pith. sign in

arxiv: 2605.19992 · v1 · pith:ZO3BD6HGnew · submitted 2026-05-19 · 📡 eess.SY · cs.SY· math.OC

Robust synchronization for multi-agent systems governed by PDEs with observable and unobservable disturbances

Yongchun Bi , Jun Zheng , Guchuan Zhu , Jiye Zhang This is my paper

Pith reviewed 2026-05-20 03:50 UTC · model grok-4.3

classification 📡 eess.SY cs.SYmath.OC
keywords robust synchronizationmulti-agent systemsparabolic PDEsdisturbance observerinput-to-state stabilityboundary disturbancesdistributed control
0
0 comments X

The pith

A disturbance observer and distributed controllers achieve robust synchronization for multi-agent parabolic PDE systems using only boundary measurements.

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

This paper develops a control scheme for multi-agent systems whose individual dynamics follow parabolic partial differential equations, in the presence of both observable boundary disturbances and unobservable disturbances in the domain or at other boundaries. A disturbance observer is built from boundary output data to estimate the observable Dirichlet disturbances while staying robust to the unobservable ones; distributed controllers then use the reference trajectory and local outputs to drive every agent toward synchronization. Exponential convergence to the reference occurs when unobservable disturbances are absent, while bounded errors and input-to-state stability are shown to hold when they appear. Readers interested in distributed physical systems would care because many networked processes such as heat diffusion or fluid flow are described by such equations, and practical implementations rarely enjoy disturbance-free conditions or full-state access.

Core claim

Using only boundary output measurements, a disturbance observer is designed to estimate observable Dirichlet boundary disturbances while ensuring robustness of the observer error system with unobservable disturbances occurring in the domain. Distributed synchronization controllers are then constructed from the reference signal and local output information to enable all agents to track the reference trajectory. Exponential tracking is achieved in the absence of unobservable disturbances, while robustness is preserved when additional unobservable disturbances occur during controller implementation. The impact of unobservable Dirichlet-Robin boundary disturbances is analyzed by proving bounded,

What carries the argument

Disturbance observer for observable Dirichlet boundary disturbances together with distributed synchronization controllers, whose closed-loop error system is analyzed via the generalized Lyapunov method and recursion technique after well-posedness is established by the lifting technique and semigroup theory.

If this is right

  • Exponential tracking of the reference trajectory by every agent when unobservable disturbances are absent.
  • Preservation of robustness with bounded synchronization errors when unobservable disturbances appear during operation.
  • Input-to-state stability of the overall closed-loop system with respect to the combined effect of all disturbances.
  • Bounded solutions of the synchronization error system under unobservable Dirichlet-Robin boundary disturbances.

Where Pith is reading between the lines

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

  • The same observer-plus-controller structure could be tested on networks governed by other linear parabolic PDEs arising in diffusion or reaction processes.
  • Performance under time-varying or stochastic unobservable disturbances remains an open question that follows directly from the ISS result.
  • The boundary-measurement requirement suggests possible hardware implementations using only edge sensors in spatially distributed plants.

Load-bearing premise

The underlying PDE systems must admit well-posed solutions under the chosen boundary conditions so that the generalized Lyapunov method and recursion technique can establish boundedness and input-to-state stability for the assumed classes of observable and unobservable disturbances.

What would settle it

A concrete numerical simulation or analysis in which the synchronization error grows unbounded or the input-to-state stability estimate fails when unobservable disturbances are added would disprove the robustness claims.

Figures

Figures reproduced from arXiv: 2605.19992 by Guchuan Zhu, Jiye Zhang, Jun Zheng, Yongchun Bi.

Figure 1
Figure 1. Figure 1: Evolution of u ref(x, t) (top) and ku ref(·, t)kC([0,1]) (bottom) for the reference system (2) when D0 = 1. 0 0.5 1 1.5 2 0 0.5 1 1.5 2 0 0.5 1 1.5 2 0 0.5 1 1.5 2 0 0.5 1 1.5 2 0 0.5 1 1.5 2 [PITH_FULL_IMAGE:figures/full_fig_p019_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Evolution of the estimation error |q (i) (t) − qˆ (i) (0, t)| with different in-domain disturbance f (i) (x, t) when D1 = 1: A = 0 (top), A = 2 (middle), and A = 4 (bottom) [PITH_FULL_IMAGE:figures/full_fig_p019_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Evolution of u (1)(x, t) for the first agent in system (1) with different f (i) (x, t), d (i) 0 (t), and d (i) 1 (t) when D0 = D1 = 1. kue (i,j) (·, t)kC([0,1]) decays to a small neighborhood of zero in the absence of f (i) , d (i) 0 , and d (i) 1 , while remaining bounded when f (i) , d (i) 0 , and d (i) 1 are involved. Moreover, as the amplitudes of f (i) , d (i) 0 , and d (i) 1 decrease, the amplitudes … view at source ↗
Figure 4
Figure 4. Figure 4: Evolution of maxi kue (i) (·, t)kC([0,1]) for the tracking error system (9) with different f (i) (x, t) (top) and d (i) 0 (t), d(i) 1 (t) (bottom) when D1 = 1 [PITH_FULL_IMAGE:figures/full_fig_p020_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Evolution of maxi,j ku (i,j) (·, t)kC([0,1]) for the synchronization error system (10) with different f (i) (x, t) (top) and d (i) 0 (t), d(i) 1 (t) (bottom) when D1 = 1. 0 1 2 0 15 30 0 1 2 0 40 80 [PITH_FULL_IMAGE:figures/full_fig_p021_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Evolution of J(t) := maxi  [PITH_FULL_IMAGE:figures/full_fig_p021_6.png] view at source ↗
read the original abstract

This paper investigates robust synchronization for multi-agent systems (MASs) governed by parabolic partial differential equations in the presence of both observable and unobservable disturbances. Using only boundary output measurements, a disturbance observer is designed to estimate observable Dirichlet boundary disturbances while ensuring robustness of the observer error system with unobservable disturbances occurring in the domain. Using only the reference signal and local output information, distributed synchronization controllers are then constructed to enable all agents to track the reference trajectory. In particular, exponential tracking is achieved in the absence of unobservable disturbances, while robustness is preserved when additional unobservable disturbances occur during controller implementation. We further analyze the impact of unobservable Dirichlet-Robin boundary disturbances on synchronization performance by proving the boundedness of solutions to the synchronization error system. Moreover, to characterize the influence of all disturbances, input-to-state stability (ISS) is established for the closed-loop system. For the involved systems, the generalized Lyapunov method and the recursion technique are extensively employed in the stability analysis, and the lifting technique and semigroup theory are used to prove the well-posedness. Simulation results validate the proposed control scheme, demonstrating effective disturbance estimation and rejection, robust synchronization, and the ISS properties under various scenarios.

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

1 major / 1 minor

Summary. The manuscript proposes a robust synchronization control scheme for multi-agent systems governed by parabolic PDEs subject to both observable Dirichlet boundary disturbances and unobservable domain disturbances. A disturbance observer is designed using only boundary output measurements to estimate the observable disturbances while remaining robust to unobservable ones; distributed controllers are then constructed from the reference signal and local outputs to achieve synchronization. The paper claims exponential tracking in the absence of unobservable disturbances, preservation of robustness when they are present, boundedness of the synchronization error system under unobservable Dirichlet-Robin boundary disturbances, and input-to-state stability (ISS) of the closed-loop system. Proofs rely on the generalized Lyapunov method combined with a recursion technique, while well-posedness is established via the lifting technique and semigroup theory. Simulation results are provided to illustrate disturbance estimation, rejection, and ISS properties.

Significance. If the well-posedness and stability results hold, the work advances robust observer-based synchronization for infinite-dimensional multi-agent systems by explicitly handling mixed observable and unobservable disturbances with boundary-only measurements. The combination of established tools (semigroup theory, lifting, generalized Lyapunov, recursion) with ISS analysis and numerical validation provides a concrete contribution to PDE control literature, with potential relevance to applications such as networked heat or fluid systems where interior disturbances are common.

major comments (1)
  1. [well-posedness analysis (referenced in abstract and stability sections)] The well-posedness analysis invokes the lifting technique and semigroup theory to guarantee existence of solutions for the closed-loop system. However, unobservable disturbances are described as occurring in the domain (interior forcing terms), whereas the lifting technique is standardly applied to boundary disturbances. For interior disturbances to preserve the generation of a C0-semigroup, the perturbation must typically satisfy admissibility or relative boundedness conditions with respect to the generator (e.g., via Lumer-Phillips or perturbation theorems). No explicit verification of these conditions for the assumed class of unobservable disturbances is apparent, which directly underpins the subsequent boundedness and ISS claims.
minor comments (1)
  1. [Abstract / Introduction] The abstract refers to the 'generalized Lyapunov method and the recursion technique' without a brief introductory reference or one-sentence explanation; adding a short pointer to the relevant literature or a high-level outline would improve accessibility for readers outside the immediate subfield.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful and constructive review of our manuscript. The comment on well-posedness is well-taken and highlights an area where additional rigor will strengthen the presentation. We address it point by point below and will revise the manuscript accordingly.

read point-by-point responses

  1. Referee: The well-posedness analysis invokes the lifting technique and semigroup theory to guarantee existence of solutions for the closed-loop system. However, unobservable disturbances are described as occurring in the domain (interior forcing terms), whereas the lifting technique is standardly applied to boundary disturbances. For interior disturbances to preserve the generation of a C0-semigroup, the perturbation must typically satisfy admissibility or relative boundedness conditions with respect to the generator (e.g., via Lumer-Phillips or perturbation theorems). No explicit verification of these conditions for the assumed class of unobservable disturbances is apparent, which directly underpins the subsequent boundedness and ISS claims.

    Authors: We agree that the well-posedness argument for the closed-loop system with interior unobservable disturbances requires explicit verification of the perturbation conditions. In the original manuscript the lifting technique was applied to the observable Dirichlet boundary disturbances, while the unobservable domain disturbances were treated as forcing terms in the abstract evolution equation. To address the referee's concern, the revised version will add a dedicated paragraph in the well-posedness section that invokes the standard bounded-perturbation theorem for C0-semigroups: we will show that the operator induced by the unobservable domain disturbance is bounded (or relatively bounded with relative bound <1) with respect to the generator of the nominal parabolic operator. This directly implies that the perturbed operator still generates a C0-semigroup, thereby justifying existence and uniqueness of mild solutions. The same argument will be used to support the subsequent boundedness and ISS results. We view this addition as a clarification rather than a change in the underlying mathematics. revision: yes

Circularity Check

0 steps flagged

No circularity: derivation applies standard external techniques to constructed systems

full rationale

The paper constructs a disturbance observer and distributed controllers, then invokes lifting technique plus semigroup theory for well-posedness of the abstract evolution equation and generalized Lyapunov method plus recursion for boundedness and ISS of the synchronization error system. These are established, externally verifiable mathematical tools applied after the controller design; no parameter is fitted to a data subset and then renamed a prediction, no central claim reduces by definition to a self-citation chain, and no ansatz is smuggled via prior work by the same authors. The analysis remains self-contained against external benchmarks such as Lumer-Phillips or ISS definitions.

Axiom & Free-Parameter Ledger

0 free parameters · 3 axioms · 0 invented entities

The central claims rest on standard assumptions from infinite-dimensional systems theory and PDE control; no explicit free parameters or newly invented entities are introduced in the abstract.

axioms (3)
  • domain assumption The multi-agent systems are governed by parabolic partial differential equations with Dirichlet boundary conditions.
    Stated directly as the system class under study.
  • standard math Semigroup theory and the lifting technique guarantee well-posedness of the closed-loop systems.
    Invoked to prove existence of solutions as described in the abstract.
  • domain assumption The generalized Lyapunov method and recursion technique apply to the observer error and synchronization error systems.
    Extensively employed for stability and boundedness analysis.

pith-pipeline@v0.9.0 · 5754 in / 1548 out tokens · 52591 ms · 2026-05-20T03:50:14.088187+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

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

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

47 extracted references · 47 canonical work pages

  1. [1]

    Aguilar, Y

    L. Aguilar, Y . Orlov, and A. Pisano. Leader-follower syn chronization and ISS analysis for a network of boundary-con trolled wave PDEs. IEEE Control Systems Letters, 5(2):683–688, 2021

    work page 2021

  2. [2]

    Y . Chen, Z. Zuo, and Y . Wang. Bipartite consensus for a net work of wave equations with time-varying disturbances. Systems & Control Letters , 136,

    work page

  3. [3]

    Y . Chen, Z. Zuo, and Y . Wang. Bipartite consensus for a net work of wave PDEs over a signed directed graph. Automatica, 129, 2021. 109640. 21

    work page 2021

  4. [4]

    Z. Chen, J. Y ang, and X. Zong. Leader-follower synchroni zation controller design for a network of boundary-control led wave PDEs with structured time-varying perturbations and general disturbances. Journal of the Franklin Institute, 358(1):834–855, 2021

    work page 2021

  5. [5]

    M. A. Demetriou. Synchronization and consensus control lers for a class of parabolic distributed parameter systems . Systems & Control Letters, 62(1):70– 76, 2013

    work page 2013

  6. [6]

    M. A. Demetriou. Design of adaptive output feedback sync hronizing controllers for networked PDEs with boundary and in-domain structured perturbations and disturbances. Automatica, 90:220–229, 2018

    work page 2018

  7. [7]

    Deutscher

    J. Deutscher. Robust cooperative output regulation for a network of parabolic PDE systems. IEEE Transactions on Automatic Control , 67(1):451–459, 2022

    work page 2022

  8. [8]

    L. C. Evans. Partial Differential Equations. American Mathematical Society, Providence, Rhode Island , 2010

    work page 2010

  9. [9]

    Frihauf and M

    P . Frihauf and M. Krstic. Leader-enabled deployment ont o planar curves: A PDE-based approach. IEEE Transactions on Automatic Control, 56(8):1791– 1806, 2011

    work page 2011

  10. [10]

    Ghods and M

    N. Ghods and M. Krstic. Multiagent deployment over a sou rce. IEEE Transactions on Control Systems Technology, 20(1):277–285, 2012

    work page 2012

  11. [11]

    B. Z. Guo and H. C. Zhou. The active disturbance rejectio n control to stabilization for multi-dimensional wave equa tion with boundary control matched disturbance. IEEE Transactions on Automatic Control, 60(1):143–157, 2015

    work page 2015

  12. [12]

    J. Han. From PID to active disturbance rejection contro l. IEEE Transactions on Industrial Electronics, 56(3):900–906, 2009

    work page 2009

  13. [13]

    P . He. Consensus of uncertain parabolic PDE agents via a daptive unit-vector control scheme. IET Control Theory & Applications , 12(18):2488–2494, 2018

    work page 2018

  14. [14]

    Konduri, P

    S. Konduri, P . R. Pagilla, and S. Darbha. V ehicle formations using directed information flow graphs. In 2013 American Control Conference (ACC), pages 3045–3050, Washington, DC, USA, 2013

    work page 2013

  15. [15]

    S. R. Kumar and D. Mukherjee. Cooperative salvo guidanc e using finite-time consensus over directed cycles. IEEE Transactions on Aerospace and Electronic Systems, 56(2):1504–1514, 2020

    work page 2020

  16. [16]

    Li and J

    L. Li and J. Liu. Consensus tracking control and vibrati on suppression for nonlinear mobile flexible manipulator mu lti-agent systems based on PDE model. Nonlinear Dynamics, 111(4):3345–3359, 2023

    work page 2023

  17. [17]

    Y . J. Liu, X. Shang, L. Tang, and S. Zhang. Finite-time co nsensus adaptive neural network control for nonlinear mult iagent systems under PDE models. IEEE Transactions on Neural Networks and Learning Systems , 36(4):6218–6228, 2025

    work page 2025

  18. [18]

    Marshall, M

    J. Marshall, M. Broucke, and B. Francis. Formations of v ehicles in cyclic pursuit. IEEE Transactions on Automatic Control, 49(11):1963–1974, 2004

    work page 1963

  19. [19]

    S. D. J. McArthur, E. M. Davidson, V . M. Catterson, A. L. D imeas, N. D. Hatziargyriou, F. Ponci, et al. Multi-agent sys tems for power engineering applications—part I: Concepts, approaches and technical c hallenges. IEEE Transactions on Power Systems, 22(4):1743–1752, 2007

    work page 2007

  20. [20]

    Meurer and M

    T. Meurer and M. Krstic. Finite-time multi-agent deplo yment: A nonlinear PDE motion planning approach. Automatica, 47(11):2534–2542, 2011

    work page 2011

  21. [21]

    Mironchenko

    A. Mironchenko. Input-to-state stability of distribu ted parameter systems. arXiv:2302.00535

    work page arXiv

  22. [22]

    Mironchenko and C

    A. Mironchenko and C. Prieur. Input-to-state stabilit y of infinite-dimensional systems: Recent results and open q uestions. SIAM Review, 62(3):529–614, 2020

    work page 2020

  23. [23]

    Morstyn, B

    T. Morstyn, B. Hredzak, and V . G. Agelidis. Cooperative multi-agent control of heterogeneous storage devices dist ributed in a DC microgrid. IEEE Transactions on Power Systems, 31(4):2974–2986, 2016

    work page 2016

  24. [24]

    Orlov, A

    Y . Orlov, A. Pilloni, A. Pisano, and E. Usai. Consensus- based leader-follower tracking for a network of perturbed d iffusion PDEs via local boundary interaction. IF AC-PapersOnLine, 49(8):228–233, 2016

    work page 2016

  25. [25]

    A. Pazy. Semigroups of Linear Operators and Applications to Partial Differential Equations. Springer, New Y ork, NY , 1983

    work page 1983

  26. [26]

    Pilloni, A

    A. Pilloni, A. Pisano, Y . Orlov, and E. Usai. Consensus-based control for a network of diffusion PDEs with boundary local interaction. IEEE Transactions on Automatic Control, 61(9):2708–2713, 2016

    work page 2016

  27. [27]

    Qi, S.-X

    J. Qi, S.-X. Tang, and C. Wang. Parabolic PDE-based mult i-agent formation control on a cylindrical surface. International Journal of Control, 92(1):77– 99, 2019

    work page 2019

  28. [28]

    J. Qi, R. V azquez, and M. Krstic. Multi-agent deploymen t in 3-D via PDE control. IEEE Transactions on Automatic Control, 60(4):891–906, 2015

    work page 2015

  29. [29]

    J. Qi, S. Wang, J.-A. Fang, and M. Diagne. Control of mult i-agent systems with input delay via PDE-based method. Automatica, 106:91–100, 2019

    work page 2019

  30. [30]

    J. Qi, J. Zhang, and Y . Ding. Wave equation-based time-v arying formation control of multiagent systems. IEEE Transactions on Control Systems Technology, 26(5):1578–1591, 2018

    work page 2018

  31. [31]

    Qiu and H

    Q. Qiu and H. Su. Exponential consensus of multiagent sy stems based on nonlinear parabolic PDEs via intermittent bo undary control. IEEE Transactions on Systems, Man, and Cybernetics: Systems , 53(8):4833–4842, 2023

    work page 2023

  32. [32]

    J. A. Rogge and D. Aeyels. V ehicle platoons through ring coupling. IEEE Transactions on Automatic Control, 53(6):1370–1377, 2008

    work page 2008

  33. [33]

    S. X. Tang, J. Qi, and J. Zhang. Formation tracking contr ol for multi-agent systems: A wave-equation based approach . International Journal of Control, Automation and Systems, 15:2704–2713, 2017

    work page 2017

  34. [34]

    S. Wang, M. Diagne, and J. Qi. Delay-adaptive compensat ion for 3-D formation control of leader-actuated multi-age nt systems. Automatica, 164, 2024. 111645

    work page 2024

  35. [35]

    X. Y an, K. Li, C. Y ang, J. Zhuang, and J. Cao. Consensus offractional-order multi-agent systems via observer-based boundary control. IEEE Transactions on Network Science and Engineering , 11(4):3370–3382, 2024

    work page 2024

  36. [36]

    Y ang, H

    C. Y ang, H. He, T. Huang, A. Zhang, J. Qiu, J. Cao, and X. Li . Consensus for non-linear multi-agent systems modelled by PDEs based on spatial boundary communication. IET Control Theory & Applications , 11(17):3196–3200, 2017. 22

    work page 2017

  37. [37]

    Q. Y ao, Q. Li, C. Huang, and H. Jahanshahi. PDE-based dis tributed adaptive fault-tolerant attitude consensus of mu ltiple flexible spacecraft. Aerospace Science and Technology, 146, 2024. 108934

    work page 2024

  38. [38]

    Y . Y u, Z. Li, X. Wang, and L. Shen. Bearing-only circumna vigation control of the multi-agent system around a moving t arget. IET Control Theory & Applications, 13(17):2747–2757, 2019

    work page 2019

  39. [39]

    J. Zhan, L. Zhang, and J. Qiao. Boundary consensus of net worked hyperbolic systems of conservation laws. IEEE Transactions on Automatic Control , 70(8):4989–5004, 2025

    work page 2025

  40. [40]

    Zhang, J

    H. Zhang, J. Li, T. Wang, Q. Zeng, and H. Y an. PDE-based ev ent-triggered containment control of multi-agent systems with input delay under spatial boundary communication. International Journal of Robust and Nonlinear Control , 33(2):753–766, 2023

    work page 2023

  41. [41]

    Zhang, R

    J. Zhang, R. V azquez, J. Qi, and M. Krstic. Multi-agent d eployment in 3-D via reaction¨Cdiffusion system with radia lly-varying reaction. Automatica, 161, 2024. 111491

    work page 2024

  42. [42]

    Zhang, L

    S. Zhang, L. Tang, and Y . J. Liu. Adaptive neural control for a network of parabolic PDEs with event-triggered mechan ism. IEEE Transactions on Parallel and Distributed Systems, 35(7):1320–1330, 2024

    work page 2024

  43. [43]

    W. Zhao, Y . Liu, and X. Y ao. PDE-based boundary adaptive consensus control of multiagent systems with input constra ints. IEEE Transactions on Neural Networks and Learning Systems , 35(9):12617–12626, 2024

    work page 2024

  44. [44]

    Zheng and G

    J. Zheng and G. Zhu. Input-to-state stability with resp ect to boundary disturbances for a class of semi-linear para bolic equations. Automatica, 97(8):271– 277, 2018

    work page 2018

  45. [45]

    Zheng and G

    J. Zheng and G. Zhu. Global ISS for the viscous Burgers’ e quation with Dirichlet boundary disturbances. IEEE Transactions on Automatic Control , 69(10):7174–7181, 2024

    work page 2024

  46. [46]

    Zheng and G

    J. Zheng and G. Zhu. Generalized Lyapunov functionals f or the input-to-state stability of infinite-dimensional sy stems. Automatica, 172, 2025. 112005

    work page 2025

  47. [47]

    X. Zong, J. Zheng, J. Wei, and J. Y ang. Synchronizing controller design for networked heat equations with boundary structured time-varying perturbations and general disturbances. Journal of Process Control, 107:94–102, 2021. 23

    work page 2021