Data-Enabled Predictive Control for Flexible Spacecraft

Amin Vahidi-Moghaddam; Daning Huang; Haowei An; Huanqing Wang; Kaixiang Zhang; Nan Li; Zhaojian Li

REVIEW 4 major objections 6 minor 32 references

Data-Enabled Predictive Control for Flexible Spacecraft

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper claims that data-enabled predictive control (DeePC) can steer a flexible spacecraft's hub angle and suppress beam vibrations using only recorded torque–angle data, with no parametric model, and that in finite-element…

desk verdict A credible but over-claimed first application of DeePC to flexible spacecraft boundary control: the simulation is plausible, but the persistence-of-excitation precondition is unverified and several claims outrun the evidence. read the letter →

arxiv 2502.09531 v1 pith:Q57C74YN submitted 2025-02-13 eess.SY cs.SY

Huanqing Wang , Kaixiang Zhang , Amin Vahidi-Moghaddam , Haowei An , Nan Li , Daning Huang , Zhaojian Li This is my paper · ORCID

classification eess.SYcs.SY

keywords data-enabledpredictivecontrolDeePCflexiblespacecraftboundaryWillemsfundamentallemmadata-drivenvibrationsuppressionsingularvaluedecomposition

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

Flexible spacecraft appendages are hard to control because their PDE-ODE dynamics resist accurate modeling and discretization can cause spillover. This paper tries to establish that the data-enabled predictive control (DeePC) framework—which builds a non-parametric model directly from recorded torque and angle trajectories via Willems' fundamental lemma—can replace the model-based Lyapunov boundary controller for this system. The authors report finite-element simulations in which DeePC tracks the desired hub angle and suppresses beam vibration across nominal, model-uncertainty, and process-noise scenarios, with lower cost than the Lyapunov benchmark and faster settling in the uncertainty case. A dimension-reduction step based on singular value decomposition keeps the optimization manageable. A sympathetic reading: the value is that a controller needing no parameter calibration can match, and in uncertain cases beat, a controller that assumes exact knowledge of the plant.

What carries the argument

The load-bearing object is the Hankel data matrix (8) built from an offline input–output trajectory, used as a non-parametric stand-in for the state-space model. Willems' fundamental lemma states that if the recorded input is persistently exciting of order n+L, every length-L trajectory of the controllable LTI system is a linear combination of the columns of this matrix; DeePC then solves the predictive-control optimization (12), which selects a coefficient vector g, a future input u, and a future output y while penalizing a slack variable for noise and adding an ℓ2 regularizer on g. The paper compresses the Hankel matrix with an SVD, keeping the range space via \bar{H}_L = H_L V_1 = W_1 \Sigma_1, which cuts the optimization dimension from about 3960 to 80 in the experiments. The same structure is what lets the controller claim to work without model identification.

What would settle it

Compute the rank of the 80 × 3961 Hankel matrix formed from the Section IV.A data and compare it with full row rank; if the rank falls short, the fundamental-lemma condition is violated. Alternatively, train DeePC on a coarse finite-element mesh and test it on a finer mesh or on the PDE-ODE model directly: if tracking or vibration suppression degrades sharply, the reported success is tied to training and evaluation sharing the same discretization.

Watch

Extended reading notes

Core claim

The central claim is that DeePC, using only input–output data collected while a PD controller drives the spacecraft, can achieve accurate angle tracking and vibration suppression for a flexible spacecraft without prior knowledge of system parameters or derivation of a mathematical model. In the paper's finite-element simulations, the regularized DeePC formulation (12) with the SVD-compressed data matrix (16) produces lower cost than the Lyapunov-based boundary controller in the nominal case (2287 vs. 2741), under model uncertainty (2623 vs. 3870), and under process noise (2290 vs. 2801); its settling time is longer in the nominal case (62.8 vs. 58.6 s) but markedly shorter under uncertainty (64.3 vs. 90.6 s). The authors take these results as strong evidence of the validity and effectiveness of the DeePC approach for flexible spacecraft.

Load-bearing premise

The load-bearing premise is that the torque sequence used to collect offline data is persistently exciting enough that the recorded input–output Hankel matrix spans every trajectory of the discretized spacecraft; the paper calls the data 'sufficiently rich' but never verifies the full-row-rank condition of Definition 1.

Editorial extensions

If this is right

A controller for a flexible spacecraft can be synthesized from torque and tip-angle data alone, eliminating the need to identify stiffness, inertia, and damping parameters.
DeePC matches or outperforms the Lyapunov boundary controller on the reported cost in all three scenarios, and keeps a fast settling time when the true mass and stiffness are doubled.
The SVD compression reduces the DeePC decision variable from thousands of entries to the rank of the data matrix, making the online optimization feasible for longer horizons.
Because the offline data is generated by a simple PD controller, the method only requires a stabilizing, not optimal, baseline for data collection.
The approach inherits the robustness of regularized DeePC: the λ_g‖g‖² penalty is what the authors credit for maintaining performance under model uncertainty and noise.

Reading between the lines

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

Editorial inference: the same pipeline should transfer to other flexible structures—solar panels, booms, manipulators—as long as a stabilizing baseline controller can generate persistently exciting data; the paper only demonstrates the antenna-hub configuration.
Editorial inference: because the fundamental lemma applies to finite-dimensional LTI systems, the paper's evidence does not by itself prove performance on the true infinite-dimensional plant; a natural test is to train on one finite-element mesh and evaluate on a finer mesh or on a different discretization order.
Editorial inference: the SVD turning-point cutoff is chosen by inspection of the singular-value distribution; a sensitivity study over the rank r would reveal how much of the reported robustness depends on that choice.
Editorial inference: online or recursive updating of the data matrix could allow DeePC to track slow parameter drift during a long mission, an extension the paper does not pursue.

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.

Desk Editor's Note

read the letter

You should know two things about this paper. First, it is a legitimate first application of DeePC to boundary control of a flexible spacecraft, and the central angle-tracking simulation is credible. Second, the claims as written are noticeably stronger than what the evidence supports, mostly in places that are fixable with extra detail and softened language rather than fatal flaws.

What is genuinely new: this is the first DeePC application to a PDE-ODE flexible spacecraft system, and the comparison against a Lyapunov boundary controller under nominal, uncertainty, and process-noise cases in a finite-element simulation is not in the cited literature. The SVD dimension reduction from the authors' earlier work is sensibly imported, and the Lyapunov appendix is thorough—the stability proof is real, not ornamental. The FE model description is good enough to reproduce.

Now the soft spots. The stress-test's persistence-of-excitation concern lands. The paper never verifies the full-row-rank condition behind the fundamental lemma, and the data collection is described only as a PD controller tracking several desired angles. That kind of damped transient is unlikely to excite all modes to the required order, especially with FE state dimension around 86 and L = 40. The SVD truncation to r = 80 plus the slack variable can mask rank deficiency, so the optimization actually solved may not be the DeePC controller the theory justifies. That is the main technical gap. The related limitation—that the same FE discretization generates both the data and the closed-loop evaluation—means the infinite-dimensional PDE is never actually controlled, which undercuts the phrase 'stabilize an infinite number of vibration modes' in the introduction. The controlled output is also defined inconsistently: the text says tip-end deflection, the algorithm uses end-tip angle, and Table I's cost function is never defined. The uncertainty and noise scenarios lack protocol details, and there are no error bars or multiple runs.

These are all addressable. The core simulation result—DeePC tracking and vibration suppression without a model—is plausible and worth taking seriously. The paper just needs to verify or at least discuss the PE condition (or show the reduced Hankel matrix still spans the relevant trajectories), define the output and cost, report the noise/uncertainty setups, and tone down the infinite-mode claim.

This paper deserves a serious referee. If it lands in your hands, the revision should focus on the PE verification and a consistent, reproducible simulation protocol. I would bring it to a reading group as a useful example of how DeePC is being pushed into new application domains, not as a definitive proof of model-free superiority.

Referee Report

4 major / 6 minor

Summary. The paper proposes a data-enabled predictive control (DeePC) scheme for a flexible spacecraft modeled as a central rigid hub with a flexible Euler–Bernoulli appendage. It collects offline input-output data under a PD controller, builds a Hankel data matrix, applies an SVD-based dimension reduction, and solves a regularized receding-horizon optimization. The controller is evaluated in a 20-element finite-element simulation and compared with a Lyapunov boundary controller in nominal, model-uncertainty, and process-noise scenarios. The reported results show comparable or better angle tracking and lower cost for DeePC, and the paper concludes that the simulations provide strong evidence for DeePC's validity and effectiveness. The manuscript also contains a detailed Lyapunov stability proof and a finite-element discretization appendix.

Significance. The application area is relevant and the FE simulation study is a useful first demonstration of DeePC for flexible spacecraft. The benchmark Lyapunov derivation in Appendix II is unusually detailed and constructive, and the FE model in Appendix I is coherent. However, the claims as stated outpace the evidence: the theoretical justification via the fundamental lemma is not verified for the collected data, the 'infinite number of vibration modes' claim is not supported by a finite-element study, and the reported superiority rests on an undefined cost function and hand-tuned hyperparameters. With these issues addressed, the empirical contribution would be a reasonable feasibility study for data-driven boundary control of flexible structures.

major comments (4)

[Section III.A, Definition 1/Lemma 1 and Section IV.A/IV.C] The persistent-excitation requirement is asserted but never verified. For the 20-element FE model of Appendix I, the state dimension n is on the order of 82, and with Tini=20, N=20, the fundamental lemma would require the scalar input to be persistently exciting of order n+L, where L=40. The offline data are collected under a PD controller tracking step commands; a damped transient response to a few setpoints is not obviously sufficiently rich, and no rank test of the Hankel matrix in (8) is reported. Because the SVD reduction in (15)-(16) with r=80 and the slack variable sigma_y in (12) can preserve feasibility even when the Hankel matrix is rank deficient, the optimization actually solved is not guaranteed to be the DeePC problem justified by Lemma 1. The convergence in Figs. 4-9 is therefore not explained by the paper's theoretical framework; the authors should either verify full row rank of the relevant Hankel matrix or use an offline excitation signal with a documented persistent-excitation property.
[Section I and Section V] The claim that DeePC can 'stabilize an infinite number of vibration modes' is not demonstrated and, as written, is not compatible with the methodology. The offline data and closed-loop evaluation are both obtained from the finite-dimensional ODE system (35) in Appendix I with 20 beam elements, not from the PDE-ODE system (1)-(4). The fundamental lemma is a finite-dimensional LTI statement, so the DeePC surrogate in (12) applies, at best, to the discretized model. This contribution claim should be restricted to the tested finite-dimensional discretization or reworded to avoid implying a result for the infinite-dimensional PDE.
[Section III.A and Algorithm 1] The controlled output is not consistently specified. Section III.A states that 'the system output is the tip-end deflection,' while Algorithm 1 and the simulation section refer to a 'spacecraft angle output sequence' and plot both theta(t) and tip deflection. Since L=40 and the Hankel matrix has 80 rows, the data matrix in (8) uses a single scalar output in addition to the scalar input. Whether the DeePC output is theta(t) or y(L,t) is essential for interpreting the claim that DeePC simultaneously tracks angle and suppresses vibration. Please define the output vector unambiguously and state which measurement enters equations (8) and (12).
[Section IV.C and Table I] The 'cost function' used to compare DeePC and the Lyapunov controller is never defined. Equation (12) defines a per-horizon DeePC objective, but the values reported in Table I (e.g., 2287 versus 2741) are presumably cumulative over the full simulation; without the exact formula and normalization, a reader cannot reproduce Table I or judge the claimed superiority. In addition, Q, R, lambda_g, lambda_y, Tini, N, and the rank cutoff r are selected per scenario with no tuning rule and no sensitivity analysis, so the comparison is not robust evidence of a general advantage. Please add the cost definition and at least one sensitivity study for the main hyperparameters.

minor comments (6)

[Section III.A, after Eq. (5)] 'xk in R^m is the state vector' should read xk in R^n; the symbol m is already used for the input dimension.
[Section IV.C, Fig. 8 caption] The caption contains the typo 'comparision' and should read 'comparison.'
[Appendix II, Eq. (54)] In the first term of the displayed expression, theta(t) appears where theta_t(t) is evidently intended; please correct this so the integration-by-parts step is transparent.
[Table I] The process-noise column omits settling times; state whether settling time is defined only for the nominal and uncertainty scenarios.
[Section IV.A] The phrase 'sufficiently rich dataset' does not connect to Definition 1; a sentence explicitly stating T, Tini, N, and the state dimension used in the persistent-excitation order would help readers verify the Hankel construction.
[Algorithm 1] Line 1 lists the inputs but not their lengths; specify T and the sampling interval so that the Hankel matrix construction is fully reproducible.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central claim is an in-silico performance comparison supported by an external lemma and an independent Lyapunov proof; self-citations are not load-bearing.

full rationale

The paper's derivation chain is not circular. DeePC's theoretical foundation is Willems' fundamental lemma (Lemma 1), an external theorem, and the paper applies it in the standard way: offline input-output data are assembled into Hankel matrices, and optimization (12) selects trajectories in the data span. The Lyapunov benchmark is derived in the paper itself in Appendix II, with a complete stability proof and explicit parameter feasibility argument; it does not borrow the paper's conclusion. The only self-citations of note are [22], a prior soft-robot application cited as background, and [28], the SVD dimension-reduction method. Neither is load-bearing for the central tracking and vibration-suppression claim: the SVD reduction is standard linear algebra, and the paper itself states the range-preservation condition in Section III.B. The unverified persistent-excitation assumption for the PD-collected data and the hand-tuned DeePC hyperparameters are legitimate correctness and generalizability concerns, but they are not instances in which a result is defined in terms of its inputs or in which a fitted parameter is relabeled as a prediction. No equation is equivalent to another by construction, and no benchmark outcome is forced by the training data. Therefore, no circular step can be exhibited, and the paper should not be penalized for circularity.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the standard DeePC toolbox (fundamental lemma, slack regularization, SVD reduction) plus the FE model as ground truth. The paper contributes no new free-parameter-free derivation; the DeePC hyperparameters and benchmark gains are choices, and the robustness claims depend on unspecified data-collection details. No invented entities are introduced.

free parameters (5)

DeePC cost weights (Q, R) = Q = 1000, R = 2.5e-4
Chosen by hand in Section IV.C with no tuning procedure or sensitivity analysis; they directly shape the reported cost comparison with the Lyapunov baseline.
DeePC regularization weights (lambda_g, lambda_y) = lambda_g = 1000, lambda_y = 300000
Hand-selected per Section IV.C; the paper states lambda_y must be 'purposefully selected' but gives no rule. The robustness claims in the uncertainty and noise cases depend on these values.
DeePC horizon lengths (Tini, N) = Tini = 20, N = 20
Chosen without stated rationale; they fix the Hankel depth L = 40 and hence the prediction horizon used in all comparisons.
SVD rank cutoff r = r = 80
Chosen by locating a 'turning point' in the singular value distribution of the same Hankel data used for control (Section III.B); the reduced matrix is 80x80, i.e., full row rank is kept.
Lyapunov benchmark gains (a1, a2, k1, k2) = a1 = 0.0428, a2 = 3000, k1 = 0.1, k2 = 2.1e-10
Selected to satisfy the stability constraints in Appendix II, not tuned for the performance comparison; k2 is effectively zero, so the comparison baseline is weak.

assumptions (4)

domain assumption Willems' fundamental lemma applies to the offline data: the recorded torque sequence is persistently exciting of order n+L for the FE-discretized LTI plant.
Section III.A invokes Lemma 1 and Section IV.A asserts the dataset is 'sufficiently rich', but the full-row-rank condition of Definition 1 is never verified, and the plant is the infinite-dimensional PDE-ODE system (1)-(4), for which the finite-dimensional lemma is invoked only after FE discretization (Appendix I).
domain assumption The FE-discretized linear Euler-Bernoulli model (Appendix I) is the ground truth for all claims about vibration suppression, and DeePC only needs to control this discretization.
The paper's claims about the physical spacecraft (intro, conclusion) are drawn entirely from simulations of the FE model; no experimental validation exists, and the model is linear despite the paper's nonlinearity framing.
ad hoc to paper The slack variable and quadratic regularization in (12) convert the nonlinearity and noise gap into a robust formulation.
The paper states DeePC 'handles the nonlinearity and non-deterministic characteristics by introducing additional terms' (Section III.A) but provides no closed-loop stability or bounded-error guarantee for the regularized problem on a nonlinear or infinite-dimensional plant; this heuristic is load-bearing for the noise and uncertainty cases.
domain assumption Classical or smooth solutions are assumed so that integration by parts and the equality conditions of the Lyapunov derivative hold.
Appendix II states 'we only consider classical/smooth solutions' to conclude y_xx = 0 everywhere from an almost-everywhere identity; standard for PDE Lyapunov analyses but unverified for the FE solutions.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Data-Enabled Predictive Control for Flexible Spacecraft." pith.science (2026). https://pith.science/paper/Q57C74YN

@misc{pith2026250209531,
  author       = {Pith},
  title        = {Pith review of: Data-Enabled Predictive Control for Flexible Spacecraft},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q57C74YN}},
  note         = {Machine review of arXiv:2502.09531}
}

read the original abstract

Spacecraft are vital to space exploration and are often equipped with lightweight, flexible appendages to meet strict weight constraints. These appendages pose significant challenges for modeling and control due to their inherent nonlinearity. Data-driven control methods have gained traction to address such challenges. This paper introduces, to the best of the authors' knowledge, the first application of the data-enabled predictive control (DeePC) framework to boundary control for flexible spacecraft. Leveraging the fundamental lemma, DeePC constructs a non-parametric model by utilizing recorded past trajectories, eliminating the need for explicit model development. The developed method also incorporates dimension reduction techniques to enhance computational efficiency. Through comprehensive numerical simulations, this study compares the proposed method with Lyapunov-based control, demonstrating superior performance and offering a thorough evaluation of data-driven control for flexible spacecraft.

Figures

Figures reproduced from arXiv: 2502.09531 by the authors.

**Figure 2.** The angle and end-point deflection of the flexible appe [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** The deformation of the flexible appendage under no con [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: The performance comparison between DeePC and Lyapun [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

**Figure 5.** Figure 5: The deformation of the flexible spacecraft with Lyapu [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

**Figure 7.** Figure 7: The deformation of the flexible spacecraft with Lyapu [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

**Figure 8.** Figure 8: The performance comparision between Lyapunov-base [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 32 canonical work pages

[1]

A review on recent development of spacecraft attitude fault tolerant control system,

S. Yin, B. Xiao, S. X. Ding, and D. Zhou, “A review on recent development of spacecraft attitude fault tolerant control system,” IEEE Transactions on Industrial Electronics , vol. 63, no. 5, pp. 3311–3320, 2016

work page 2016
[2]

Adaptive output feedback sliding-mode manoeuvr ing and vibration control of ﬂexible spacecraft with input saturat ion,

Q. Hu, “Adaptive output feedback sliding-mode manoeuvr ing and vibration control of ﬂexible spacecraft with input saturat ion,” IET Control Theory & Applications , vol. 2, pp. 467–478(11), June 2008

work page 2008
[3]

V ariable structure control and active vi bration sup- pression of ﬂexible spacecraft during attitude maneuver,

Q. Hu and G. Ma, “V ariable structure control and active vi bration sup- pression of ﬂexible spacecraft during attitude maneuver,” Aerospace Science and Technology , vol. 9, no. 4, pp. 307–317, 2005

work page 2005
[4]

Three- axis attitude stabilization of a ﬂexible satellite using no n-linear pd controller,

B. G. Baghi, M. Kabganian, R. Nadaﬁ, and E. Arabi, “Three- axis attitude stabilization of a ﬂexible satellite using no n-linear pd controller,” Transactions of the Institute of Measurement and Control , vol. 40, pp. 591 – 605, 2018

work page 2018
[5]

Appl ication of piezoelectric devices to vibration suppression,

C. C. Won, J. L. Sulla, D. W. Sparks, and W. K. Belvin, “Appl ication of piezoelectric devices to vibration suppression,” Journal of Guidance, Control, and Dynamics , vol. 17, no. 6, pp. 1333–1338, 1994

work page 1994
[6]

Robust h∞ output feedback control for attitude stabilization of a ﬂexible spacecraft,

S. Wu and S. Wen, “Robust h∞ output feedback control for attitude stabilization of a ﬂexible spacecraft,” Nonlinear Dynamics , vol. 84, pp. 405–412, 2016

work page 2016
[7]

Energy-based co ntrol of axially translating beams: varying tension, varying speed , and distur- bance adaptation,

K.-J. Y ang, K.-S. Hong, and F. Matsuno, “Energy-based co ntrol of axially translating beams: varying tension, varying speed , and distur- bance adaptation,” IEEE Transactions on Control Systems Technology , vol. 13, no. 6, pp. 1045–1054, 2005

work page 2005
[8]

Output feedback variab le structure adaptive control of a ﬂexible spacecraft,

Y . Zeng, A. Araujo, and S. Singh, “Output feedback variab le structure adaptive control of a ﬂexible spacecraft,” Acta Astronautica , vol. 44, no. 1, pp. 11–22, 1999

work page 1999

Show all 32 references

[9]

Optimal distributed control of a ﬂexible spacecraft during a large-angle maneuver,

J. D. Turner and H. M. Chun, “Optimal distributed control of a ﬂexible spacecraft during a large-angle maneuver,” Journal of Guidance, Control, and Dynamics , vol. 7, no. 3, pp. 257–264, 1984

1984
[10]

Sliding mode and shaped in put vibration control of ﬂexible systems,

Q.-l. Hu, Z. Wang, and H. Gao, “Sliding mode and shaped in put vibration control of ﬂexible systems,” IEEE Transactions on Aerospace and Electronic Systems , vol. 44, no. 2, pp. 503–519, 2008

2008
[11]

Adaptive boundary control for ﬂexi ble two-link manipulator based on partial differential equation dynami c model,

L. Zhang and J. Liu, “Adaptive boundary control for ﬂexi ble two-link manipulator based on partial differential equation dynami c model,” IET Control Theory & Applications , vol. 7, no. 1, pp. 43–51, 2013

2013
[12]

Bending and torsional vibration control of a ﬂexible plate structure using h/sub /spl inﬁn// -based robust control law,

I. Kar, T. Miyakura, and K. Seto, “Bending and torsional vibration control of a ﬂexible plate structure using h/sub /spl inﬁn// -based robust control law,” IEEE Transactions on Control Systems Technology , vol. 8, no. 3, pp. 545–553, 2000

2000
[13]

Dynamic modeling and vibration contr ol of a ﬂexible satellite,

W. He and S. S. Ge, “Dynamic modeling and vibration contr ol of a ﬂexible satellite,” IEEE Transactions on Aerospace and Electronic Systems, vol. 51, no. 2, pp. 1422–1431, 2015

2015
[14]

Vibration control for a ﬂexible satellite system with output constraints,

T. Meng, W. He, H. Y ang et al. , “Vibration control for a ﬂexible satellite system with output constraints,” Nonlinear Dynamics, vol. 85, no. 4, pp. 2673–2686, 2016

2016
[15]

Bou ndary control of ﬂexible satellite vibration in planar motion,

H. K. Rad, H. Salarieh, A. Alasty, and R. V atankhah, “Bou ndary control of ﬂexible satellite vibration in planar motion,” Journal of Sound and Vibration , vol. 432, pp. 549–568, 2018

2018
[16]

Pde model-based boundary cont rol of a spacecraft with double ﬂexible appendages under prescr ibed performance,

J. Ma, H. Wen, and D. Jin, “Pde model-based boundary cont rol of a spacecraft with double ﬂexible appendages under prescr ibed performance,” Advances in Space Research , vol. 65, no. 1, pp. 586– 597, 2020

2020
[17]

A note on persistency of excitation,

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

2005
[18]

Data-enabled p redictive control: In the shallows of the deepc,

J. Coulson, J. Lygeros, and F. D¨ orﬂer, “Data-enabled p redictive control: In the shallows of the deepc,” in 2019 18th European Control Conference (ECC) , 2019, pp. 307–312

2019
[19]

Decen tralized data- enabled predictive control for power system oscillation da mping,

L. Huang, J. Coulson, J. Lygeros, and F. D¨ orﬂer, “Decen tralized data- enabled predictive control for power system oscillation da mping,” 2021

2021
[20]

Data-dr iven predic- tive control with online adaption: Application to a fuel cel l system,

L. Schmitt, J. Beerwerth, M. Bahr, and D. Abel, “Data-dr iven predic- tive control with online adaption: Application to a fuel cel l system,” IEEE Transactions on Control Systems Technology , vol. 32, no. 1, pp. 61–72, 2024

2024
[21]

Data- enabled predictive control for quadcopters,

E. Elokda, J. Coulson, P . N. Beuchat, J. Lygeros, and F. D ¨ orﬂer, “Data- enabled predictive control for quadcopters,” Int. J. Robust Nonlinear Control, vol. 31, no. 18, pp. 8916–8936, Dec 2021

2021
[22]

Mechanical design and data-enabled predictive c ontrol of a planar soft robot,

H. Wang, K. Zhang, K. Lee, Y . Mei, K. Zhu, V . Srivastava, J . Sheng, and Z. Li, “Mechanical design and data-enabled predictive c ontrol of a planar soft robot,” IEEE Robotics and Automation Letters , vol. 9, no. 9, pp. 7923–7930, 2024

2024
[23]

Deep-lcc: Data-enab led predic- tive leading cruise control in mixed trafﬁc ﬂow,

J. Wang, Y . Zheng, K. Li, and Q. Xu, “Deep-lcc: Data-enab led predic- tive leading cruise control in mixed trafﬁc ﬂow,” IEEE Transactions on Control Systems Technology , vol. 31, no. 6, p. 2760–2776, Nov. 2023

2023
[24]

Initial experiment s on the end- point control of a ﬂexible one-link robot,

J. Robert H. Cannon and E. Schmitz, “Initial experiment s on the end- point control of a ﬂexible one-link robot,” The International Journal of Robotics Research , vol. 3, no. 3, pp. 62–75, 1984

1984
[25]

Improving regulation of a sing le-link ﬂexible manipulator with strain feedback,

S. Ge, T. Lee, and G. Zhu, “Improving regulation of a sing le-link ﬂexible manipulator with strain feedback,” IEEE Transactions on Robotics and Automation , vol. 14, no. 1, pp. 179–185, 1998

1998
[26]

Data-driven model predictive control with stability and robustness gua rantees,

J. Berberich, J. K¨ ohler, M. A. M¨ uller, and F. Allg¨ ower, “Data-driven model predictive control with stability and robustness gua rantees,” IEEE Transactions on Automatic Control , vol. 66, no. 4, pp. 1702– 1717, 2021

2021
[27]

Quadrati c regularization of data-enabled predictive control: Theory and applicatio n to power converter experiments,

L. Huang, J. Zhen, J. Lygeros, and F. D¨ orﬂer, “Quadrati c regularization of data-enabled predictive control: Theory and applicatio n to power converter experiments,” IF AC-PapersOnLine, vol. 54, no. 7, pp. 192– 197, 2021

2021
[28]

Dimension reduc tion for efﬁcient data-enabled predictive control,

K. Zhang, Y . Zheng, C. Shang, and Z. Li, “Dimension reduc tion for efﬁcient data-enabled predictive control,” IEEE Control Systems Letters, vol. 7, pp. 3277–3282, 2023

2023
[29]

On the stability of pd control for a two-lin k rigid-ﬂexible manipulator,

A. S. Yigit, “On the stability of pd control for a two-lin k rigid-ﬂexible manipulator,” Journal of Dynamic Systems Measurement and Control- transactions of The Asme , vol. 116, pp. 208–215, 1994

1994
[30]

Bou ndary con- trol of anti-symmetric vibration of satellite with ﬂexible appendages in planar motion with exponential stability,

H. K. Rad, H. Salarieh, A. Alasty, and R. V atankhah, “Bou ndary con- trol of anti-symmetric vibration of satellite with ﬂexible appendages in planar motion with exponential stability,” Acta Astronautica , vol. 147, pp. 219–230, 2018

2018
[31]

P . P . Friedmann, G. A. Lesieutre, and D. Huang, Structural Dynamics. Cambridge University Press, 2023, ch. 4

2023
[32]

Krstic and A

M. Krstic and A. Smyshlyaev, Boundary Control of PDEs: A Course on Backstepping Designs . Philadelphia, PA: SIAM, 2008, vol. 16

2008

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

A review on recent development of spacecraft attitude fault tolerant control system,

S. Yin, B. Xiao, S. X. Ding, and D. Zhou, “A review on recent development of spacecraft attitude fault tolerant control system,” IEEE Transactions on Industrial Electronics , vol. 63, no. 5, pp. 3311–3320, 2016

work page 2016

[2] [2]

Adaptive output feedback sliding-mode manoeuvr ing and vibration control of ﬂexible spacecraft with input saturat ion,

Q. Hu, “Adaptive output feedback sliding-mode manoeuvr ing and vibration control of ﬂexible spacecraft with input saturat ion,” IET Control Theory & Applications , vol. 2, pp. 467–478(11), June 2008

work page 2008

[3] [3]

V ariable structure control and active vi bration sup- pression of ﬂexible spacecraft during attitude maneuver,

Q. Hu and G. Ma, “V ariable structure control and active vi bration sup- pression of ﬂexible spacecraft during attitude maneuver,” Aerospace Science and Technology , vol. 9, no. 4, pp. 307–317, 2005

work page 2005

[4] [4]

Three- axis attitude stabilization of a ﬂexible satellite using no n-linear pd controller,

B. G. Baghi, M. Kabganian, R. Nadaﬁ, and E. Arabi, “Three- axis attitude stabilization of a ﬂexible satellite using no n-linear pd controller,” Transactions of the Institute of Measurement and Control , vol. 40, pp. 591 – 605, 2018

work page 2018

[5] [5]

Appl ication of piezoelectric devices to vibration suppression,

C. C. Won, J. L. Sulla, D. W. Sparks, and W. K. Belvin, “Appl ication of piezoelectric devices to vibration suppression,” Journal of Guidance, Control, and Dynamics , vol. 17, no. 6, pp. 1333–1338, 1994

work page 1994

[6] [6]

Robust h∞ output feedback control for attitude stabilization of a ﬂexible spacecraft,

S. Wu and S. Wen, “Robust h∞ output feedback control for attitude stabilization of a ﬂexible spacecraft,” Nonlinear Dynamics , vol. 84, pp. 405–412, 2016

work page 2016

[7] [7]

Energy-based co ntrol of axially translating beams: varying tension, varying speed , and distur- bance adaptation,

K.-J. Y ang, K.-S. Hong, and F. Matsuno, “Energy-based co ntrol of axially translating beams: varying tension, varying speed , and distur- bance adaptation,” IEEE Transactions on Control Systems Technology , vol. 13, no. 6, pp. 1045–1054, 2005

work page 2005

[8] [8]

Output feedback variab le structure adaptive control of a ﬂexible spacecraft,

Y . Zeng, A. Araujo, and S. Singh, “Output feedback variab le structure adaptive control of a ﬂexible spacecraft,” Acta Astronautica , vol. 44, no. 1, pp. 11–22, 1999

work page 1999