Pith. sign in

REVIEW 2 major objections 5 minor 38 references

Data-driven Kernel-based Predictive Control with Stability and Robustness Guarantees

T0 review · 2 major / 5 minor · reviewed 2026-07-12 · grok-4.5

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

Pith's one-line read Kernel multi-step predictors built only from input-output data give recursive feasibility and practical closed-loop stability for nonlinear predictive control when the horizon is long enough and representation error is small enough.

desk verdict Solid recursive-feasibility proofs for kernel multi-step DDPC; the uncertifiable d-bar is a real but standard soft spot, not a collapse of the argument. read the letter →

arxiv 2607.02851 v1 pith:HZ2SNRCG submitted 2026-07-03 eess.SY cs.SY

Wenjie Liu , Yifei Li , Gang Wang , Lihua Xie This is my paper · ORCID
classification eess.SYcs.SY
keywords data-drivencontrolkernelmethodsmodelpredictivenonlinearsystemsrecursivefeasibilitypracticalstabilityonlinedictionaryapproximatelineardependence
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 develops a data-driven kernel-based predictive controller that never needs an explicit state-space model of an unknown nonlinear plant. Past input-output windows are turned into a multi-step predictor via the representer theorem, so every future output element is written as a finite kernel expansion whose residual is bounded. When that residual is small and the prediction horizon is long enough, the resulting receding-horizon scheme is shown to be recursively feasible and practically stable about a desired equilibrium, without terminal costs or terminal sets. The same argument is extended to measurement noise by folding noise and representation error into one uncertainty radius, and to slowly time-varying plants by refreshing a fixed-size kernel dictionary online under an approximate-linear-dependence rule. The result matters because it supplies the first closed-loop guarantees for kernelized nonlinear data-driven MPC that work from pure input-output trajectories and remain valid under mild non-stationarity.

What carries the argument

The implicit multi-step kernel predictor obtained from the representer theorem: each future output coordinate is written as a linear combination of kernel evaluations on the offline (or online) dictionary, with a deterministic residual bound that is treated as a single additive uncertainty inside a robust data-driven MPC problem.

What would settle it

On a compact nonlinear plant for which a high-fidelity multi-step residual bound can be computed offline, run the controller with a horizon shorter than the derived threshold or with an intentionally larger residual; recursive feasibility or the practical Lyapunov decrease should then fail.

Watch

Extended reading notes

Core claim

For noise-free data the DDKPC optimization is recursively feasible and the closed-loop system is practically stable whenever the prediction horizon exceeds an explicit length that depends only on the sublevel set of the cost and the kernel residual is smaller than a matching threshold; the same practical Lyapunov decrease continues to hold after the residual is enlarged to absorb bounded measurement noise or the bounded-rate drift of a slowly time-varying plant.

Load-bearing premise

Every multi-step input-output map must live in a known reproducing-kernel Hilbert space whose norm is bounded by a known constant, so that a uniform residual size can be certified before the controller is run.

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

2 major / 5 minor

Summary. The paper develops a data-driven kernel-based predictive control (DDKPC) scheme for unknown nonlinear input-output systems. A multi-step predictor is constructed from offline (and later online) trajectories via the representer theorem; the resulting nonconvex program is analyzed without terminal ingredients. Theorem 1 establishes recursive feasibility and practical closed-loop stability for the noise-free case when the horizon is long enough and the uniform kernel-error bound is small enough. The same Lyapunov argument is extended to measurement noise via a unified uncertainty bound (Theorem 2) and to slowly time-varying plants via an ALD-managed fixed-budget dictionary with periodic predictor updates (Theorem 3). A penalty-relaxation formulation is proposed for real-time solution, and two numerical examples (1-D vehicle, payload-carrying quadruped) illustrate tracking performance.

Significance. If the claims hold, the paper supplies one of the first closed-loop recursive-feasibility and practical-stability certificates for a kernelized multi-step DDPC scheme that does not rely on terminal ingredients or full-state measurements. The online ALD extension and the unified-noise robustification are natural and useful increments over existing kernelized DDPC and GP-MPC literature. The candidate-solution construction in the proof of Theorem 1 is explicit and follows standard practical-MPC reasoning; the numerical examples (especially the contact-rich quadruped residual controller) give concrete evidence that the formulation can be implemented. The main limitation is that the stability margins remain non-constructive with respect to the kernel-error bound, which is standard but should be stated more carefully for a data-driven claim.

major comments (2)
  1. Theorem 1 (and likewise Theorems 2–3) asserts recursive feasibility and practical stability once L > L_Ȳ and d̄ ≤ d0. The proof constructs a candidate whose cost excess is bounded by α_Y(d̄) (see (24), (28) and the definition of α_Y after (28)). That excess is absorbed only when d̄ is smaller than a threshold that itself depends on the unknown RKHS-norm bound Γ of Assumption 2 and the uniform bound of Assumption 3. Remark 2 candidly admits both quantities are “generally intractable” for black-box plants. Consequently the existence claim is non-constructive: from data alone one cannot verify d̄ ≤ d0 nor compute a concrete L_Ȳ that guarantees the claimed decrease. The same uncertifiable scale reappears as w̄ and d̂. The paper should either (i) supply a data-driven upper bound / validation procedure for d̄ (or an a-posteriori residual check that certifies the Lyapunov decrease online), or (
  2. Section IV-A replaces the hard kernel constraint (19b) by a quadratic penalty and then assumes a numerical solver returns a point whose objective suboptimality δ_J and residual δ_K are “sufficiently small” to be absorbed into α_Y. No quantitative relation is given between (δ_J, δ_K) and the constants that appear in a_L and α_Y of Theorem 1. Because the program remains nonconvex, global optimality is not guaranteed; the closed-loop certificate therefore rests on an unstated numerical-accuracy hypothesis. A short corollary that makes the admissible (δ_J, δ_K) explicit in terms of the Lyapunov margin would close this gap.
minor comments (5)
  1. Assumption 1 (finite-memory input-output realization) is classical but non-trivial for general nonlinear systems; a short pointer to concrete classes (NARX, systems with well-defined relative degree) already present in Remark 1 could be moved into the assumption statement itself for clarity.
  2. In the vehicle example the stacked input cost uses both ru and rΔu, while the theoretical stage cost is separable; a one-sentence remark that the analysis extends verbatim to any positive-definite quadratic form on the stacked input would avoid confusion.
  3. Figure 3 (middle panel) shows that tracking error is non-monotonic in the update period T0; a brief discussion of how T0 should be chosen in practice would strengthen the online section.
  4. Notation for the stacked windows (x[t1,t2]) and the multi-index a of the multi-step maps is dense; a short table of symbols would help the reader.
  5. The conference precursor [27] is cited; it would be useful to state explicitly which theorems are new relative to that version (the robust and online results appear to be the main additions).

Circularity Check

1 steps flagged · score 1.0 of 10

No significant circularity: closed-loop guarantees are conditional on independent Lyapunov/UIOSS assumptions and an external kernel-error bound that is not fitted to the claimed decrease.

  1. self citation load bearing [Introduction, paragraph on preliminary conference version; also Lemma 2 citation]
    "A preliminary conference version of this paper was submitted to IFAC conference [27]. Compared with the conference version, this paper makes three substantial extensions. First, it provides a detailed proof of the nominal recursive feasibility and practical stability result, which was only outlined in the conference version due to space limitations."

    The conference outline is cited for the high-level claim, but the present paper supplies the complete proof of Theorem 1 (and extensions). The citation is therefore not load-bearing for the derivation; it is ordinary priority acknowledgment and does not make the stability margin circular.

full rationale

The derivation chain for Theorem 1 (and its robust/online extensions) proceeds from the representer-based multi-step predictor (13), the deterministic error bound of Lemma 1 (external citation), and the slack-augmented OCP (19). Recursive feasibility and practical Lyapunov decrease (22) are then obtained by constructing a candidate solution that shifts the prior optimum and appends a local stabilizer from Assumption 5; the resulting cost excess is bounded by class-K functions of the a-priori uniform error d-bar (Assumptions 2–3). Those assumptions are independent of the claimed decrease and are not obtained by fitting the target residual. The same structure holds for Theorems 2–3 after absorbing noise or slow variation into a unified bound. Self-citations appear only for the conference outline [27] and related DDPC formulations; the full proofs are self-contained in the present manuscript and do not reduce the stability claim to an unverified prior result by the same authors. No equation equates a predicted quantity to a fitted constant by construction, and no uniqueness or ansatz is smuggled in to force the result. The uncertifiability of d-bar (Remark 2) is a verifiability issue, not circularity.

Assumptions & free parameters 4 free parameters · 7 assumptions · 2 invented entities

The central stability claims rest on a chain of domain assumptions (finite-memory realization, RKHS membership with known norm bound, local stabilizability, UIOSS, bounded noise, slow time-variation) plus several free design parameters (horizon, regularization weights, kernel hyperparameters, dictionary budget). No new physical entities are postulated; the invented objects are algorithmic (the DDKPC problem and the ALD dictionary).

free parameters (4)
  • prediction horizon L
    Must exceed an a-priori lower bound L_Y-bar that depends on unknown Lyapunov constants; chosen by the designer and critical for a_L > 0.
  • kernel regularization γ and hyperparameters (σ_f, ℓ)
    Fitted or hand-tuned; they determine both the Gram matrix and the size of the representation error d-bar that enters every residual term.
  • penalty weights λ_h, λ_g and uncertainty scale d-bar (or w-bar, d-hat)
    Hand-chosen scaling factors that balance stage cost against slack and g-regularization; appear directly in the practical Lyapunov residual α_Y.
  • ALD novelty threshold ν and dictionary budget M_dict
    Control which online samples are retained; affect the online prediction error bound d_on that must remain small for Theorem 3.
assumptions (7)
  • domain assumption Finite-memory input-output realization of length η exists (Assumption 1 / 8)
    Required to replace the unknown state by a window of past inputs and outputs; classical for NARX systems but not universal.
  • domain assumption Each multi-step map q[a] lies in a known RKHS with ||q[a]||_H ≤ Γ (Assumption 2)
    Enables the representer theorem and the deterministic error bound of Lemma 1; Γ is treated as known yet generally uncomputable.
  • domain assumption Uniform bound d(ζ) ≤ d-bar on the kernel representation error (Assumption 3)
    Used to construct candidate solutions and to bound the practical residual; Remark 2 notes that only an order-of-magnitude estimate is needed in practice.
  • domain assumption Existence of a local Lipschitz feedback κ and Lyapunov function V_s satisfying the decrease (17) (Assumption 5 / 11)
    Used only for analysis to prove local feasibility of the candidate solution; not required for online implementation.
  • domain assumption Uniform input-output-to-state stability Lyapunov function V_o (Assumption 6 / 12)
    Supplies the lower bound on the composite Lyapunov function Y_L and the one-step decrease used throughout the proofs.
  • domain assumption Bounded measurement noise ||n_t|| ≤ n-bar (Assumption 7)
    Allows aggregation of noise and representation error into a single uncertainty bound w-bar.
  • domain assumption Bounded rate of variation of the multi-step maps (Assumption 9)
    Controls the mismatch between frozen and true time-varying predictors over one update period T_0.
invented entities (2)
  • DDKPC optimization problem (19)/(30)/(38)
    purpose: Encodes the kernel multi-step predictor, input constraints and regularization into a single receding-horizon program whose optimal first move is applied to the plant.
    The problem statement itself is the algorithmic contribution; it is not a physical entity and has no independent existence outside the paper.
  • ALD-managed fixed-budget online dictionary independent evidence
    purpose: Keeps the kernel Gram matrix of bounded size while admitting only sufficiently novel closed-loop samples.
    Standard ALD criterion is reused; the paper’s novelty is its periodic freeze-and-rebuild schedule inside the DDKPC loop.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Data-driven Kernel-based Predictive Control with Stability and Robustness Guarantees." pith.science (2026). https://pith.science/paper/HZ2SNRCG

@misc{pith2026260702851,
  author       = {Pith},
  title        = {Pith review of: Data-driven Kernel-based Predictive Control with Stability and Robustness Guarantees},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HZ2SNRCG}},
  note         = {Machine review of arXiv:2607.02851}
}
read the original abstract

In this paper, we provide a theoretical analysis of the closed-loop properties of a data-driven kernel-based predictive control (DDKPC) scheme developed solely from input-output data. The proposed formulation integrates a robust data-driven predictive control framework with a multi-step predictor for nonlinear systems constructed via kernel-based methods. This predictor implicitly captures the system's nonlinear behavior using the representer theorem. For the nominal case with noise-free data, we prove that the DDKPC scheme guarantees recursive feasibility and closed-loop stability, provided that the prediction horizon is sufficiently long and the kernel representation error is sufficiently small. To facilitate real-time implementation, we introduce a penalty relaxation formulation to alleviate the computational burden inherently caused by nonconvex implicit constraints. Furthermore, the framework is robustified against measurement noise by aggregating the representation mismatch and the bounded noise into a unified uncertainty bound. Finally, we extend the DDKPC framework to slowly time-varying nonlinear systems by periodically reconstructing the kernel predictor from a fixed-budget online dictionary managed by the approximate linear dependency (ALD) criterion. Under suitable conditions on the rate of variation of the input-output evolution and the online prediction error, recursive feasibility and practical closed-loop stability are preserved. The effectiveness of the proposed approach is illustrated through numerical examples.

Figures

Figures reproduced from arXiv: 2607.02851 by the authors.

Figure 1
Figure 1. Closed-loop input and output trajectories under Algorithm 1, comparing [PITH_FULL_IMAGE:figures/full_fig_p013_1.png] view at source ↗
Figure 2
Figure 2. Time-varying parameters c1(t) and c2(t). the standard linear DDPC, the kernel-based constraint (19b) is replaced by the standard Hankel matrix formulation: h Hη+L(u d ) Hη+L(y d ) i gt = h u¯[0,L−1]|t y¯[0,L−1]|t+h[0,L−1]|t i . (56) [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗
Figure 4
Figure 4. Velocity control architecture and PyBullet simulation of the payload [PITH_FULL_IMAGE:figures/full_fig_p014_4.png] view at source ↗
Figures from the paper (2 more)
Figure 6
Figure 6. Figure 6: Forward-velocity tracking under a payload-switching condition: static [PITH_FULL_IMAGE:figures/full_fig_p015_6.png]
Figure 7
Figure 7. Figure 7: Planar body-velocity tracking with a fixed [PITH_FULL_IMAGE:figures/full_fig_p015_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 2 linked inside Pith

  1. [1]

    J. B. Rawlings, D. Q. Mayne, and M. M. Diehl,Model Predictive Control: Theory and Design, 2nd ed. Nob Hill Publishing, 2019

    2019

  2. [2]

    Efficient and robust time-optimal trajectory planning and control for agile quadrotor flight,

    Z. Zhou, G. Wang, J. Sun, J. Wang, and J. Chen, “Efficient and robust time-optimal trajectory planning and control for agile quadrotor flight,” IEEE Robot. Autom. Lett., vol. 8, no. 12, pp. 7913–7920, 2023

    2023

  3. [3]

    A note on persistency of excitation,

    J. C. Willems, I. Markovsky, P. Rapisarda, and B. L. M. De Moor, “A note on persistency of excitation,”Syst. Control Lett., vol. 56, no. 4, pp. 325–329, May, 2005

    2005

  4. [4]

    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 Trans. Autom. Control, vol. 65, no. 3, pp. 909–924, Mar. 2020

    2020

  5. [5]

    Data-enabled policy optimization for direct adaptive learning of the LQR,

    F. Zhao, F. D ¨orfler, A. Chiuso, and K. You, “Data-enabled policy optimization for direct adaptive learning of the LQR,”IEEE Trans. Autom. Control, vol. 70, no. 11, pp. 7217–7232, 2025

    2025

  6. [6]

    Koopman-based data-driven control of networked nonlinear system under an event-triggered communication,

    X. Wang, M. Shen, L. Li, X. Zhao, Q.-G. Wang, and Z. H. Zhu, “Koopman-based data-driven control of networked nonlinear system under an event-triggered communication,”IEEE Trans. Consum. Electron., vol. 72, no. 1, pp. 938–944, 2026

    2026

  7. [7]

    Data-driven control for multi-input systems with partly unknown models under aperiodic sampling,

    Z. Yuan, G. Chen, T. Yang, J. Xia, and J. H. Park, “Data-driven control for multi-input systems with partly unknown models under aperiodic sampling,”IEEE Internet Things J., vol. 13, no. 12, pp. 27 130–27 141, 2026

    2026

  8. [8]

    Data-driven tracking control design with noisy data and its application in DC microgrid control,

    Z.-J. Wei, K.-Z. Liu, P. Lin, and X.-M. Sun, “Data-driven tracking control design with noisy data and its application in DC microgrid control,”IEEE Trans. Power Electron., vol. 40, no. 12, pp. 17 776–17 788, 2025

    2025

Show all 38 references
  1. [9]

    Data-driven mode detection and stabilization of unknown switched linear systems,

    J. Eising, S. Liu, S. Mart ´ınez, and J. Cort´es, “Data-driven mode detection and stabilization of unknown switched linear systems,”IEEE Trans. Autom. Control, vol. 70, no. 6, pp. 3830–3845, 2024

    2024

  2. [10]

    Data-enabled predictive control: In the shallows of the DeePC,

    J. Coulson, J. Lygeros, and F. D ¨orfler, “Data-enabled predictive control: In the shallows of the DeePC,” inProc. Eur. Control Conf., Naples, Italy, Italy, June, 25-28 2019, pp. 307–312

    2019

  3. [11]

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

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

    2021

  4. [12]

    Robust data-driven predictive control for unknown linear systems with bounded disturbances,

    K. Hu and T. Liu, “Robust data-driven predictive control for unknown linear systems with bounded disturbances,”IEEE Trans. Autom. Control, vol. 70, no. 10, pp. 6529–6544, 2025

    2025

  5. [13]

    Data-driven control from poisoned data: Fundamental limitations and secure DeePC,

    T. Shinohara, H. Sandberg, and K. H. Johansson, “Data-driven control from poisoned data: Fundamental limitations and secure DeePC,”arXiv preprint arXiv:2606.20301, 2026

    2026 arXiv

  6. [14]

    Data-driven resilient predictive control under denial-of-service,

    W. Liu, J. Sun, G. Wang, F. Bullo, and J. Chen, “Data-driven resilient predictive control under denial-of-service,”IEEE Trans. Autom. Control, vol. 68, no. 8, pp. 4722–4737, Aug. 2023. 16

    2023

  7. [15]

    Data-driven self-triggered control via trajectory prediction,

    ——, “Data-driven self-triggered control via trajectory prediction,”IEEE Trans. Autom. Control, vol. 68, no. 11, pp. 6951–6958, Nov. 2023

    2023

  8. [16]

    Data-driven control of network systems: Accounting for communication adaptivity and security,

    G. Wang, W. Liu, Y . Li, X. Wang, J. Sun, and J. Chen, “Data-driven control of network systems: Accounting for communication adaptivity and security,”Sci. China Inf. Sci., vol. 69, no. 2, p. 121201, 2026

    2026

  9. [17]

    Linear tracking MPC for nonlinear systems—part II: The data-driven case,

    J. Berberich, J. K ¨ohler, M. A. M ¨uller, and F. Allg ¨ower, “Linear tracking MPC for nonlinear systems—part II: The data-driven case,”IEEE Trans. Autom. Control, vol. 67, no. 9, pp. 4406–4421, 2022

    2022

  10. [18]

    Data-driven nonlinear predictive control for feedback linearizable systems,

    M. Alsalti, V . G. Lopez, J. Berberich, F. Allg ¨ower, and M. A. M ¨uller, “Data-driven nonlinear predictive control for feedback linearizable systems,”IFAC-PapersOnLine, vol. 56, no. 2, pp. 617–624, 2023

    2023

  11. [19]

    Neural data–enabled predictive control,

    M. Lazar, “Neural data–enabled predictive control,”IFAC-PapersOnLine, vol. 58, no. 15, pp. 91–96, 2024

    2024

  12. [20]

    Robust and kernelized data- enabled predictive control for nonlinear systems,

    L. Huang, J. Lygeros, and F. D ¨orfler, “Robust and kernelized data- enabled predictive control for nonlinear systems,”IEEE Trans. Control Syst. Technol., vol. 32, no. 2, pp. 611–624, 2023

    2023

  13. [21]

    Data- driven predictive control for a class of nonlinear systems with polynomial terms,

    A. Azarbahram, M. Al Khatib, V . K. Mishra, and N. Bajcinca, “Data- driven predictive control for a class of nonlinear systems with polynomial terms,”IFAC-PapersOnLine, vol. 58, no. 21, pp. 226–231, 2024

    2024

  14. [22]

    Data-enabled predictive control for nonlinear systems based on a koopman bilinear realization,

    Z. Xiong, Z. Yuan, K. Miao, H. Wang, J. Cort ´es, and A. Pa- pachristodoulou, “Data-enabled predictive control for nonlinear systems based on a koopman bilinear realization,” inProc. IEEE Conf. Decis. Control, Rio de Janeiro, Brazil, Dec. 09-12 2025, pp. 6498–6503

    2025

  15. [23]

    A generalized representer theorem,

    B. Sch ¨olkopf, R. Herbrich, and A. J. Smola, “A generalized representer theorem,” inProc. Int. Conf. Comput. Learn. Theory. Cham, Switzerlend: Springer, 2001, pp. 416–426

    2001

  16. [24]

    Online data-driven adaptive control for unknown linear time-varying systems,

    S. Liu, K. Chen, and J. Eising, “Online data-driven adaptive control for unknown linear time-varying systems,” inProc. IEEE Conf. Decis. Control, Singapore, Singapore, Dec. 13-15 2023, pp. 8775–8780

    2023

  17. [25]

    Online learning-based model predictive control with Gaussian process models and stability guarantees,

    M. Maiworm, D. Limon, and R. Findeisen, “Online learning-based model predictive control with Gaussian process models and stability guarantees,” Int. J. Robust Nonlin. Control, vol. 31, no. 18, pp. 8785–8812, 2021

    2021

  18. [26]

    Gaussian processes for dynamics learning in model predictive control,

    A. Scampicchio, E. Arcari, A. Lahr, and M. N. Zeilinger, “Gaussian processes for dynamics learning in model predictive control,”Annu. Rev. Control, vol. 60, p. 101034, 2025

    2025

  19. [27]

    Data-driven kernel-based predictive control with closed-loop guarantees,

    W. Liu, Y . Li, G. Wang, and L. Xie, “Data-driven kernel-based predictive control with closed-loop guarantees,” inProc. IFAC WC, Busan, Republic of Korea, Aug. 23-28 2026

    2026

  20. [28]

    Deterministic error bounds for kernel-based learning techniques under bounded noise,

    E. T. Maddalena, P. Scharnhorst, and C. N. Jones, “Deterministic error bounds for kernel-based learning techniques under bounded noise,” Automatica, vol. 134, p. 109896, 2021

    2021

  21. [29]

    Theory of reproducing kernels,

    N. Aronszajn, “Theory of reproducing kernels,”Trans. Am. Math. Soc., vol. 68, no. 3, pp. 337–404, 1950

    1950

  22. [30]

    Optimal kernel regression bounds under energy-bounded noise,

    A. Lahr, J. K ¨ohler, A. Scampicchio, and M. N. Zeilinger, “Optimal kernel regression bounds under energy-bounded noise,”arXiv preprint arXiv:2505.22235, 2025

    2025

  23. [31]

    Isidori,Nonlinear control systems

    A. Isidori,Nonlinear control systems. London, UK: Springer, 1995

    1995

  24. [32]

    Learning controllers from data via kernel-based interpolation,

    Z. Hu, C. De Persis, and P. Tesi, “Learning controllers from data via kernel-based interpolation,” inIEEE Conf. Decision Control. IEEE, 2023, pp. 8509–8514

    2023

  25. [33]

    Input–output-to-state stability for discrete-time systems,

    C. Cai and A. R. Teel, “Input–output-to-state stability for discrete-time systems,”Automatica, vol. 44, no. 2, pp. 326–336, Sept. 2008

    2008

  26. [34]

    Data-driven stabilizer design and closed-loop analysis of general nonlinear systems via Taylor’s expansion,

    M. Guo, C. De Persis, and P. Tesi, “Data-driven stabilizer design and closed-loop analysis of general nonlinear systems via Taylor’s expansion,” arXiv:2209.01071, 2022

    2022 arXiv

  27. [35]

    Optimization algorithms with superlinear convergence rate,

    H. Wang, Y . Xu, Z. Guo, and H. Zhang, “Optimization algorithms with superlinear convergence rate,”IEEE Trans. Autom. Control, vol. 71, no. 1, pp. 684–691, 2026

    2026

  28. [36]

    Input-to-state stability for nonlinear time- varying systems via averaging,

    D. Ne ˇsi´c and A. R. Teel, “Input-to-state stability for nonlinear time- varying systems via averaging,”Math. Control Signals Systems, vol. 14, no. 3, pp. 257–280, Sep. 2001

    2001

  29. [37]

    Kernel-based least squares policy iteration for reinforcement learning,

    X. Xu, D. Hu, and X. Lu, “Kernel-based least squares policy iteration for reinforcement learning,”IEEE Trans. Neural. Netw., vol. 18, no. 4, pp. 973–992, 2007

    2007

  30. [38]

    Learning a contact-adaptive controller for robust, efficient legged locomotion,

    X. Da, Z. Xie, D. Hoeller, B. Boots, A. Anandkumar, Y . Zhu, B. Babich, and A. Garg, “Learning a contact-adaptive controller for robust, efficient legged locomotion,” inProc. Conf. robot learning. PMLR, 2021, pp. 883–894

    2021

Pith tools

Reviewed July 12, 2026 · model on record in the stance chip above.