mechanics, Vol. 451, 2002, pp. 35–97. doi: 10.1017/S0022112001006255

    work page doi:10.1017/s0022112001006255 2002

  8. [8]

    Flow control: New challenges for a new Renaissance,

    Bewley, T. R., “Flow control: New challenges for a new Renaissance,”Progress in Aerospace Sciences, Vol. 37, No. 1, 2001, pp. 21–58. doi:10.1016/S0376-0421(00)00016-6

    work page doi:10.1016/s0376-0421(00)00016-6 2001

  9. [9]

    Transition delay using control theory,

    Bagheri, S., and Henningson, D. S., “Transition delay using control theory,”Philosophical Transactions: Mathematical, Physical and Engineering Sciences, Vol. 369, 2011. doi:10.1098/rsta.2010.0358. 20

    work page doi:10.1098/rsta.2010.0358 2011

  10. [10]

    A Linear Systems Approach to Flow Control,

    Kim, J., and Bewley, T. R., “A Linear Systems Approach to Flow Control,”Annual Review of Fluid Mechanics, Vol. 39, No. 1, 2007, pp. 383–417. doi:10.1146/annurev.fluid.39.050905.110153

    work page doi:10.1146/annurev 2007

  11. [11]

    Optimal and robust control and estimation of linear paths to transition,

    Bewley, T. R., and Liu, S., “Optimal and robust control and estimation of linear paths to transition,”Journal of Fluid Mechanics, Vol. 365, 2018, pp. 305–349. doi:10.1017/S0022112098001281

    work page doi:10.1017/s0022112098001281 2018

  12. [12]

    FeedbackControlofTransitionalChannelFlowusingBalancedProperOrthogonalDecomposition,

    Ilak,M.,andRowley,C.W.,“FeedbackControlofTransitionalChannelFlowusingBalancedProperOrthogonalDecomposition,” AIAA Theoretical Fluid Mechanics Conference, Seattle, WA, AIAA paper 2008-4230, 2008

    work page 2008

  13. [13]

    Linearfeedbackcontroloftransientenergygrowthandcontrol performance limitations in subcritical plane Poiseuille flow,

    Martinelli, F., Quadrio, M., Mckernan, J., andWhidborne, J.F., “Linearfeedbackcontroloftransientenergygrowthandcontrol performance limitations in subcritical plane Poiseuille flow,”Physics of Fluids, 2011. doi:10.1063/1.3540672

    work page doi:10.1063/1.3540672 2011

  14. [14]

    Linear feedback control and estimation of transition in plane channel flow,

    Hogberg, M., Bewley, T. R., and Henningson, D. S., “Linear feedback control and estimation of transition in plane channel flow,”Journal of Fluid Mechanics, Vol. 481, 2003, pp. 149–175. doi:10.1017/S0022112003003823

    work page doi:10.1017/s0022112003003823 2003

  15. [15]

    Finite Dimensional Optimal Control of Poiseuille Flow,

    Joshi, S. S., Speyer, J. L., and Kim, J., “Finite Dimensional Optimal Control of Poiseuille Flow,”Journal of Guidance, Control, and Dynamics, Vol. 22, No. 2, 1999. doi:10.2514/2.4383

    work page doi:10.2514/2.4383 1999

  16. [16]

    Performance Limitations of Observer-Based Feedback for Transient Energy Growth Suppression,

    Hemati, M. S., and Yao, H., “Performance Limitations of Observer-Based Feedback for Transient Energy Growth Suppression,” AIAA Journal, 2018. doi:10.2514/1.j056877

    work page doi:10.2514/1.j056877 2018

  17. [17]

    Revisiting the separation principle for improved transition control,

    Yao, H., and Hemati, M. S., “Revisiting the separation principle for improved transition control,”AIAA Paper 2018-3693, 2018. doi:10.2514/6.2018-3693

    work page doi:10.2514/6.2018-3693 2018

  18. [18]

    Advances in Output Feedback Control of Transient Energy Growth in a Linearized Channel Flow,

    Yao, H., and Hemati, M. S., “Advances in Output Feedback Control of Transient Energy Growth in a Linearized Channel Flow,” AIAA Paper 2019-0882, 2019. doi:10.2514/6.2019-0882

    work page doi:10.2514/6.2019-0882 2019

  19. [19]

    On the Minimization of Maximum Transient Energy Growth,

    Whidborne, J. F., and McKernan, J., “On the Minimization of Maximum Transient Energy Growth,”IEEE Transactions on Automatic Control, Vol. 52, No. 9, 2007, pp. 1762–1767. doi:10.1109/TAC.2007.900854

    work page doi:10.1109/tac.2007.900854 2007

  20. [20]

    Three-dimensional optimal perturbations in viscous shear flow,

    Butler, K. M., and Farrell, B. F., “Three-dimensional optimal perturbations in viscous shear flow,”Physics of Fluids A: Fluid Dynamics, Vol. 4, No. 8, 1992, pp. 1637–1650. doi:10.1063/1.858386

    work page doi:10.1063/1.858386 1992

  21. [21]

    Componentwise energy amplification in channel flows,

    Jovanovic, M. R., and Bameih, B., “Componentwise energy amplification in channel flows,”Journal of Fluid Mechanics, Vol. 534, 2005, p. 145–183. doi:10.1017/S0022112005004295

    work page doi:10.1017/s0022112005004295 2005

  22. [22]

    doi:10.1137/1.9781611970777

    Boyd, S., El Ghaoui, L., Feron, E., and Balakrishnan, V.,Linear Matrix Inequalities in System and Control Theory, Society for Industrial and Applied Mathematics, 1994. doi:10.1137/1.9781611970777

    work page doi:10.1137/1.9781611970777 1994

  23. [23]

    Reduced Order Models for Closed Loop Control: Comparison between POD, BPOD, and Global Modes,

    Barbagallo, A., Sipp, D., and Schmid, P. J., “Reduced Order Models for Closed Loop Control: Comparison between POD, BPOD, and Global Modes,”Progress in Flight Physics, Vol. 3, 2012, pp. 503–512. doi:10.1051/eucass/201203503

    work page doi:10.1051/eucass/201203503 2012

  24. [24]

    Modelling for robust feedback control of fluid flows,

    Jones, B. L., Heins, P. H., Kerrigan, E. C., Morrison, J. F., and Sharma, A. S., “Modelling for robust feedback control of fluid flows,”Journal of Fluid Mechanics, Vol. 769, 2015, pp. 687–722. doi:10.1017/jfm.2015.84. 21

    work page doi:10.1017/jfm.2015.84 2015

  25. [25]

    Modal Analysis of Fluid Flows: An Overview,

    Taira, K., Brunton, S. L., Dawson, S. T. M., Rowley, C. W., Colonius, T., Mckeon, B. J., Schmidt, O. T., Gordeyev, S., Theofilis, V., and Ukeiley, L. S., “Modal Analysis of Fluid Flows: An Overview,”AIAA Journal, 2017. doi:10.2514/1.J056060

    work page doi:10.2514/1.j056060 2017

  26. [26]

    Advances in Measuring the Environmental and Social Impacts of Environmental Programs

    Rowley, C. W., and Dawson, S. T. M., “Model Reduction for Flow Analysis and Control,”Annual Review of Fluid Mechanics, Vol. 49, 2017, pp. 387–417. doi:10.1146/annurev-fluid-010816-060042

    work page doi:10.1146/annurev- 2017

  27. [27]

    Model reduction for fluids using balanced proper orthogonal decomposition,

    Rowley, C. W., “Model reduction for fluids using balanced proper orthogonal decomposition,”International Journal on Bifurcation and Chaos, Vol. 15, No. 3, 2005, pp. 997–1013. doi:10.1142/S0218127405012429

    work page doi:10.1142/s0218127405012429 2005

  28. [28]

    Modeling of transitional channel flow using balanced proper orthogonal decomposition,

    Ilak, M., and Rowley, C. W., “Modeling of transitional channel flow using balanced proper orthogonal decomposition,”Physics of Fluids, Vol. 20, 2008, p. 34103. doi:10.1063/1.2840197

    work page doi:10.1063/1.2840197 2008

  29. [29]

    Principal component analysis in linear systems: Controllability, observability, and model reduction,

    Moore, B., “Principal component analysis in linear systems: Controllability, observability, and model reduction,”IEEE Transactions on Automatic Control, Vol. 26, No. 1, 1981, pp. 17–32. doi:10.1109/TAC.1981.1102568

    work page doi:10.1109/tac.1981.1102568 1981

  30. [30]

    C.,Approximation of Large-Scale Dynamical Systems, Society for Industrial and Applied Mathematics, 2006

    Antoulas, A. C.,Approximation of Large-Scale Dynamical Systems, Society for Industrial and Applied Mathematics, 2006

    work page 2006

  31. [31]

    Balanced Model Reduction via the Proper Orthogonal Decomposition,

    Willcox, K., and Peraire, J., “Balanced Model Reduction via the Proper Orthogonal Decomposition,”AIAA Journal, Vol. 40, No. 11, 2002. doi:10.2514/2.1570

    work page doi:10.2514/2.1570 2002

  32. [32]

    CVX: Matlab Software for DisciplinedConvex Programming, version2.1,

    Grant, M., and Boyd, S., “CVX: Matlab Software for DisciplinedConvex Programming, version2.1,”http://cvxr.com/cvx, Mar. 2014

    work page 2014

  33. [33]

    Computation of System Balancing Transformations and Other Applications of Simultaneous Diagonalization Algorithms,

    Laub, A. J., Heath, M. T., Paige, C. C., and Ward, R. C., “Computation of System Balancing Transformations and Other Applications of Simultaneous Diagonalization Algorithms,”IEEE Transactions on Automatic Control, Vol. AC-32, No. 2, 1987, pp. 115–122. doi:10.1109/TAC.1987.1104549

    work page doi:10.1109/tac.1987.1104549 1987

  34. [34]

    P.,Chebyshev and Fourier Spectral Methods, Dover Publications, 2000

    Boyd, J. P.,Chebyshev and Fourier Spectral Methods, Dover Publications, 2000. doi:10.1007/978-0-387-77674-3

    work page doi:10.1007/978-0-387-77674-3 2000

  35. [35]

    A linear state-space representation of plane Poiseuille flow for control design: a tutorial,

    McKernan, J., Papadakis, G., and Whidborne, J. F., “A linear state-space representation of plane Poiseuille flow for control design: a tutorial,”International Journal of Modelling, Identification and Control, Vol. 1, No. 4, 2006, p. 272. doi:10.1504/IJMIC.2006.012615

    work page doi:10.1504/ijmic.2006.012615 2006

  36. [36]

    Control of Plane Poiseuille Flow: A Theoretical and Computational Investigation,

    McKernan, J., “Control of Plane Poiseuille Flow: A Theoretical and Computational Investigation,” Ph.D. thesis, Cranfield University, 2006

    work page 2006

  37. [37]

    Skogestad, S., and Postlethwaite, I.,Multivariable feedback control : analysis and design, John Wiley, 2005

    work page 2005

  38. [38]

    Computing the maximum transient energy growth,

    Whidborne, J. F., and Amar, N., “Computing the maximum transient energy growth,”BIT Numerical Mathematics, Vol. 51, No. 2, 2011, pp. 447–457. doi:10.1007/s10543-011-0326-4. 22 Appendix A. Limitations of Modal truncation for MTEG-minimizing control Modal truncation based on the eigendecomposition of the matrixA = VΛV−1 in (1) is a common approach for reduc...

    work page doi:10.1007/s10543-011-0326-4 2011