Dissipativity-Based Data-Driven Decentralized Control of Interconnected Systems

Ahmed Aboudonia; Andrea Martinelli; Florian D\"orfler; Jaap Eising; John Lygeros; Taiki Nakano

REVIEW 2 major objections 3 minor 52 references

Dissipativity-Based Data-Driven Decentralized Control of Interconnected Systems

T0 review · 2 major / 3 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read The paper proves that local noisy data alone, plus neighbor output measurements, are enough to synthesize decentralized stabilizing controllers for an unknown interconnected system.

desk verdict Solid data-driven decentralized control pipeline with a real but fixable gap in how Algorithm 1 handles the inertia condition. read the letter →

arxiv 2509.14047 v2 pith:T6D5T6ID submitted 2025-09-17 eess.SY cs.SY

Taiki Nakano , Ahmed Aboudonia , Jaap Eising , Andrea Martinelli , Florian Dörfler , John Lygeros This is my paper · ORCID

classification eess.SYcs.SY MSC 93A1493B5293C5593D30

keywords decentralizedcontroldata-drivendissipativitytheoryinterconnectedsystemslinearmatrixinequalitiesstate-feedbackdiscrete-timeLTImicrogrids

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

The paper tries to show that stabilizing controllers for an interconnected system can be designed from local data alone, without any model of the subsystems or of the interconnections. It gives a two-step pipeline: first, a data-driven LMI renders each closed-loop subsystem dissipative with respect to a chosen supply rate; second, a data-driven decentralized LMI checks that these supply rates are compatible with the unknown symmetric interconnection. Because the supply-rate matrices appear linearly in both LMIs, they can be treated as decision variables, so local dissipativity design and global stability certification are solved jointly. The resulting local state-feedback gains stabilize the nominal global closed-loop system for every system and every interconnection consistent with the measured data. A special case exploiting diffusive coupling computes a data-consistent upper bound on the coupling strength and uses it in a simpler stability condition.

What carries the argument

The machinery is the matrix S-lemma together with the dual quadratic matrix inequality. The data-consistent set of systems Sigma_i is described by a QMI built from local data and the noise bound Phi_i; the S-lemma turns 'dissipative for all systems in Sigma_i' into the single LMI (20) with multiplier alpha_i. Similarly, the interconnection data, through the dual QMI, turn 'stability inequality holds for all consistent interconnection matrices' into LMI (33) with multiplier tau_i. The supply-rate matrices (F_i, G_i, H_i), which enter linearly, act as the coupling variables, while the storage function P_i and the gain-related matrix L_i are the other decision variables. For diffusive coupling,

What would settle it

Generate random systems whose noise stays inside the assumed quadratic bounds, solve the LMIs (20), (27b), (33), and simulate the noiseless closed loop with the true system and true interconnection. If any feasible controller yields a diverging trajectory, the claimed sufficiency chain is false; because the paper proves sufficiency, a single such counterexample settles the question.

Watch

Extended reading notes

Core claim

The central claim is that, under quadratic bounds on process noise, measurement noise, and interconnection noise, the feasibility of the data-driven LMIs (20), (27b), and (33), with the supply-rate matrices (F_i, G_i, H_i) as decision variables, certifies that each closed-loop subsystem is dissipative with a supply rate that, together with the dual QMI of the interconnection data, guarantees asymptotic stability of the nominal global closed-loop system. Concretely, the computed gain K_i = L_i P_i^{-1} and the storage function V_i(x_i) = x_i^T P_i^{-1} x_i witness dissipativity, and the stability inequality holds for every interconnection matrix consistent with the neighbor data. This turns r

Load-bearing premise

The guarantee rests on knowing quadratic bounds for the unmeasured noise and interconnection noise, plus sufficiently rich data; if those bounds mischaracterize the real noise, the data-consistent sets are wrong and the certified controller may not stabilize the actual system.

Editorial extensions

If this is right

Controllers can be synthesized entirely from local data and neighbor output measurements; no global model, no identification step, and no centralized optimization are needed.
The closed-loop guarantee is robust: it holds for all systems in the data-consistent set Sigma_i and all interconnections consistent with the data, not just the observed trajectory.
Since the supply-rate matrices are decision variables, the design automatically finds a dissipativity certificate compatible with the interconnection, making the local and global conditions jointly feasible rather than iterated.
Because the nominal system is asymptotically stable, the actual noisy system is input-to-state stable with respect to process, measurement, and interconnection noise, so bounded noise leads to bounded state deviations.
For networks coupled by diffusion, the quadratic program (39) yields the tightest data-consistent upper bound on the weighted degree, which is then plugged into the stability condition to design the local gains.

Reading between the lines

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

If the same pipeline extends to directed interconnections with a suitable local dissipativity condition, the data-driven decomposition could apply to non-symmetric networks; the paper only treats symmetric coupling M = M^T.
Treating F_i, G_i, H_i as free variables suggests a natural way to tune performance: add objective terms on these matrices to shape the supply rate, for instance toward passivity or L2 gain, without changing the algorithm's structure. The paper leaves this open.
The feasibility drop of Algorithm 2 as noise or coupling density increases hints that the diffusive-coupling shortcut is most useful in lightly loaded, low-noise regimes; the authors conjecture this but do not quantify the conservatism.
Because the local conditions are independent LMIs, the approach could run online in a plug-and-play fashion when a new subsystem joins, provided its data and neighbor set are available; this is not tested in the paper.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Colleague — here's my take on arXiv:2509.14047.

The paper does something genuinely useful: it gives a data-driven, fully decentralized way to stabilize an interconnected LTI system when both local dynamics and the interconnection matrix are unknown. The key trick is to treat the supply-rate matrices (F_i, G_i, H_i) as decision variables, so the local dissipativity LMI and the decentralized stability LMI can be solved jointly. That is a real extension of the van Waarde et al. S-lemma machinery and of Martinelli et al.'s discrete-time dissipativity interconnection condition. The diffusive-coupling variant, with a data-driven upper bound on the weighted degree, is neat.

The math is generally sound. Theorems 1 and 2 are standard S-lemma/dual-QMI applications, and the proofs check out. The microgrid experiments are careful: 100 random systems, varying noise/data/topology, and a comparison against a centralized structured controller. The computational savings are real, though the baseline comparison is predictable because it requires global data.

The soft spot is the inertia condition (17). Lemma 1 requires the supply-rate matrix to have inertia (p_i,0,p_i). In Algorithm 1, that condition is listed as a constraint, but the LMI relaxation omits it; the text acknowledges this, and the implementation solves without it and checks afterwards. So the guarantee is conditional: if the SDP returns matrices violating (17), the equivalence between the primal dissipativity LMI and the dual form breaks, and the chain to global stability is not rigorous. This isn't fatal — in the experiments it held every time — but it is an internal gap that should be fixed, either with a convex relaxation that enforces the inertia or by explicitly stating it as a verification step and analyzing failure rates.

The other assumption — the noise QMI bounds Phi_i and Psi_i — is standard for this literature and not a hidden flaw. It is user-specified, but that's explicit.

Worth refereeing. I would accept it with a request to address the inertia-condition handling. The paper is honest, the derivations are reproducible, and the code is available. A serious referee can push on the inertia issue and on the conservatism of the decentralized condition.

If I worked in data-driven networked control, I'd cite it.

Referee Report

2 major / 3 minor

Summary. The paper proposes a data-driven, decentralized state-feedback synthesis method for interconnected discrete-time LTI systems. The authors first derive a local LMI condition (Theorem 1) that, for a given data set satisfying a QMI noise bound, yields a gain K_i rendering the nominal closed-loop subsystem dissipative with respect to a supply rate parameterized by (F_i,G_i,H_i). They then use a dual QMI argument (Theorem 2) to translate an interconnection stability condition from [15] into a decentralized data-driven LMI using measurements of v_i and \tilde{y}_i. By treating (F_i,G_i,H_i) as decision variables, the two LMIs are combined into Algorithm 1; a specialized diffusive-coupling variant (Algorithm 2) replaces the interconnection LMI with a condition based on the maximum weighted degree consistent with the data. Numerical experiments on a 50-DGU microgrid, including comparisons with a centralized structured data-driven controller, are reported.

Significance. If the claims hold, the paper is a clean contribution to direct data-driven control of networked systems: it is fully decentralized, avoids system identification, handles process/measurement/interconnection noise via QMIs, and provides a unified S-lemma/dissipativity framework. The benchmarks for data length and computation time are meaningful. However, the load-bearing gap in the treatment of the inertia condition (17) tempers the significance until it is resolved.

major comments (2)

[Algorithm 1 (Section IV-C), Theorem 1 (Section IV-A), and Section VI] The inertia condition (17) is essential for the equivalence in Lemma 1 between the primal dissipativity LMI (4) and the dual form (22) used in the proof of Theorem 1. However, Algorithm 1 lists (17) as a constraint while the immediately following text admits that step 3 is an LMI 'when we omit the inertia condition (17)'. Section VI confirms that the implementation solves the LMIs without (17) and checks it a posteriori. Since (17) is not convex, the relaxed LMI can return (F_i,G_i,H_i) for which (22) is not equivalent to dissipativity; the subsequent stability certificate via Proposition 4/Theorem 2 then has no basis. The paper provides no convex characterization of (17), nor does it report how often the post-hoc check succeeds in the experiments. As stated, Algorithm 1 is not an LMI algorithm, and the reported 'feasible instances' in Table II may overstate the number of systems for whi
[Theorem 1 (Section IV-A) and Algorithm 1] Theorem 1's 'if and only if' uses the converse of the matrix S-lemma (Proposition 1), which requires J_i to have at least one positive eigenvalue. This assumption is stated in the theorem but is not part of the constraints in Algorithm 1 and is not checked in the numerical implementation. The sufficiency direction holds without the converse, so the algorithm's stability guarantee is not destroyed, but the theorem overclaims an equivalence that the algorithm does not actually verify. The authors should either weaken the statement to a sufficient condition or add the positive-eigenvalue check to the algorithm and report its success rate.

minor comments (3)

[Abstract and Section IV-C] The abstract claims that 'both conditions take the form of linear matrix inequalities', but Algorithm 1 includes the non-convex inertia condition (17). The sentence 'Step 3 is an LMI condition when we omit the inertia condition (17)' should be reconciled with the abstract and the title's framing.
[Section V-B and Algorithm 2] The computation of d_max_i via (39) assumes that Theta_i is invertible. The paragraph on the invertibility assumption explains that a small perturbation of Psi_i restores it, but the numerical section does not state whether Theta_i was actually invertible in all generated instances, nor whether the perturbation was needed.
[Section VI] The authors report 'feasible instances' for Algorithms 1 and 2 in Table II, but they also state that the inertia condition is checked after solving the LMIs. To avoid misleading claims, the table should report how many of those feasible instances also passed the inertia check, or clarify that the reported numbers refer only to LMI feasibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the data-driven LMIs are derived from independent S-lemma/dissipativity theorems; stability guarantees are not fitted outputs.

full rationale

The derivation chain is self-contained in the sense required by the circularity rules. Theorem 1 starts from the data equation (18) and Assumption 1, characterizes the data-consistent set Sigma_i by the QMI (21), and uses the matrix S-lemma (Proposition 1) together with the dual dissipativity LMI (Lemma 1 from [30]) to prove that feasibility of LMI (20) is equivalent to existence of a gain K_i rendering all systems in Sigma_i dissipative. Theorem 2 similarly uses the dual QMI (Proposition 2) to make the model-based stability condition (27a) robust over all interconnections consistent with Assumption 2 and the data equation (30). The supply-rate matrices (F_i, G_i, H_i) are free decision variables (certificates), not parameters fitted to reproduce a pre-specified stability outcome, and the gains are not chosen to match the conclusion. The only author-overlapping citation that is load-bearing is Proposition 4 from [15]; it is a parameter-free model-based sufficient condition whose assumptions do not include the data-driven target, so under the review rules it counts as independent support rather than circularity. The paper's own caveat that Step 3 of Algorithm 1 is an LMI only when the inertia condition (17) is omitted, and the numerical practice of checking (17) after solving the LMIs, is a genuine correctness/computational limitation, but it is not a circularity: it does not define the stability conclusion in terms of the data inputs or smuggle the theorem into its assumptions.

Assumptions & free parameters 1 free parameters · 7 assumptions · 0 invented entities

The central claim rests on user-provided noise bounds, symmetric interconnections, and technical conditions for the S-lemma. These are standard assumptions in data-driven control; they are not fitted to the target result and no new physical entity is introduced.

free parameters (1)

alpha (diffusive coupling parameter) = alpha = 1 (in the numerical experiments)
User-specified scalar in Algorithm 2's condition (38); the feasible region and resulting controllers depend on this choice.

assumptions (7)

domain assumption Noise matrix W_i satisfies W_i^T in Z_{N_i}(Phi_i) for a known Phi_i (Assumption 1).
Section III-B. Defines Sigma_i; all guarantees are relative to this bound.
domain assumption Noise matrix Xi_i satisfies Xi_i^T in Z_{N_tilde_i}(Psi_i) for a known Psi_i (Assumption 2).
Section III-B. Defines Sigma_{M_tilde^r_i}.
domain assumption Interconnection matrix M is symmetric, M = M^T.
Section III-A. Used in Lemma 2 to decompose the global stability condition (25a) into local inequalities (27a).
domain assumption For each subsystem, input and output dimensions match (p_i = p_i) and i in N_i.
Section III-A. Simplifies the square-system formulation.
standard math Matrix S-lemma (Proposition 1) and dual QMI (Proposition 2) from [39] are valid.
Used in the proofs of Theorems 1 and 2 to convert set inclusions into LMIs.
domain assumption J_i has at least one positive eigenvalue; Theta_i is invertible.
Arguments in Sections IV-A and IV-B show these can be ensured by adding epsilon I to Phi_11,i or Psi_11,i, which slightly enlarges the noise bound.
domain assumption Local data matrix [X_i^T U_i^T V_i^T]^T has full row rank (excitation).
Section IV-A notes the LMI (20) can only be feasible if X_i has full row rank; this is an excitation condition.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Dissipativity-Based Data-Driven Decentralized Control of Interconnected Systems." pith.science (2026). https://pith.science/paper/T6D5T6ID

@misc{pith2026250914047,
  author       = {Pith},
  title        = {Pith review of: Dissipativity-Based Data-Driven Decentralized Control of Interconnected Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/T6D5T6ID}},
  note         = {Machine review of arXiv:2509.14047}
}

read the original abstract

We propose data-driven decentralized control algorithms for stabilizing interconnected discrete-time linear time-invariant systems. We first derive a data-driven condition to synthesize a local controller that ensures the dissipativity of the local subsystems. Then, we propose data-driven decentralized stability conditions for the global system based on the dissipativity of each local system. Since both conditions take the form of linear matrix inequalities and are based on dissipativity theory, this yields a unified pipeline, resulting in a data-driven decentralized control algorithm. As a special case, we also consider stabilizing systems interconnected through diffusive coupling and propose a control algorithm. We validate the effectiveness and the scalability of the proposed control algorithms in numerical examples in the context of microgrids.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 4 linked inside Pith

[15]

Interconnection of (Q, S, R)-Dissipative Systems in Discrete Time,

A. Martinelli, A. Aboudonia, and J. Lygeros, “Interconnection of (Q, S, R)-Dissipative Systems in Discrete Time,”arXiv preprint arXiv:2311.08088, 2023

arXiv 2023
[1]

Decentralized control: An overview,

L. Bakule, “Decentralized control: An overview,”Annual reviews in control, vol. 32, no. 1, pp. 87–98, 2008

2008
[2]

D. D. Siljak,Decentralized control of complex systems. Courier Corporation, 2013

2013
[3]

Decentralized control: Status and outlook,

L. Bakule, “Decentralized control: Status and outlook,”Annual Reviews in Control, vol. 38, no. 1, pp. 71–80, 2014

2014
[4]

van der Schaft,L2-gain and passivity techniques in nonlinear control

A. van der Schaft,L2-gain and passivity techniques in nonlinear control. Springer, 2000

2000
[5]

Dissipative dynamical systems part I: General theory,

J. C. Willems, “Dissipative dynamical systems part I: General theory,” Archive for rational mechanics and analysis, vol. 45, no. 5, pp. 321–351, 1972

1972
[6]

Passivity-based control of multi-agent systems,

N. Chopra and M. W. Spong, “Passivity-based control of multi-agent systems,”Advances in robot control: from everyday physics to human- like movements, pp. 107–134, 2006

2006
[7]

Arcak, C

M. Arcak, C. Meissen, and A. Packard,Networks of dissipative sys- tems: compositional certification of stability, performance, and safety. Springer, 2016

2016

Show all 52 references

[8]

Ortega, A

R. Ortega, A. Loria, P. J. Nicklasson, H. Sira-Ramirez, R. Ortega, A. Loría, P. J. Nicklasson, and H. Sira-Ramírez,Passivity-based Control of Euler-Lagrange Systems. Springer, 1998

1998
[9]

Discrete-time lossless systems, feedback equivalence and passivity,

C. I. Byrnes and W. Lin, “Discrete-time lossless systems, feedback equivalence and passivity,” inProceedings of 32nd IEEE Conference on Decision and Control, pp. 1775–1781, IEEE, 1993

1993
[10]

Losslessness, feedback equivalence, and the global stabilization of discrete-time nonlinear systems,

C. I. Byrnes and W. Lin, “Losslessness, feedback equivalence, and the global stabilization of discrete-time nonlinear systems,”IEEE Transac- tions on automatic control, vol. 39, no. 1, pp. 83–98, 1994

1994
[11]

Dissipa- tivity and feedback dissipativity properties of general nonlinear discrete- time systems,

E. M. Navarro-López, H. Sira-Ramírez, and E. Fossas-Colet, “Dissipa- tivity and feedback dissipativity properties of general nonlinear discrete- time systems,”European journal of control, vol. 8, no. 3, pp. 265–274, 2002

2002
[12]

A Generalized Passivity Theory over Abstract Time Domains,

A. Moreschini, M. Bin, A. Astolfi, and T. Parisini, “A Generalized Passivity Theory over Abstract Time Domains,”IEEE Transactions on Automatic Control, 2024

2024
[13]

Stability of interconnected switched systems using QSR dissipativity with multiple supply rates,

M. J. McCourt and P. J. Antsaklis, “Stability of interconnected switched systems using QSR dissipativity with multiple supply rates,” in2012 American Control Conference (ACC), pp. 4564–4569, IEEE, 2012

2012
[14]

Passivity-based decen- tralized control for discrete-time large-scale systems,

A. Aboudonia, A. Martinelli, and J. Lygeros, “Passivity-based decen- tralized control for discrete-time large-scale systems,” in2021 American Control Conference (ACC), pp. 2037–2042, IEEE, 2021
[16]

System identification,

L. Ljung, “System identification,” inSignal analysis and prediction, pp. 163–173, Springer, 1998

1998
[17]

Identification for control: From the early achievements to the revival of experiment design,

M. Gevers, “Identification for control: From the early achievements to the revival of experiment design,”European journal of control, vol. 11, no. 4-5, pp. 335–352, 2005

2005
[18]

A note on persistency of excitation,

J. C. Willems, P. Rapisarda, I. Markovsky, and B. L. De Moor, “A note on persistency of excitation,”Systems & Control Letters, vol. 54, no. 4, pp. 325–329, 2005

2005
[19]

J. C. Willems and J. W. Polderman,Introduction to mathematical systems theory: a behavioral approach, vol. 26. Springer Science & Business Media, 1997

1997
[20]

Data-driven simulation and control,

I. Markovsky and P. Rapisarda, “Data-driven simulation and control,” International Journal of Control, vol. 81, no. 12, pp. 1946–1959, 2008

1946
[21]

A data-driven predictive controller design based on reduced Hankel matrix,

H. Yang and S. Li, “A data-driven predictive controller design based on reduced Hankel matrix,” in2015 10th Asian Control Conference (ASCC), pp. 1–7, IEEE, 2015

2015
[22]

Data-enabled predictive con- trol: In the shallows of the DeePC,

J. Coulson, J. Lygeros, and F. Dörfler, “Data-enabled predictive con- trol: In the shallows of the DeePC,” in2019 18th European Control Conference (ECC), pp. 307–312, IEEE, 2019

2019
[23]

Data informativity: a new perspective on data-driven analysis and control,

H. J. van Waarde, J. Eising, H. L. Trentelman, and M. K. Camlibel, “Data informativity: a new perspective on data-driven analysis and control,” IEEE Transactions on Automatic Control, vol. 65, no. 11, pp. 4753– 4768, 2020

2020
[24]

Formulas for data-driven control: Stabilization, optimality, and robustness,

C. De Persis and P. Tesi, “Formulas for data-driven control: Stabilization, optimality, and robustness,”IEEE Transactions on Automatic Control, vol. 65, no. 3, pp. 909–924, 2019

2019
[25]

Data-driven model predictive control with stability and robustness guarantees,

J. Berberich, J. Köhler, M. A. Müller, and F. Allgöwer, “Data-driven model predictive control with stability and robustness guarantees,”IEEE Transactions on Automatic Control, vol. 66, no. 4, pp. 1702–1717, 2020. NAKANOet al.: DISSIPATIVITY -BASED DATA-DRIVEN DECENTRALIZED CON...

2020
[26]

From noisy data to feedback controllers: Nonconservative design via a matrix S-lemma,

H. J. van Waarde, M. K. Camlibel, and M. Mesbahi, “From noisy data to feedback controllers: Nonconservative design via a matrix S-lemma,” IEEE Transactions on Automatic Control, vol. 67, no. 1, pp. 162–175, 2020

2020
[27]

On Lyapunov functions and data-driven dissipativity,

T. Maupong, J. C. Mayo-Maldonado, and P. Rapisarda, “On Lyapunov functions and data-driven dissipativity,”IFAC-PapersOnLine, vol. 50, no. 1, pp. 7783–7788, 2017

2017
[28]

Provably robust verification of dissipativity properties from data,

A. Koch, J. Berberich, and F. Allgöwer, “Provably robust verification of dissipativity properties from data,”IEEE Transactions on Automatic Control, vol. 67, no. 8, pp. 4248–4255, 2021

2021
[29]

On the one-shot data-driven verifica- tion of dissipativity of LTI systems with general quadratic supply rate function,

T. E. Rosa and B. Jayawardhana, “On the one-shot data-driven verifica- tion of dissipativity of LTI systems with general quadratic supply rate function,” in2021 European Control Conference (ECC), pp. 1291–1296, IEEE, 2021

2021
[30]

Data-driven dissipativity analysis: application of the matrix S-lemma,

H. J. van Waarde, M. K. Camlibel, P. Rapisarda, and H. L. Trentelman, “Data-driven dissipativity analysis: application of the matrix S-lemma,” IEEE Control Systems Magazine, vol. 42, no. 3, pp. 140–149, 2022

2022
[31]

Synthesis of Dissipative Systems Using Input-State Data,

E. T. Nguyen and H. J. van Waarde, “Synthesis of Dissipative Systems Using Input-State Data,” in2024 European Control Conference (ECC), pp. 2959–2964, IEEE, 2024

2024
[32]

Data-Driven State-Feedback Controller Synthesis for Dissipativity: A Dualization-Based Approach,

P. Kristovi ´c and A. Joki ´c, “Data-Driven State-Feedback Controller Synthesis for Dissipativity: A Dualization-Based Approach,” in2024 American Control Conference (ACC), pp. 1219–1224, IEEE, 2024

2024
[33]

Algebraic approach to synthesis of data-driven control design for dissipativity,

Y . Tanaka, O. Kaneko, and T. Sueyoshi, “Algebraic approach to synthesis of data-driven control design for dissipativity,”SICE Journal of Control, Measurement, and System Integration, vol. 17, no. 1, pp. 247–255, 2024

2024
[34]

Data-driven control of complex networks,

G. Baggio, D. S. Bassett, and F. Pasqualetti, “Data-driven control of complex networks,”Nature communications, vol. 12, no. 1, p. 1429, 2021

2021
[35]

Data-driven control of distributed event-triggered network systems,

X. Wang, J. Sun, G. Wang, F. Allgöwer, and J. Chen, “Data-driven control of distributed event-triggered network systems,”IEEE/CAA Journal of Automatica Sinica, vol. 10, no. 2, pp. 351–364, 2023

2023
[36]

Distributed data-driven control of network systems,

F. Celi, G. Baggio, and F. Pasqualetti, “Distributed data-driven control of network systems,”IEEE Open Journal of Control Systems, vol. 2, pp. 93–107, 2023

2023
[37]

A compositional dissipativity approach for data-driven safety verification of large-scale dynamical systems,

A. Lavaei, S. Soudjani, and E. Frazzoli, “A compositional dissipativity approach for data-driven safety verification of large-scale dynamical systems,”IEEE Transactions on Automatic Control, vol. 68, no. 12, pp. 7240–7253, 2023

2023
[38]

Data-Driven Decentralized Control Design for Discrete-Time Large-Scale Systems,

J. Liao, S. Lu, T. Wang, and W. Xiang, “Data-Driven Decentralized Control Design for Discrete-Time Large-Scale Systems,”arXiv preprint arXiv:2411.10243, 2024

2024 arXiv
[39]

Quadratic matrix inequalities with applications to data-based control,

H. J. van Waarde, M. K. Camlibel, J. Eising, and H. L. Trentelman, “Quadratic matrix inequalities with applications to data-based control,” SIAM Journal on Control and Optimization, vol. 61, no. 4, pp. 2251– 2281, 2023

2023
[40]

From data to reduced-order models of complex dynamical systems,

A. M. Burohman, “From data to reduced-order models of complex dynamical systems,”Ph.D. dissertation, University of Groningen, 2023

2023
[41]

On relationships among passivity, positive realness, and dissipativity in linear systems,

N. Kottenstette, M. J. McCourt, M. Xia, V . Gupta, and P. J. Antsaklis, “On relationships among passivity, positive realness, and dissipativity in linear systems,”Automatica, vol. 50, no. 4, pp. 1003–1016, 2014

2014
[42]

Bullo, J

F. Bullo, J. Cortés, F. Dörfler, and S. Martínez,Lectures on network systems, vol. 1. CreateSpace, 2018

2018
[43]

Robust data- driven state-feedback design,

J. Berberich, A. Koch, C. W. Scherer, and F. Allgöwer, “Robust data- driven state-feedback design,” in2020 American Control Conference (ACC), pp. 1532–1538, IEEE, 2020

2020
[44]

Input-to-state stability for discrete-time nonlinear systems,

Z.-P. Jiang and Y . Wang, “Input-to-state stability for discrete-time nonlinear systems,”Automatica, vol. 37, no. 6, pp. 857–869, 2001

2001
[45]

R. A. Horn and C. R. Johnson,Matrix analysis. Cambridge university press, 2012

2012
[46]

Boyd and L

S. Boyd and L. Vandenberghe,Convex optimization. Cambridge univer- sity press, 2004

2004
[47]

Yalmip: A toolbox for modeling and optimization in mat- lab,

J. Lofberg, “Yalmip: A toolbox for modeling and optimization in mat- lab,” in2004 IEEE international conference on robotics and automation (IEEE Cat. No. 04CH37508), pp. 284–289, IEEE, 2004

2004
[48]

Version 9.0., 2019

MOSEK ApS,The MOSEK optimization toolbox for MATLAB manual. Version 9.0., 2019

2019
[49]

Plug-and-play voltage and frequency control of islanded microgrids with meshed topology,

S. Riverso, F. Sarzo, and G. Ferrari-Trecate, “Plug-and-play voltage and frequency control of islanded microgrids with meshed topology,”IEEE Transactions on Smart Grid, vol. 6, no. 3, pp. 1176–1184, 2014

2014
[50]

A consensus-based secondary control layer for stable current sharing and voltage balancing in DC microgrids,

M. Tucci, L. Meng, J. M. Guerrero, and G. Ferrari-Trecate, “A consensus-based secondary control layer for stable current sharing and voltage balancing in DC microgrids,”arXiv preprint arXiv:1603.03624, 2016

2016 arXiv
[51]

Controller architectures: Tradeoffs between performance and structure,

M. R. Jovanovi ´c and N. K. Dhingra, “Controller architectures: Tradeoffs between performance and structure,”European Journal of Control, vol. 30, pp. 76–91, 2016

2016
[52]

Data-driven structured robust control of linear systems,

J. Miller, J. Eising, F. Dörfler, and R. S. Smith, “Data-driven structured robust control of linear systems,”arXiv preprint arXiv:2411.11542, 2024

2024 arXiv

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.

[1] [15]

Interconnection of (Q, S, R)-Dissipative Systems in Discrete Time,

A. Martinelli, A. Aboudonia, and J. Lygeros, “Interconnection of (Q, S, R)-Dissipative Systems in Discrete Time,”arXiv preprint arXiv:2311.08088, 2023

arXiv 2023

[2] [1]

Decentralized control: An overview,

L. Bakule, “Decentralized control: An overview,”Annual reviews in control, vol. 32, no. 1, pp. 87–98, 2008

2008

[3] [2]

D. D. Siljak,Decentralized control of complex systems. Courier Corporation, 2013

2013

[4] [3]

Decentralized control: Status and outlook,

L. Bakule, “Decentralized control: Status and outlook,”Annual Reviews in Control, vol. 38, no. 1, pp. 71–80, 2014

2014

[5] [4]

van der Schaft,L2-gain and passivity techniques in nonlinear control

A. van der Schaft,L2-gain and passivity techniques in nonlinear control. Springer, 2000

2000

[6] [5]

Dissipative dynamical systems part I: General theory,

J. C. Willems, “Dissipative dynamical systems part I: General theory,” Archive for rational mechanics and analysis, vol. 45, no. 5, pp. 321–351, 1972

1972

[7] [6]

Passivity-based control of multi-agent systems,

N. Chopra and M. W. Spong, “Passivity-based control of multi-agent systems,”Advances in robot control: from everyday physics to human- like movements, pp. 107–134, 2006

2006

[8] [7]

Arcak, C

M. Arcak, C. Meissen, and A. Packard,Networks of dissipative sys- tems: compositional certification of stability, performance, and safety. Springer, 2016

2016