Pith. sign in

REVIEW 3 major objections 5 minor 40 references

State Synchronization for Homogeneous Networks of Non-introspective Agents in Presence of Input Saturation -A Scale-free Protocol Design

T0 review · 3 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A single protocol, designed without knowing the network size or topology, achieves regulated state synchronization for saturated homogeneous multi-agent systems on any graph whose agents are reachable from a root set.

desk verdict Plausible extension of prior low-gain work, but the semi-global 'scale-free' claim is stronger than the proof; the abstract overstates what the theorems establish. read the letter →

arxiv 1908.06535 v1 pith:NTRGB66K submitted 2019-08-18 eess.SY cs.SYeess.SP

Zhenwei Liu , Ali Saberi , Anton A. Stoorvogel , Donya Nojavanzadeh This is my paper · ORCID
classification eess.SYcs.SYeess.SP MSC 93A1693C1093D0593D2093B52
keywords multi-agentsystemsstatesynchronizationinputsaturationscale-freeprotocollow-gainfeedbackadaptivenonlinearsemi-globalregulationnon-introspectiveagents
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper sets out to prove that a network of identical linear agents, each with saturating actuators, can be driven to track a common reference trajectory by a protocol that is designed from the agent model alone and requires no knowledge of the number of agents or the structure and spectrum of the communication graph. The authors establish two results: an adaptive nonlinear protocol that achieves global regulated state synchronization for every initial condition, and a parametrized linear low-gain protocol that achieves semi-global regulation for any fixed compact set of initial conditions. Both cover full-state coupling and partial-state coupling, and both work for any directed graph in which every agent is reachable from at least one root agent that can observe the reference. If the results are right, a single one-shot protocol design can be reused when agents are added or the communication topology changes, as long as the reachability condition persists.

What carries the argument

The mechanism that carries the argument is the combination of the expanded Laplacian $\tilde{L} = L + \mathrm{diag}\{\iota_i\}$, whose eigenvalues all have positive real parts precisely when every agent is reachable from the root set $\mathcal{C}$, with the low-gain Riccati solution defined by $A^T P_\rho + P_\rho A - P_\rho B B^T P_\rho + \rho P_\rho = 0$ (or its semi-global counterpart with $+ \varepsilon I$). The additional information exchange $\hat\zeta_i = \sum_j a_{ij}(\xi_i - \xi_j)$, where $\xi_i$ collects internal protocol variables, is what creates the decoupled error system $\dot e = (I \otimes A - \tilde{L} \otimes I)e$, which is Hurwitz because $A$ is at most weakly unstable. The Riccati solution $P_\rho \to 0$ as $\rho \to 0$ supplies the small feedback gains that keep the saturated input in its linear region, and in the adaptive protocol the scheduling law $\varepsilon(\chi_i) = \max\{\rho \in (0,1] : \chi_i^T P_\rho \chi_i \, \mathrm{tr}(B^T P_\rho B) \le 1\}$ guarantees the control never saturates, converting the global problem into a time-varying low-gain stabilization problem.

What would settle it

Run the adaptive protocol (14)-(16) on a large directed chain, say 1,000 agents with a single root agent, using the paper's triple-integrator example and initial states far outside the unit ball. The claim predicts that the one-shot protocol drives every agent to the exosystem trajectory with no input ever saturating; if any input is saturated for a positive time interval, or if the regulation errors fail to converge, the global scale-free claim is refuted.

Watch

Extended reading notes

Core claim

The central claim is that regulated state synchronization under input saturation is scale-free solvable for homogeneous non-introspective agents: for any number $N$ and any directed graph in which every node is reachable from a nonempty root set $\mathcal{C}$, the same protocol built only from the agent triple $(A,B,C)$ makes every agent state $x_i(t)$ converge to the exosystem trajectory $x_r(t)$. The global result uses the nonlinear adaptive protocol (14)-(16), whose scheduling law $\varepsilon(\chi_i)$ chooses the instantaneous low-gain parameter so that the control signal lies inside the saturation limits for all time; the semi-global result uses the linear protocol (27)-(28) with a fixed small low-gain parameter chosen from the prescribed compact set of initial conditions. In both cases saturation is never activated along closed-loop trajectories, so the saturated system behaves linearly, and the proof reduces to coupled error systems driven by the expanded Laplacian $\tilde{L} = L + \mathrm{diag}\{\iota_i\}$ and by the low-gain Riccati solution. The load-bearing property is that all eigenvalues of $A$ lie in the closed left half-plane, which makes each block $A - \lambda_i I$ Hurwitz for the positive-real-part eigenvalues $\lambda_i$ of $\tilde{L}$.

Load-bearing premise

The load-bearing premise is that the agents are at most weakly unstable—none of the agent's natural modes grows without bound—together with the standard stabilizability and detectability assumptions; if any mode of A were genuinely unstable, the low-gain strategy that keeps the saturated inputs unsaturated would no longer work, and the claimed scale-free regulation would collapse.

Editorial extensions

If this is right

  • A single protocol implementation can be deployed in a fleet without redesigning gains when agents join or leave, provided every new agent remains reachable from some root agent.
  • The designer never needs to estimate the Laplacian spectrum or the graph's algebraic connectivity; only the reachability condition of Definition 1 must be checked.
  • The adaptive global protocol removes the need to choose a low-gain parameter at all, while the linear semi-global protocol trades that tuning freedom for a simpler fixed-gain implementation.
  • The results cover both full-state coupling and partial-state coupling, the latter using an observer gain and additional exchange of protocol states and inputs.
  • Regulated synchronization means the network does not merely agree with itself; it tracks a reference trajectory generated by the same dynamics, which is the form needed when agents must follow a planned path.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Going beyond the paper, the adaptive protocol's independence from tuning suggests a natural testable extension to switching topologies: if the graph at every instant lies in $\mathcal{G}_N^\mathcal{C}$, the same one-shot protocol may retain its guarantee, although the proof here treats a fixed graph.
  • A second extension, not proven here, is whether the semi-global $\varepsilon^*$ can be chosen uniformly for all graphs in $\mathcal{G}_N^\mathcal{C}$; the proof's bounds depend on the smallest positive real part among the eigenvalues of $\tilde{L}$, so a uniform choice may be impossible unless such a lower bound is available.
  • The same low-gain-plus-expanded-Laplacian decoupling should carry over to discrete-time agents with saturation and to output synchronization problems, since neither the error decoupling nor the saturation-avoidance step uses continuous-time structure.
  • The paper's setting assumes saturation level one and a normalized input; rescaling the saturation level or the input matrix tests whether the scale-free property survives actuator scaling, which is a practical concern not addressed in the text.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. This manuscript studies regulated state synchronization in homogeneous linear multi-agent systems with input saturation and non-introspective agents. It proposes an adaptive nonlinear protocol for global synchronization and a linear low-gain protocol for semi-global synchronization, with separate versions for full-state and partial-state coupling. The claimed contribution is that the protocol design uses only the agent model (A,B,C) and requires no knowledge of the graph or of the Laplacian spectrum, and that it is scale-free in the number of agents. Theorems 1 and 2 address global results, Theorems 3 and 4 address semi-global results, and a numerical example illustrates the global partial-state protocol on three graphs of different sizes.

Significance. If fully proven, the global result would be a useful contribution: a single scheduled low-gain protocol that avoids saturation by construction and works for arbitrary N and for all graphs in G_N^C is an elegant construction, and the simulations support the ease of implementation. The semi-global linear protocol is a natural low-gain design that would complement the existing literature. However, the manuscript as written has a load-bearing circular step in the semi-global proof, an incomplete global proof, and a mismatch between the abstract's 'scale-free / no spectrum knowledge' wording and the theorem statements, which allow epsilon* to depend on the graph. These issues are central and require a major revision.

major comments (3)
  1. [Section IV-A, proof of Theorem 3, Eqs. (30)-(32)] The saturation-inactivity argument is circular. Equation (30) gives dot e = (I otimes A - Lbar otimes I) e only under the assumption that sigma(u)=u, yet inequalities (31) and (32), which are then used to conclude that saturation is never active, are derived from this unsaturated error dynamics. The actual error dynamics obtained from (29) contains the saturation nonlinearity, so the proof does not rule out an interval on which saturation is active before the bound (32) is available. The same circular step is reused in the proof of Theorem 4 in Section IV-B. A valid proof must analyze the saturated closed-loop dynamics directly or establish an invariant-region argument that does not presuppose the conclusion.
  2. [Abstract and Problem 2 vs. Theorem 3] Even if the circularity is repaired, the proof does not support the paper's strong scale-free and no-spectrum-knowledge claims for the linear protocol. The existence of epsilon_1 in (31) and epsilon* in (35) depends on the decay rate of e, which is min_j Re lambda_j(Lbar). Graphs in G_N^C can have arbitrarily small edge weights (for example, a root-follower pair connected by an edge of weight a tending to 0), making this decay rate arbitrarily small and forcing epsilon* to shrink accordingly. The theorem as stated only asserts, for each graph, existence of an epsilon*, so it does not imply a single protocol parameter that works for all graphs with uniformly bounded initial conditions. The proof gives no way to choose epsilon without spectral information. Please either prove a uniform bound over G_N^C or qualify the abstract and introduction to state that epsilon* may depend on the graph's spectral properties.
  3. [Section III-A, proof of Theorem 1, after Eq. (21)] The proof of the global result is not complete. First, the statement that (16) guarantees saturation is inactive by construction should be justified with the explicit inequality ||B^T P_rho x||^2 <= tr(B^T P_rho B) x^T P_rho x, which is not stated. Second, the bounds involving z1, z2, z3 and beta_1, beta_2 in L1 are asserted without derivation; moreover, the notation ||.||_1 is used both for instantaneous absolute values and for L1 signal norms, and the claimed monotonicity property relating V_i and epsilon_alpha is not proved. Finally, the conclusion V_i -> 0 from dV_i/dt <= -alpha_tilde V_i + beta_tilde(t)(V_i+1)^{1/2} with beta_tilde in L1 is stated as clear but deserves a proof. Since Theorem 1 is a central claim, this gap must be closed. The proof of Theorem 2 in Section III-B refers back to this analysis and is only a sketch; it should be written out in full as well.
minor comments (5)
  1. [Section IV heading] The heading contains a typo: 'Scalabale' should be 'Scalable'.
  2. [Section I, paragraph 5] In the sentence about reference [39], 'studid' should be 'studied'.
  3. [Section II, Eq. (18) and later] The matrix Lbar is used from Eq. (18) onward but is never defined; it should be introduced explicitly (it appears to be the expanded Laplacian Ltilde).
  4. [Section IV-B, after Eq. (39)] The communicated information is written as zeta_hat_i = (zeta_hat_i2^T, zeta_hat_i2^T)^T; the second block should be zeta_hat_i1.
  5. [Theorem 4 statement] The compact sets should be written as subsets: S_a subset R^n, S_e subset R^n, and S_c subset R^{2n}, rather than elements of those spaces.

Circularity Check

1 steps flagged · score 4.0 of 10

Semi-global Theorem 3 proof contains a local circularity: the saturation-inactivity conclusion is derived from bounds obtained under the assumption that saturation is inactive.

  1. other [Theorem 3 proof, Section IV-A, equations (30)-(37)]
    "Let e = x̃−χ, if the saturation is not active, we can obtain ... e˙ = (I⊗A−L̄⊗I)e (30)... Then, there exists an ε1 such that for ε<ε1 ... ||(I⊗B^T Pε)e||2 < 1 (32)... Next, we prove (33)... if the saturation is not active, then Vdot ≤ ||I⊗(B^T Pε)e||2. Integrating both sides ... using (32), which proves (33)... we find that the saturation never gets activated, i.e. σ(−(I⊗B^TPε)(x̃−e)) = −(I⊗B^TPε)(x̃−e) (37)"

    The bound (32) is asserted for the error dynamics (30), which are derived under the hypothesis 'if the saturation is not active.' The proof then integrates the Lyapunov function V using the same saturation-free dynamics and invokes (32) to conclude that saturation never gets activated, i.e. (37). Thus the conclusion 'saturation never activated' is established by assuming saturation is inactive when deriving the very bounds used to prove it. No first-activation or finite-horizon contradiction argument is supplied to break the loop. This is a local logical circularity in the proof of the semi-global claim. It does not by itself make the global adaptive protocol circular, and the theorem may be repairable, but as written the derivation of (31)-(33) presupposes the target property.

full rationale

The paper's central contribution is a controller design, not a data fit or a renamed empirical pattern. The adaptive global protocols (14)-(16) and (23) are self-contained synthesis procedures; their proofs do not reduce to the problem statement. The low-gain ARE properties are cited to published parameter-free results [18], [40], and [17, Lemma 6.1] is a standard external lemma, so the self-citations are not load-bearing in a circular way. No fitted parameter is renamed as a prediction, and no uniqueness theorem is imported from the authors' prior work to forbid alternatives. The main circularity is localized to the proof of Theorem 3 (and by reference Theorem 4): inequality (32) is obtained from the unsaturated error dynamics (30), and the same inequality is then used to prove that saturation never activates. That is a proof-local loop rather than an equivalence of the theorem with its assumptions, so the score is moderate. Separately, the abstract's claim that the protocol does not need 'any knowledge of the directed network topology and the spectrum' and is 'scale-free' is stronger than the formal statement of Problem 2, where ε* is allowed to depend on N and on the graph; however, that mismatch is a rigor/scope issue, not a circularity. Overall, the derivation is not circular by construction, but the semi-global theorem as written contains a genuine circular step.

Assumptions & free parameters 2 free parameters · 6 assumptions · 0 invented entities

The central claim rests on standard low-gain and adaptive tools, plus the domain assumptions of neutrally stable agents, graph reachability from the root set, and the availability of additional information exchange. No data-fitted constants or new physical entities are introduced. The most fragile items are Assumption 1, which excludes exponentially unstable agents, and the external lemmas from the same authors' prior work that are invoked without full proof.

free parameters (2)
  • Semi-global low-gain parameter epsilon = unspecified, exists sufficiently small
    Theorems 3 and 4 require existence of epsilon* such that all epsilon in (0,epsilon*] work. The value depends on the compact initial-condition sets and the network, but is not computed or fitted. It is the standard low-gain design knob from [18], not a data-fitted constant.
  • Adaptive scheduling variable epsilon(chi_i) = max rho in (0,1] satisfying chi_i^T P_rho chi_i tr B^T P_rho B <= 1
    State-dependent gain in the global protocols (14) and (23); it is an online solution of the algebraic constraint (16), not a constant fitted to data. The proof needs this variable to stay bounded away from zero, which is not fully established.
assumptions (6)
  • domain assumption Assumption 1: eigenvalues of A in closed left half plane; (A,B) stabilizable; (A,C) detectable.
    Restricts agent dynamics to neutrally stable or stable systems; the low-gain design, P_rho tending to zero, and stability of A minus lambda I rely on this. Location: Section II, Assumption 1.
  • domain assumption Graph condition: every node is reachable from the root set C, equivalently all eigenvalues of the expanded Laplacian L_bar have positive real parts.
    Ensures the error dynamics e_dot = (I tensor A - L_bar tensor I)e are asymptotically stable and L_bar is invertible. Location: Definition 1 and the paragraph after it.
  • domain assumption Exosystem trajectory is generated by xr_dot = A xr with the same A, and at least one agent receives relative output to the exosystem.
    Internal model principle for regulated synchronization; the agent model must contain the exosystem dynamics. Location: equations (6)-(8).
  • domain assumption Additional information exchange among protocol variables chi_i and u_i is available over the same graph, as in (11), (17), (24)-(25).
    The proposed protocols are not purely output feedback; they require communication of internal controller states. This is stated in Problems 1 and 2, but is an extra capability compared to standard diffusive coupling.
  • standard math Lemma 6.1 of [17] bounds the derivative of P_epsilon(t) by a multiple of |dV/dt|.
    Invoked in the proof of Theorem 1 to handle the time-varying Lyapunov function; the lemma is not proved in this paper and is authored by the same group. Location: proof of Theorem 1, after equation (22).
  • standard math Properties of the parametric ARE: P_rho is increasing in rho, P_rho tends to zero as rho tends to zero, and the solution is unique.
    Imported from references [40] and [18]; the whole low-gain proof depends on these properties. Location: equations (15)-(16) and (28).

how reviews work

0 comments
Cite this review

Pith. "Pith review of State Synchronization for Homogeneous Networks of Non-introspective Agents in Presence of Input Saturation -A Scale-free Protocol Design." pith.science (2026). https://pith.science/paper/NTRGB66K

@misc{pith2026190806535,
  author       = {Pith},
  title        = {Pith review of: State Synchronization for Homogeneous Networks of Non-introspective Agents in Presence of Input Saturation -A Scale-free Protocol Design},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NTRGB66K}},
  note         = {Machine review of arXiv:1908.06535}
}
read the original abstract

This paper studies global and semi-global regulated state synchronization of homogeneous networks of non-introspective agents in presence of input saturation based on additional information exchange where the reference trajectory is given by a so-called exosystem which is assumed to be globally reachable. Our protocol design methodology does not need any knowledge of the directed network topology and the spectrum of the associated Laplacian matrix. Moreover, the proposed protocol is scalable and achieves synchronization for any arbitrary number of agents.

Figures

Figures reproduced from arXiv: 1908.06535 by the authors.

Figure 1
Figure 1. The directed communication network, Case [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 5
Figure 5. Results for MAS with associated communication graph [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Results for MAS with associated communication graph [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 40 canonical work pages

  1. [1]

    Chowdhury and H

    D. Chowdhury and H. K. Khalil. Synchronization in networks of identical linear systems with reduced information. In American Control Conference, pages 5706–5711, Milwaukee, WI, 2018

    work page 2018

  2. [2]

    H. Chu, J. Yuan, and W. Zhang. Observer-based consensus tracking for linear multi-agent systems with input saturation.IET Control Theory and Applications, 9(14):2124–2131, 2015

    work page 2015

  3. [3]

    J. Fu, G. Wen, W. Yu, and Z. Ding. Finite-time consensus for second- order multi-agent systems with input saturation.IEEE Trans. Circ. & Syst.-II Express briefs, 65(11):1758–1762, 2018

    work page 2018

  4. [4]

    Godsil and G

    C. Godsil and G. Royle. Algebraic graph theory, volume 207 of Graduate Texts in Mathematics. Springer-Verlag, New York, 2001. 0 2 4 6 8 10 12 14 16 18 20 time(s) 0 50 100states Regulated State Synchronization, Case 1 0 2 4 6 8 10 12 14 16 18 20 time(s) -4 -2 0 errors 0 2 4 6 8 10 12 14 16 18 20 time(s) -1 0 1 inputs 0 2 4 6 8 10 12 14 16 18 20 time(s) 0.5...

    work page 2001

  5. [5]

    H.F. Grip, A. Saberi, and A.A. Stoorvogel. On the existence of virtual exosystems for synchronized linear networks. Automatica, 49(10):3145–3148, 2013

    work page 2013

  6. [6]

    H. Kim, H. Shim, J. Back, and J. Seo. Consensus of output-coupled linear multi-agent systems under fast switching network: averaging approach. Automatica, 49(1):267–272, 2013

    work page 2013

  7. [7]

    Y. Li, J. Xiang, and W. Wei. Consensus problems for linear time- invariant multi-agent systems with saturation constraints.IET Control Theory and Applications, 5(6):823–829, 2011

    work page 2011

  8. [8]

    Z. Liu, A. Saberi, A. A. Stoorvogel, and M. Zhang. Passivity-based state synchronization of homogeneous multiagent systems via static protocol in the presence of input saturation.Int. J. Robust & Nonlinear Control, 28(7):2720–2741, 2018

    work page 2018

Show all 40 references
  1. [9]

    Z. Liu, M. Zhang, A. Saberi, and A. A. Stoorvogel. State synchroniza- tion of multi-agent systems via static or adaptive nonlinear dynamic protocols. Automatica, 95:316–327, 2018

    2018

  2. [10]

    Z. Liu, M. Zhang, A. Saberi, and A.A. Stoorvogel. Passivity based state synchronization of homogeneous discrete-time multi-agent sys- tems via static protocol in the presence of input delay. European Journal of Control, 41:16–24, 2018. 0 2 4 6 8 10 12 14 16 18 20 time(s) 0 50 ...

    2018

  3. [11]

    Z. Meng, Z. Zhao, and Z. Lin. On global leader-following consensus of identical linear dynamic systems subject to actuator saturation.Syst. & Contr. Letters, 62(2):132–142, 2013

    2013

  4. [12]

    Olfati-Saber, J.A

    R. Olfati-Saber, J.A. Fax, and R.M. Murray. Consensus and coopera- tion in networked multi-agent systems.Proc. of the IEEE, 95(1):215– 233, 2007

    2007

  5. [13]

    Olfati-Saber and R.M

    R. Olfati-Saber and R.M. Murray. Consensus problems in networks of agents with switching topology and time-delays.IEEE Trans. Aut. Contr., 49(9):1520–1533, 2004

    2004

  6. [14]

    W. Ren. On consensus algorithms for double-integrator dynamics. IEEE Trans. Aut. Contr., 53(6):1503–1509, 2008

    2008

  7. [15]

    Ren and R.W

    W. Ren and R.W. Beard. Consensus seeking in multiagent systems under dynamically changing interaction topologies.IEEE Trans. Aut. Contr., 50(5):655–661, 2005

    2005

  8. [16]

    Ren and Y.C

    W. Ren and Y.C. Cao. Distributed coordination of multi-agent networks. Communications and Control Engineering. Springer-Verlag, London, 2011

    2011

  9. [17]

    Saberi, P

    A. Saberi, P. Hou, and A.A. Stoorvogel. On simultaneous global exter- nal and global internal stabilization of critically unstable linear systems with saturating actuators.IEEE Trans. Aut. Contr., 45(6):1042–1052, 2000

    2000

  10. [18]

    Saberi, A.A

    A. Saberi, A.A. Stoorvogel, and P. Sannuti. Internal and external stabilization of linear systems with constraints. Birkhäuser, Boston, MA, 2012

    2012

  11. [19]

    Scardovi and R

    L. Scardovi and R. Sepulchre. Synchronization in networks of identical linear systems. Automatica, 45(11):2557–2562, 2009

    2009

  12. [20]

    J.H. Seo, J. Back, H. Kim, and H. Shim. Output feedback consensus for high-order linear systems having uniform ranks under switching topology. IET Control Theory and Applications, 6(8):1118–1124, 2012

    2012

  13. [21]

    Consensusofhigh-orderlinearsystems using dynamic output feedback compensator: low gain approach

    J.H.Seo,H.Shim,andJ.Back. Consensusofhigh-orderlinearsystems using dynamic output feedback compensator: low gain approach. Automatica, 45(11):2659–2664, 2009

    2009

  14. [22]

    L. Shi, Y. Li, and Z. Lin. Semi-global leader-following output consensus of heterogeneous multi-agent systems with input saturation. Int. J. Robust & Nonlinear Control, 28(16):4916–4930, 2018

    2018

  15. [23]

    Su and M.Z.Q

    H. Su and M.Z.Q. Chen. Multi-agent containment control with input saturation on switching topologies. IET Control Theory and Applications, 9(3):399–409, 2015

    2015

  16. [24]

    Su, M.Z.Q

    H. Su, M.Z.Q. Chen, and G. Chen. Robust semi-global coordinated tracking of linear multi-agent systems with input saturation.Int. J. Robust & Nonlinear Control, 25(14):2375–2390, 2015

    2015

  17. [25]

    Su, M.Z.Q

    H. Su, M.Z.Q. Chen, J. Lam, and Z. Lin. Semi-global leader-following consensus of linear multi-agent systems with input saturation via low gain feedback. IEEE Trans. Circ. & Syst.-I Regular papers, 60(7):1881–1889, 2013

    2013

  18. [26]

    H. Su, G. Jia, and M.Z.Q. Chen. Semi-global containment control of multi-agent systems with intermittent input saturation.Journal of the Franklin Institute, 352(9):3504–3525, 2015

    2015

  19. [27]

    Su and J

    Y. Su and J. Huang. Stability of a class of linear switching systems with applications to two consensus problem.IEEE Trans. Aut. Contr., 57(6):1420–1430, 2012

    2012

  20. [28]

    K. Takaba. A dynamic protocol for local synchronization of linear multi-agent systems subject to input saturation. InProc. 54th CDC, pages 4923–4927, Osaka, Japan, 2015

    2015

  21. [29]

    S.E. Tuna. LQR-based coupling gain for synchronization of linear systems. Available: arXiv:0801.3390v1, 2008

    2008 arXiv

  22. [30]

    S.E. Tuna. Conditions for synchronizability in arrays of coupled linear systems. IEEE Trans. Aut. Contr., 55(10):2416–2420, 2009

    2009

  23. [31]

    X. Wang, H. Su, X. Wang, and G. Chen. An overview of coordinated control for multi-agent systems subject to input saturation.Perspectives in Science, 7:133–139, 2016

    2016

  24. [32]

    Wieland, J.S

    P. Wieland, J.S. Kim, and F. Allgöwer. On topology and dynamics of consensus among linear high-order agents. International Journal of Systems Science, 42(10):1831–1842, 2011

    2011

  25. [33]

    Wieland, R

    P. Wieland, R. Sepulchre, and F. Allgöwer. An internal model principle is necessary and sufficient for linear output synchronization. Automatica, 47(5):1068–1074, 2011

    2011

  26. [34]

    C.W. Wu. Synchronization in complex networks of nonlinear dynam- ical systems. World Scientific Publishing Company, Singapore, 2007

    2007

  27. [35]

    T. Yang, Z. Meng, D.V. Dimarogonas, and K.H. Johansson. Global consensus for discrete-time multi-agent systems with input saturation constraints. Automatica, 50(2):499–506, 2014

    2014

  28. [36]

    Yang, A.A

    T. Yang, A.A. Stoorvogel, H.F. Grip, and A. Saberi. Semi-global regulation of output synchronization for heterogeneous networks of non-introspective, invertible agents subject to actuator saturation.Int. J. Robust & Nonlinear Control, 24(3):548–566, 2014

    2014

  29. [37]

    X. Yi, T. Yang, J. Wu, and K. H. Johansson. Distributed event- triggered control for global consensus of multi-agent systems with input saturation. Automatica, 100:1–9, 2019

    2019

  30. [38]

    Zhang, M.Z.Q

    L. Zhang, M.Z.Q. Chen, and H. Su. Observer-based semi-global consensus of discrete-time multi-agent systems with input saturation. Transactions of the Institute of Measurement and Control, 38(6):665– 674, 2016

    2016

  31. [39]

    Semiglobal state synchronization for continuous-or discrete-time multiagent systems subject to actuator saturation

    Meirong Zhang, Ali Saberi, and Anton A Stoorvogel. Semiglobal state synchronization for continuous-or discrete-time multiagent systems subject to actuator saturation. International journal of robust and nonlinear control, 28(16):4966–4980, 2018

    2018

  32. [40]

    Zhou, G.R

    B. Zhou, G.R. Duan, and Z. Lin. A parametric Lyapunov equation approach to the design of low gain feedback.IEEE Trans. Aut. Contr., 53(6):1548–1554, 2008

    2008

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.