arxiv: 2605.10431 · v1 · submitted 2026-05-11 · 📡 eess.SY · cs.SY· math.OC

Recognition: 1 theorem link

· Lean Theorem

Hierarchical 2-degree-of-freedom control combining Youla-Kucera parameterization and model predictive control

Hans Henrik Niemann, John Bagterp J{\o}rgensen, Zhiheng Zhao

Pith reviewed 2026-05-12 05:08 UTC · model grok-4.3

classification 📡 eess.SY cs.SYmath.OC

keywords Youla-Kucera parameterizationmodel predictive controloffset-free control2-degree-of-freedomcoprime factorizationH2 optimal controlhierarchical control

0 comments

The pith

A 2DOF control structure uses Youla-Kucera parameterization to separate cascaded MPC optimization from H2-designed offset-free control.

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

This paper proposes a hierarchical two-degree-of-freedom control scheme that combines Youla-Kucera parameterization with model predictive control. The parameterization introduces an auxiliary feedforward channel for system optimization through cascaded MPC and a separate channel for controller parameterization to achieve offset-free performance using an H2 optimal design. Readers interested in advanced control would care because the method unifies optimization and regulation in one framework based on coprime factorizations. It builds on existing nominal controllers without requiring new stability proofs for the added features.

Core claim

The paper establishes that the Youla-Kucera parameterization, employing the coprime factorization of the nominal system and controller, introduces an auxiliary feedforward channel for cascaded MPC system optimization and a controller parameterization channel for offset-free MPC achieved by designing the YK parameter through H2 optimal controller design.

What carries the argument

Youla-Kucera parameterization via coprime factorization of the nominal plant and controller, which splits the control into a feedforward optimization channel and a parameterization channel for offset-free regulation.

If this is right

The feedforward channel implements cascaded MPC for optimization.
The parameterization channel achieves offset-free MPC via H2 optimal YK parameter design.
The overall structure forms a hierarchical 2DOF controller.
Stability and performance properties are inherited from the nominal controller and the H2 design.

Where Pith is reading between the lines

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

This approach may allow easier integration of economic MPC layers with regulatory control in industrial applications.
Extensions could include robust variants if uncertainty is incorporated into the factorization.
It suggests a general way to add optimization to existing feedback controllers without redesigning the base loop.

Load-bearing premise

The coprime factorization of the nominal system and controller must exist and support the H2 optimal design to ensure offset-free performance.

What would settle it

A closed-loop test showing nonzero steady-state error under a step reference or constant disturbance would disprove the offset-free property of the proposed controller.

Figures

Figures reproduced from arXiv: 2605.10431 by Hans Henrik Niemann, John Bagterp J{\o}rgensen, Zhiheng Zhao.

**Figure 3.** Figure 3: Block diagram for designing YK parameter. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

**Figure 4.** Figure 4: Simulation results under different control configurations. “ref [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗

read the original abstract

A hierarchical 2DOF (2-degree-of-freedom) structure combining Youla-Kucera (YK) parameterization and model predictive control (MPC) is presented in this paper. The YK parameterization employs the coprime factorization of the nominal system and controller, thereby introducing an auxiliary feedforward channel dedicated to system optimization and a controller parameterization channel. The feedforward channel is utilized to implement cascaded MPC for system optimization. The controller parameterization channel is utilized to achieve offset-free MPC by designing an appropriate YK parameter through the H2 optimal controller design.

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.

Desk Editor's Note private letter to a colleague

The paper gives a modular split of YK channels so MPC handles optimization in the feedforward path while H2 design on the parameterization path targets offset-free tracking, but the offset-free part rests on an assumption that may not hold without extra augmentation.

read the letter

The core idea is straightforward: use coprime factorization to create two channels in a 2DOF setup, put cascaded MPC in the auxiliary feedforward channel for system optimization, and tune the YK parameter Q with an H2-optimal controller in the feedback channel to get offset-free behavior. This keeps the stabilizing controller intact while layering the MPC on top in a structured way. The description stays close to standard YK and H2 results, which makes the construction easy to follow if you already know those tools. The modular assignment of roles is the clearest part of the contribution and could serve as a template for engineers who want to embed optimization without rebuilding the entire loop from scratch. The main weakness is the offset-free claim. Standard H2 minimization on the nominal coprime factors does not automatically place a zero at DC in the sensitivity function. Without an explicit integrator in the plant model or a disturbance model inside the H2 problem, constant disturbances or step references will generally produce nonzero steady-state error. The MPC feedforward channel cannot compensate for that because it sits outside the feedback loop for error correction. The abstract gives no indication that this augmentation was done, and the stress-test note flags exactly this point. If the full paper has no simulations with step disturbances or model mismatch, or no proof that the closed-loop DC gain is zero, the central claim does not hold up. This is the kind of paper that fits a control-engineering reading group or a special issue on MPC architectures. Readers already working with YK parameterization will see the structure quickly and can judge whether the missing augmentation step is addressed in the derivations or examples. It is incremental rather than foundational, but the presentation is clean enough that a serious referee could check the details and the validation in one pass. I would send it to peer review rather than desk-reject it.

Referee Report

1 major / 1 minor

Summary. The manuscript proposes a hierarchical 2DOF control architecture that integrates Youla-Kucera (YK) parameterization with model predictive control (MPC). Using coprime factorization of the nominal plant and stabilizing controller, the structure creates an auxiliary feedforward channel in which cascaded MPC is implemented for system optimization, while the controller parameterization channel sets the YK parameter Q via H2-optimal design to achieve offset-free MPC performance.

Significance. If the offset-free claim is rigorously established, the approach would provide a structured method for embedding predictive optimization within a 2DOF framework while preserving steady-state tracking properties through YK parameterization, without requiring redesign of the base controller. This could be useful for applications requiring both performance optimization and disturbance rejection.

major comments (1)

[Abstract / controller parameterization channel] Abstract and the description of the controller parameterization channel: the claim that H2-optimal design of the YK parameter Q achieves offset-free MPC is not automatically true under standard assumptions. The closed-loop map from reference (or constant disturbance) to tracking error must have DC gain exactly zero. Standard H2 minimization on the nominal coprime factors (X + MQ, Y - NQ) does not place a zero at DC in the sensitivity function unless the H2 problem explicitly augments the plant with an integrator or includes a disturbance model. The manuscript must derive the relevant transfer functions and show this property holds, or state the augmentation used.

minor comments (1)

[Abstract] The abstract supplies no equations, stability conditions, or simulation results; the full manuscript should include at least the coprime-factor definitions, the explicit H2 cost functional, and the resulting closed-loop sensitivity at DC.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive comments and the opportunity to strengthen the manuscript. We address the major comment point by point below.

read point-by-point responses

Referee: [Abstract / controller parameterization channel] Abstract and the description of the controller parameterization channel: the claim that H2-optimal design of the YK parameter Q achieves offset-free MPC is not automatically true under standard assumptions. The closed-loop map from reference (or constant disturbance) to tracking error must have DC gain exactly zero. Standard H2 minimization on the nominal coprime factors (X + MQ, Y - NQ) does not place a zero at DC in the sensitivity function unless the H2 problem explicitly augments the plant with an integrator or includes a disturbance model. The manuscript must derive the relevant transfer functions and show this property holds, or state the augmentation used.

Authors: We agree that the offset-free property is not automatic and requires explicit verification. The current manuscript states the property but does not derive the closed-loop maps or specify any augmentation in the H2 design of Q. We will revise the paper to add the necessary derivations of the sensitivity function and error maps, showing that the DC gain is zero when the H2 optimization is performed on a plant augmented with integrators (standard for enforcing the internal model principle). This will be inserted in the controller parameterization channel section, with a brief note added to the abstract. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation invokes standard external techniques

full rationale

The paper's central construction uses the established Youla-Kucera parameterization via coprime factorization of the nominal plant and controller, then applies standard H2-optimal design to select the YK parameter Q in the controller channel while placing cascaded MPC in the auxiliary feedforward channel. These are treated as independent, pre-existing building blocks rather than quantities fitted or defined inside the paper itself. No equation or claim reduces by construction to a tautology, a renamed fit, or a self-citation chain whose validity depends on the present work. The offset-free claim is asserted via the H2 design step, but that step is not shown to be circular; it is simply an application of a known method whose correctness is external to the manuscript.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No free parameters, axioms, or invented entities are identifiable from the abstract; the approach relies on standard coprime factorization and H2 optimization drawn from prior literature.

pith-pipeline@v0.9.0 · 5404 in / 1197 out tokens · 58861 ms · 2026-05-12T05:08:27.756354+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear
The YK parameterization employs the coprime factorization of the nominal system and controller... offset-free MPC by designing an appropriate YK parameter through the H2 optimal controller design.

Reference graph

Works this paper leans on

17 extracted references · 17 canonical work pages

[1]

Modern Wiener–Hopf design of optimal controllers Part I: The single-input-output case,

D. Youla, J. Bongiorno, and H. Jabr, “Modern Wiener–Hopf design of optimal controllers Part I: The single-input-output case,”IEEE Transactions on Automatic Control, vol. 21, no. 1, pp. 3–13, Feb. 1976

work page 1976
[2]

Modern Wiener-Hopf design of optimal controllers–Part II: The multivariable case,

D. Youla, H. Jabr, and J. Bongiorno, “Modern Wiener-Hopf design of optimal controllers–Part II: The multivariable case,”IEEE Trans- actions on Automatic Control, vol. 21, no. 3, pp. 319–338, Jun. 1976

work page 1976
[3]

Stability of discrete linear feedback systems,

V . Ku ˇcera, “Stability of discrete linear feedback systems,”IFAC Pro- ceedings Volumes, vol. 8, no. 1, Part 1, pp. 573–578, 1975, 6th IFAC World Congress (IFAC 1975) - Part 1: Theory, Boston/Cambridge, MA, USA, August 24-30, 1975

work page 1975
[4]

Advances in Youla-Kucera parametrization: A review,

I. Mahtout, F. Navas, V . Milanes, and F. Nashashibi, “Advances in Youla-Kucera parametrization: A review,”Annual Reviews in Control, vol. 49, pp. 81–94, 2020

work page 2020
[5]

T.-T. Tay, I. Mareels, and J. B. Moore,High Performance Control, ser. Systems & Control: Foundations & Applications. Boston, MA: Birkh¨auser Boston, 1998

work page 1998
[6]

Moving horizon estimation and control,

J. B. Jørgensen, “Moving horizon estimation and control,” Ph.D. dis- sertation, Department of Chemical Engineering, Technical University of Denmark, 2004

work page 2004
[7]

Numerical methods for large scale moving horizon estimation and control,

J. B. Jørgensen, J. B. Rawlings, and S. B. Jørgensen, “Numerical methods for large scale moving horizon estimation and control,”IFAC Proceedings Volumes, vol. 37(9), pp. 895–900, 2004

work page 2004
[8]

Finite horizon MPC for systems in innovation form,

J. B. Jørgensen, J. K., and J. B. Rawlings, “Finite horizon MPC for systems in innovation form,” in50th IEEE Conference on Decision and Control and European Control Conference (CDC-ECC), 2011, pp. 1896–1903

work page 2011
[9]

J. M. Maciejowski,Predictive Control with Constraints. Prentice- Hall, 2002

work page 2002
[10]

MPC for uncertain systems using the Youla parameterizations,

S. C. Thomsen, H. Niemann, and N. K. Poulsen, “MPC for uncertain systems using the Youla parameterizations,” in2008 47th IEEE Conference on Decision and Control. Cancun, Mexico: IEEE, 2008, pp. 3421–3426

work page 2008
[11]

A Youla parameter approach to robust constrained linear model predictive control,

Q. Cheng, B. Kouvaritakis, M. Cannon, and J. Rossiter, “A Youla parameter approach to robust constrained linear model predictive control,” 12 2009, pp. 2771–2776

work page 2009
[12]

Linear multivariable feedback systems,

J. Maciejowski, “Linear multivariable feedback systems,” inEncy- clopedia of Systems and Control Engineering (First Edition), first edition ed., Z. Ding, Ed. Oxford: Elsevier, 2026, pp. 66–82

work page 2026
[13]

Zhou and J

K. Zhou and J. C. Doyle,Essentials of Robust Control. Prentice Hall, 1998, vol. 104

work page 1998
[14]

Offset-free MPC ex- plained: novelties, subtleties, and applications,

G. Pannocchia, M. Gabiccini, and A. Artoni, “Offset-free MPC ex- plained: novelties, subtleties, and applications,”IFAC-PapersOnLine, vol. 48, no. 23, pp. 342–351, 2015, 5th IFAC Conference on Nonlinear Model Predictive Control NMPC 2015

work page 2015
[15]

Comparing model-based control strategies for a quadruple tank system: Decentralized PID, LMPC, and NMPC,

A. H. D. Christensen, T. K. S. Ritschel, J. L. Svensen, S. Hørsholt, J. K. Huusom, and J. B. Jørgensen, “Comparing model-based control strategies for a quadruple tank system: Decentralized PID, LMPC, and NMPC,”arXiv:2509.11235, 2025

work page arXiv 2025
[16]

The quadruple-tank process: a multivariable laboratory process with an adjustable zero,

K. Johansson, “The quadruple-tank process: a multivariable laboratory process with an adjustable zero,”IEEE Transactions on Control Systems Technology, vol. 8, no. 3, pp. 456–465, 2000

work page 2000
[17]

Dual Youla parameterisation,

H. Niemann, “Dual Youla parameterisation,”IEE Proceedings - Con- trol Theory and Applications, vol. 150, pp. 493–497, 2003

work page 2003