Pith. sign in

REVIEW 3 major objections 5 minor 80 references

Two-component controller design to safeguard data-driven predictive control

T0 review · 3 major / 5 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 For output tracking on second-order control-affine systems, a model-free funnel controller guarantees the prescribed error bound even while a data-driven predictive controller is learning, provided the predictive input stays bounded.

desk verdict A tutorial-style paper that applies the authors' funnel-safeguard architecture to EDMD-MPC and DeePC; the fill-distance data-collection idea is genuinely new but explicitly unfinished, and Theorem 1's proof is a sketch that needs tightening. read the letter →

arxiv 2505.19131 v1 pith:BNGPUSUK submitted 2025-05-25 math.OC

Lea Bold , Lukas Lanza , Karl Worthmann This is my paper · ORCID
classification math.OC MSC 93C1093B5293C5793D15
keywords two-componentcontrollerfunnelcontroldata-drivenpredictivedata-enabled(DeePC)EDMD-basedMPCKoopmanoperatorfilldistancesafelearning
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 proposes a two-component controller for output tracking with prescribed error bounds on a class of second-order control-affine systems, with no model of the plant required. One component is a data-driven predictive controller, exemplified by data-enabled predictive control (DeePC) and by model predictive control built on extended dynamic mode decomposition (EDMD); the other is a model-free funnel controller that an event-triggered activation function switches on only when the tracking error leaves a safe region. The central claim, Theorem 1, is that this combination keeps the tracking error strictly inside the prescribed funnel for all time, no matter how poor the learned model is, as long as the predictive input is bounded and the initial auxiliary errors start inside the funnel. The same funnel mechanism doubles as a data collector: steering along a reference that visits prescribed virtual observation points keeps samples within a chosen cluster radius, so the fill distance needed for kernel-EDMD approximation error bounds can be achieved. This is a route to safe learning-based predictive control without an offline training phase.

What carries the argument

The load-bearing mechanism is the two-component feedback law $u = \mu(t,x) + a_\tau(t,e_2) u_{\mathrm{FC}}(t)$, with the model-free funnel law $u_{\mathrm{FC}} = -e_2/(1 - \|e_2\|^2)$ built from the scaled error variables $e_1 = \sigma(t)(y - y_{\mathrm{ref}})$ and $e_2 = \sigma(t)\dot{e} + e_1/(1 - \|e_1\|^2)$. The activation function $a_\tau = \max\{0, \max_{s \in [t-\tau,t]} \|e_2(s)\| - \lambda\}$, with threshold $\lambda \in (0,1)$ and dwell time $\tau > 0$, keeps the safeguard switched off in the safe region and prevents chattering when it switches on. The funnel feedback is high-gain: as $\|e_2\|$ approaches 1 the denominator vanishes, which forces the auxiliary variables back into the unit ball and hence keeps $\|y(t) - y_{\mathrm{ref}}(t)\| < 1/\sigma(t)$. For the EDMD instance, the second key object is the bilinear surrogate $x^+ = [I_n, 0] K_u^{\Delta t} \Psi(x)$, formed by lifting states into observables and approximating the Koopman operator from data; the fill distance $h_\mathcal{X}$ and cluster radius $\varepsilon_c$ of the collected samples certify the surrogate's approximation error.

What would settle it

Run the set-point transition of the forced oscillator example of Section 5 with an EDMD dictionary chosen to be deliberately poor, for instance one that omits the coordinate functions or uses a kernel whose fill distance exceeds the threshold of Theorem 3; if the computed input $\mu$ grows without bound and the closed-loop output leaves the prescribed funnel while Assumptions 1 and 2 hold and the initial auxiliary errors are inside the unit ball, then the boundedness premise of Theorem 1 is violated in practice.

Watch

Extended reading notes

Core claim

For systems of the form $\dot{x}_1 = x_2$, $\dot{x}_2 = g_0(x) + G(x)u$, $y = x_1$ with sign-definite input distribution $G$, the paper claims that the two-component feedback $u = \mu + a_\tau u_{\mathrm{FC}}$, where $\mu$ is any bounded input produced by a data-driven predictive controller and $u_{\mathrm{FC}}$ is the funnel feedback, guarantees $\|y(t) - y_{\mathrm{ref}}(t)\| < 1/\sigma(t)$ for all $t \geq 0$ whenever the funnel function and reference are feasible and the auxiliary variables $e_1, e_2$ start inside the unit ball. The proof is given only as a sketch and deferred to prior work: outside the safe $\lambda$-region the funnel feedback acts with increasing gain as $\|e_2\|$ approaches 1, driving the auxiliary variables back inside, while inside the safe region the activation function keeps the funnel inactive. The paper also claims that this safeguarding property supports safe online learning: EDMD-based MPC can start with a single data point and still track a set-point transition within the prescribed error bounds, with the funnel controller active only briefly during the transient. Finally, the same feedback mechanism is used to steer the system so that samples land near prescribed virtual observation points, making the fill-distance-dependent kernel-EDMD approximation error bound applicable.

Load-bearing premise

The theorem depends on the predictive controller's input being bounded, an assumption the paper states rather than derives from the learned surrogate and the optimization that computes the input, and it also presumes the reference trajectory can be routed through the prescribed virtual observation points at sampling instants.

Editorial extensions

If this is right

  • For any data-driven predictive controller whose input is bounded, output constraint satisfaction holds during runtime, so the safeguarded schemes need no offline training phase.
  • Steering the system along a reference through prescribed virtual observation points with $\sigma \geq 3/\varepsilon_c$ keeps sampled states within $\varepsilon_c$ of those points, so the fill-distance-dependent error bounds of Theorem 3 apply to the learned EDMD surrogate.
  • The EDMD-based MPC can be initialized with a single data point and still complete a set-point transition inside the prescribed error bounds, with the funnel controller activating only a few times.
  • DeePC combined with the funnel controller inherits the same safeguarding guarantee, and the data needed for its Hankel matrices can be collected online rather than in advance.
  • Because the funnel component is model-free and independent of the prediction scheme, the two-component architecture applies to any learning-based predictive controller, not just the two examples worked out here.

Reading between the lines

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

  • A natural hardening step would be to make the boundedness of $\mu$ a certified property, for example by adding hard input constraints and proving feasibility of the MPC problem under the surrogate's error bounds; without that, the theorem's main assumption is left to the optimizer's good behavior.
  • The fill-distance argument turns data collection into a planning problem: choose virtual observation points and a reference so that a target approximation error is reached with a minimal number of samples, and use the activation function's record of interventions as a data-quality flag.
  • The safety argument is likely not tied to funnel control specifically: prescribed performance control and other high-gain safeguards for the same system class should combine with the same activation function, so the architecture is a template for safe learning rather than a single controller.
  • A quantitative trade-off worth testing is between exploration speed and data quality: larger $\sigma$ shrinks the neighborhood visited around the reference, making fill-distance control easier but slowing the coverage of the state-space region.
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.

Referee Report

3 major / 5 minor

Summary. The paper proposes a two-component feedback architecture for output tracking of control-affine systems of relative degree two: a learning-based predictive controller (exemplified by DeePC and by EDMD-based MPC) provides the nominal input, and a model-free funnel controller is added through an activation function that switches it on only when a scaled error variable leaves a safe λ-region. Theorem 1 claims that, whenever the predictive input is bounded, the closed loop satisfies the prescribed error bound ||y(t)-y_ref(t)|| < 1/σ(t) for all time. The paper further sketches a fill-distance-based data collection procedure intended to guarantee a desired approximation accuracy for kernel-EDMD surrogates, and illustrates the approach numerically on a forced Van der Pol oscillator.

Significance. If the safeguarding property were fully proven and the boundedness/feasibility of µ were enforced, the architecture would be a useful step toward safe learning-based predictive control without offline training: it combines two established components in a modular way, and the explicit link to fill-distance guarantees for kernel EDMD is original. The paper is clearly written and the numerical examples illustrate the behavior. However, the main theoretical claim is only sketched and relies on assumptions not certified by the algorithms, so the central 'safe learning' guarantee is not yet established in the manuscript. No machine-checked proofs, parameter-free derivations, or reproducibility artifacts are provided; the numerical results are illustrative only.

major comments (3)
  1. [Section 3, Theorem 1 and proof sketch] The proof of Theorem 1 is not self-contained: it invokes [31, Thm. 1], [10, Thm. 1.9] and [46, Thm. 5.1], none of which includes the activation factor aτ defined in (6). In the critical zone ||e2(t)|| ∈ [λ, 1), aτ(t, e2) = max{0, max_{s∈[t−τ,t]} ||e2(s)|| − λ} can be arbitrarily small: for an error that has just entered the zone, the most recent maximum is close to ||e2(t)||, so aτ is close to zero at the boundary. The standard funnel-control proof dominates a bounded disturbance by the full gain −e2/(1−||e2||^2); with the gain multiplied by a vanishing factor, that domination argument does not go through. The claim that there exists ε̃ ∈ [λ, 1) with ||ei(t)|| ≤ ε̃ for i = 1, 2 requires a positive lower bound on aτ during the activation intervals, which the manuscript does not provide. This gap is load-bearing because all subsequent safe-learning statements rest on Theorem 1. If the result is already proven under these conditions in the authors' earlier work, the manuscript should state that explicitly and give the precise theorem mapping; as written, the proof is only a sketch.
  2. [Section 2.2, Algorithm 1, and Eq. (8)] Theorem 1 assumes that the predictive input µ is bounded, but neither controller component enforces this. Algorithm 1 solves a finite-horizon OCP in Step 2 without a feasibility certificate; if the EDMD surrogate is poor, the OCP may be infeasible, in which case no µ is produced and control law (4) is undefined. The DeePC formulation (8) imposes the input bound ||u(i)|| ≤ umax, but it does not prove that the optimization is feasible at every step, nor that the resulting µ satisfies the boundedness hypothesis of Theorem 1. The numerical example imposes U = [−2, 2] in Algorithm 1, but this is an implementation choice not covered by the theorem's assumptions. To substantiate the central claim, the authors should either add an explicit fallback (e.g., µ = 0 whenever the OCP is infeasible) and include it in Theorem 1, or prove recursive feasibility and an input bound for the chosen predictive component.
  3. [Section 4.2.3, Fig. 3] The fill-distance data-collection scheme is not a theorem. The crucial step 'Define reference yref and sampling time Δt such that S_{i=0}^D {(yref(iΔt), yrefdot(iΔt))} = X' is an assumption on the design, not a construction, and the manuscript itself states in Section 6 that the scheme 'will be thoroughly analyzed' in future research. Equation (14) bounds ||x(t) − (yref(t), yrefdot(t))|| by 3/σ; to conclude x(iΔt) ∈ B_{εc}(x_i), one must also know that the reference evaluated at iΔt equals x_i, which is imposed but not established for a function yref ∈ W^{2,∞}. Thus the statement in Fig. 3 that the EDMD-based surrogate satisfies the bounds of Theorem 3 is premature. The authors should either provide the construction/proof or clearly mark this scheme as a conjecture/outlook rather than part of the established contributions.
minor comments (5)
  1. [Section 3, proof sketch] The proof sketch refers to 'multiplying uFC by ατ', but the notation used elsewhere, including Eq. (6), is aτ; the notation should be made consistent.
  2. [Fig. 3] The set notation 'Sd_i=1' in Fig. 3 should be replaced by a union symbol or explicitly defined, and the parameter D used in the text below the figure should be defined in the figure caption.
  3. [Section 5.2] The reference y_t_ref(t) is defined with a parameter t̂ that is not formally introduced, and the symbols y10_ref and y16_ref in Sections 5.2.1 and 5.2.2 are used without explicit formulas; please clarify their definitions.
  4. [Section 4.2.3] When citing [46, Lem. 2.1] for the bounds on ||e1(t)|| and ||e2(t)||, the text should state explicitly that this lemma concerns the pure funnel controller (5), not the two-component controller (4), to avoid ambiguity.
  5. [Caption of Fig. 9] The caption 'Set-point transition ,initializing with 1 data point' contains a spacing typo and should read 'Set-point transition, initializing with 1 data point'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the safety guarantee is a funnel-control theorem with a bounded exogenous input, not an artifact of the learned model.

full rationale

The central safety claim (Theorem 1) is conditional: it assumes Assumptions 1 and 2, initial conditions inside the funnel, and a bounded predictive input µ, and asserts that the tracking error remains below 1/σ. The proof is only sketched, but the sketch explicitly points to the external funnel-control theorem [10, Thm. 1.9] as the operative mechanism, with [31, Thm. 1] and [46, Thm. 5.1] cited for the activation-function modification and sampled-data aspects. Because the base theorem is external and the data-driven controller is not used to prove safety, the guarantee does not reduce to a fitted parameter or to a definition. The EDMD/DeePC components appear only as instances of the predictive component; their prediction quality is not the source of the safety bound. The fill-distance sampling scheme in Fig. 3 is explicitly described as schematic, and its details are deferred to future research ("These aspects are topics of future research"), so any incompleteness there is a correctness or rigor gap rather than a circular derivation. The boundedness of µ is an assumption, not a derived consequence; if Algorithm 1 can return unbounded or infeasible controls, that undermines the theorem's applicability, but it does not make the theorem circular. No circular step can be exhibited from the paper's own equations or citations.

Assumptions & free parameters 7 free parameters · 6 assumptions · 0 invented entities

The central theorem relies on standard funnel-control assumptions plus an unproven boundedness condition on the predictive input. The fill-distance data-collection scheme adds another unproven premise about the reference trajectory. Tuning parameters in the numerical example are hand-chosen and not fitted to data. No new entities are introduced.

free parameters (7)
  • Activation threshold λ = 0.75
    Chooses the safe region within the funnel; set by hand in Section 5.
  • Dwell-time τ = Δ t/2 = 0.025
    Avoids chattering; set by hand in Section 5.
  • MPC weights Q, R = Q=diag(10^4,1), R=10^-4
    Stage cost weights in Algorithm 1; hand-chosen for the numerical example.
  • Prediction horizon N = 30
    MPC horizon; hand-chosen.
  • Sampling time Δ t = 0.05
    Discretization for EDMD; hand-chosen.
  • Funnel function σ(t) = piecewise: 1/2.3 for t≤4, 1/(2e^{-2(t-4)}+0.3) for t>4
    Prescribed time-varying error bound; hand-chosen.
  • Observable dictionary = monomials of degree ≤3 (10 observables)
    EDMD surrogate quality depends on this choice; no selection rule is given.
assumptions (6)
  • domain assumption Assumption 1: The input distribution matrix G is sign definite (positive definite).
    Required for the funnel control feedback law to work; stated in Section 2.1.
  • domain assumption Assumption 2: The signals y, ẏ, y_ref, and ẏ_ref are continuously available to the controller.
    Stated in Section 3; needed for the continuous-time funnel controller.
  • ad hoc to paper The predictive control input µ is bounded for all times.
    Imposed in Theorem 1 without proof; the EDMD-MPC solution may not be bounded if the surrogate is poor.
  • domain assumption The system model (1) is of relative degree two with no internal dynamics.
    The entire theory is developed for this system class; extensions to higher relative degree are only mentioned in a remark.
  • ad hoc to paper For the fill-distance scheme, the reference trajectory can be designed to pass through prescribed virtual observation points at sampling instants.
    Assumed in Fig. 3; not proven and explicitly left for future research.
  • standard math The kernel-EDMD error bounds of Theorem 2 and Theorem 3 hold under the stated smoothness and fill-distance conditions.
    These are prior results from self-cited references, used as black boxes.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Two-component controller design to safeguard data-driven predictive control." pith.science (2026). https://pith.science/paper/BNGPUSUK

@misc{pith2026250519131,
  author       = {Pith},
  title        = {Pith review of: Two-component controller design to safeguard data-driven predictive control},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BNGPUSUK}},
  note         = {Machine review of arXiv:2505.19131}
}
read the original abstract

We design a two-component controller to achieve reference tracking with output constraints - exemplified on systems of relative degree two. One component is a data-driven or learning-based predictive controller, which uses data samples to learn a model and predict the future behavior of the system. We exemplify this component concisely by data-enabled predictive control (DeePC) and by model predictive control based on extended dynamic mode decomposition (EDMD). The second component is a model-free high-gain feedback controller, which ensures satisfaction of the output constraints if that cannot be guaranteed by the predictive controller. This may be the case, for example, if too little data has been collected for learning or no (sufficient) guarantees on the approximation accuracy derived. In particular, the reactive/adaptive feedback controller can be used to support the learning process by leading safely through the state space to collect suitable data, e.g., to ensure a sufficiently-small fill distance. Numerical examples are provided to illustrate the combination of EDMD-based model predictive control and a safeguarding feedback for the set-point transitions including the transition between the set points within prescribed bounds.

Figures

Figures reproduced from arXiv: 2505.19131 by the authors.

Figure 1
Figure 1. Schematic illustration of tracking error, funnel [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Structure of the two-component controller. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Flowchart illustrating the procedure to gener [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Stabilizing the origin. Visualization of the [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Stabilizing the origin. Visualization of the ac [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 8
Figure 8. Figure 8: Set-point transition, initializing with 1 data [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]
Figure 6
Figure 6. Figure 6: Set-point transition. Visualization of the out [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 9
Figure 9. Figure 9: Set-point transition ,initializing with 1 data [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

80 extracted references · 74 canonical work pages

  1. [1]

    Data-based system analysis and control of flat nonlinear systems

    Mohammad Alsalti, Julian Berberich, Victor G Lopez, Frank Allg¨ ower, and Matthias A M¨ uller. “Data-based system analysis and control of flat nonlinear systems”. In: 60th IEEE Conference on Decision and Control (CDC) . 2021

    work page 2021

  2. [2]

    Control barrier functions: The- ory and applications

    Aaron D Ames, Samuel Coogan, Magnus Egerst- edt, Gennaro Notomista, Koushil Sreenath, and Paulo Tabuada. “Control barrier functions: The- ory and applications”. In: 18th European control conference (ECC). 2019

    work page 2019

  3. [3]

    Two- degree-of-freedom PID controllers

    Mituhiko Araki and Hidefumi Taguchi. “Two- degree-of-freedom PID controllers”. In: Interna- tional Journal of Control, Automation, and Sys- tems 1.4 (2003)

    work page 2003

  4. [4]

    Control-Coherent Koopman Modeling: A Phys- ical Modeling Approach

    H Harry Asada and Jose A Solano-Castellanos. “Control-Coherent Koopman Modeling: A Phys- ical Modeling Approach”. In: arXiv preprint arXiv:2403.16306 (2024)

    arXiv 2024

  5. [5]

    Provably safe and robust learning-based model predictive control

    Anil Aswani, Humberto Gonzalez, S Shankar Sas- try, and Claire Tomlin. “Provably safe and robust learning-based model predictive control”. In: Au- tomatica 49.5 (2013)

    work page 2013

  6. [6]

    Analysis of stable model inversion meth- ods for constrained underactuated mechanical sys- tems

    Guaraci Bastos Jr, Robert Seifried, and Olivier Br¨ uls. “Analysis of stable model inversion meth- ods for constrained underactuated mechanical sys- tems”. In: Mechanism and Machine Theory 111 (2017)

    work page 2017

  7. [7]

    A low-complexity global approximation-free control scheme with pre- scribed performance for unknown pure feedback systems

    Charalampos P Bechlioulis and George A Rovithakis. “A low-complexity global approximation-free control scheme with pre- scribed performance for unknown pure feedback systems”. In: Automatica 50.4 (2014)

    work page 2014

  8. [8]

    Prescribed performance adaptive control for multi-input multi-output affine in the control nonlinear systems

    Charalampos P Bechlioulis and George A Rovithakis. “Prescribed performance adaptive control for multi-input multi-output affine in the control nonlinear systems”. In: IEEE Transactions on automatic control 55.5 (2010)

    work page 2010

Show all 80 references
  1. [9]

    Robust Funnel Model Pre- dictive Control for output tracking with prescribed performance

    Thomas Berger, Dario Dennst¨ adt, Lukas Lanza, and Karl Worthmann. “Robust Funnel Model Pre- dictive Control for output tracking with prescribed performance”. In: SIAM Journal on Control and Optimization 62.4 (2024)

    2024

  2. [10]

    Funnel control of nonlinear systems

    Thomas Berger, Achim Ilchmann, and Eugene P Ryan. “Funnel control of nonlinear systems”. In: Mathematics of Control, Signals, and Systems 33 (2021)

    2021

  3. [11]

    Reinforcement learning and op- timal control

    Dimitri Bertsekas. Reinforcement learning and op- timal control. Vol. 1. Athena Scientific, 2019

    2019

  4. [12]

    Stability and feasibility of state constrained MPC without stabilizing terminal constraints

    Andrea Boccia, Lars Gr¨ une, and Karl Worthmann. “Stability and feasibility of state constrained MPC without stabilizing terminal constraints”. In: Sys- tems & control letters 72 (2014)

    2014

  5. [13]

    Data-driven MPC with stability guarantees using extended dynamic mode decom- position

    Lea Bold, Lars Gr¨ une, Manuel Schaller, and Karl Worthmann. “Data-driven MPC with stability guarantees using extended dynamic mode decom- position”. In: IEEE Transactions on Automatic Control (2024)

    2024

  6. [14]

    Kernel-based Koop- man approximants for control: Flexible sam- pling, error analysis, and stability

    Lea Bold, Friedrich M Philipp, Manuel Schaller, and Karl Worthmann. “Kernel-based Koop- man approximants for control: Flexible sam- pling, error analysis, and stability”. In: Preprint arXiv:2412.02811 (2024)

    2024

  7. [15]

    On Koopman-based surrogate models for non- holonomic robots

    Lea Bold, Mario Rosenfelder, Hannes Es- chmann, Henrik Ebel, and Karl Worthmann. “On Koopman-based surrogate models for non- holonomic robots”. In: IF AC-PapersOnLine58.21 (2024). 11

    2024

  8. [16]

    Kernel EDMD for data-driven nonlinear Koopman MPC with stability guaran- tees

    Lea Bold, Manuel Schaller, Irene Schimperna, and Karl Worthmann. “Kernel EDMD for data-driven nonlinear Koopman MPC with stability guaran- tees”. In: Preprint arxiv:2501.08709 (2025)

    2025 arXiv

  9. [17]

    Ad- vantages of bilinear Koopman realizations for the modeling and control of systems with unknown dy- namics

    Daniel Bruder, Xun Fu, and Ram Vasudevan. “Ad- vantages of bilinear Koopman realizations for the modeling and control of systems with unknown dy- namics”. In: IEEE Robotics and Automation Let- ters 6.3 (2021)

    2021

  10. [18]

    Koopman invariant subspaces and finite linear representations of non- linear dynamical systems for control

    Steven L Brunton, Bingni W Brunton, Joshua L Proctor, and J Nathan Kutz. “Koopman invariant subspaces and finite linear representations of non- linear dynamical systems for control”. In: PloS one 11.2 (2016)

    2016

  11. [19]

    Modern Koopman theory for dynamical systems

    Steven L Brunton, Marko Budiˇ si´ c, Eurika Kaiser, and J Nathan Kutz. “Modern Koopman theory for dynamical systems”. In: SIAM Review 64.2 (2022)

    2022

  12. [20]

    Hamilton–jacobi reachability: Some recent theoretical advances and applications in unmanned airspace management

    Mo Chen and Claire J Tomlin. “Hamilton–jacobi reachability: Some recent theoretical advances and applications in unmanned airspace management”. In: Annual Review of Control, Robotics, and Au- tonomous Systems 1.1 (2018)

    2018

  13. [21]

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

    Jeremy Coulson, John Lygeros, and Florian D¨ orfler. “Data-enabled predictive control: In the shallows of the DeePC”. In: 18th European Con- trol Conference (ECC). 2019

    2019

  14. [22]

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

    Claudio De Persis and Pietro Tesi. “Formulas for data-driven control: Stabilization, optimality, and robustness”. In: IEEE Transactions on Automatic Control 65.3 (2019)

    2019

  15. [23]

    Learning con- trollers for nonlinear systems from data

    Claudio De Persis and Pietro Tesi. “Learning con- trollers for nonlinear systems from data”. In: An- nual Reviews in Control 56.100915 (2023)

    2023

  16. [24]

    Bridging direct and indirect data- driven control formulations via regularizations and relaxations

    Florian D¨ orfler, Jeremy Coulson, and Ivan Markovsky. “Bridging direct and indirect data- driven control formulations via regularizations and relaxations”. In: IEEE Transactions on Automatic Control 68.2 (2022)

    2022

  17. [25]

    Experimental val- idation for the combination of funnel control with a feedforward control strategy

    Svenja Dr¨ ucker, Lukas Lanza, Thomas Berger, Timo Reis, and Robert Seifried. “Experimental val- idation for the combination of funnel control with a feedforward control strategy”. In: Multibody Sys- tem Dynamics (2024)

    2024

  18. [26]

    Behav- ioral theory for stochastic systems? A data-driven journey from Willems to Wiener and back again

    Timm Faulwasser, Ruchuan Ou, Guanru Pan, Philipp Schmitz, and Karl Worthmann. “Behav- ioral theory for stochastic systems? A data-driven journey from Willems to Wiener and back again”. In: Annual Reviews in Control 55 (2023)

    2023

  19. [27]

    Koopman NMPC: Koopman-based learning and nonlinear model predictive control of control-affine systems

    Carl Folkestad and Joel W Burdick. “Koopman NMPC: Koopman-based learning and nonlinear model predictive control of control-affine systems”. In: IEEE International Conference on Robotics and Automation (ICRA). 2021

    2021

  20. [28]

    A com- prehensive survey on safe reinforcement learning

    Javier Garcıa and Fernando Fern´ andez. “A com- prehensive survey on safe reinforcement learning”. In: Journal of Machine Learning Research 16.1 (2015)

    2015

  21. [29]

    Reprojection methods for Koopman-based modelling and prediction

    Pieter van Goor, Robert Mahony, Manuel Schaller, and Karl Worthmann. “Reprojection methods for Koopman-based modelling and prediction”. In: 62nd IEEE Conference on Decision and Control (CDC). 2023

    2023

  22. [30]

    Bilin- earization, reachability, and optimal control of control-affine nonlinear systems: A Koopman spec- tral approach

    Debdipta Goswami and Derek A. Paley. “Bilin- earization, reachability, and optimal control of control-affine nonlinear systems: A Koopman spec- tral approach”. In: IEEE Transactions on Auto- matic Control 67.6 (2021)

    2021

  23. [31]

    Reinforce- ment Learning for Docking Maneuvers with Pre- scribed Performance

    Simon Gottschalk, Lukas Lanza, Karl Worth- mann, and Kerstin Lux-Gottschalk. “Reinforce- ment Learning for Docking Maneuvers with Pre- scribed Performance”. In: IF AC-PapersOnLine 58.17 (2024)

    2024

  24. [32]

    Analysis of unconstrained non- linear MPC schemes with time varying control horizon

    Lars Gr¨ une, J¨ urgen Pannek, Martin Seehafer, and Karl Worthmann. “Analysis of unconstrained non- linear MPC schemes with time varying control horizon”. In: SIAM Journal on Control and Op- timization 48.8 (2010)

    2010

  25. [33]

    Learning-based model predictive control: Toward safe learning in control

    Lukas Hewing, Kim P Wabersich, Marcel Menner, and Melanie N Zeilinger. “Learning-based model predictive control: Toward safe learning in control”. In: Annual Review of Control, Robotics, and Au- tonomous Systems 3.1 (2020)

    2020

  26. [34]

    Koopman form of nonlinear systems with inputs

    Lucian Cristian Iacob, Roland T´ oth, and Maarten Schoukens. “Koopman form of nonlinear systems with inputs”. In: Automatica 162 (2024)

    2024

  27. [35]

    Tracking with prescribed transient behaviour

    Achim Ilchmann, Eugene P Ryan, and Christopher J. Sangwin. “Tracking with prescribed transient behaviour”. In: ESAIM: Control, Optimisation and Calculus of Variations 7 (2002)

    2002

  28. [36]

    Residual reinforcement learning for robot control

    Tobias Johannink, Shikhar Bahl, Ashvin Nair, Jianlan Luo, Avinash Kumar, Matthias Loskyll, Juan Aparicio Ojea, Eugen Solowjow, and Sergey Levine. “Residual reinforcement learning for robot control”. In: IEEE International Conference on Robotics and Automation (ICRA) . 2019

    2019

  29. [37]

    Optimal and autonomous control using reinforcement learn- ing: A survey

    Bahare Kiumarsi, Kyriakos G Vamvoudakis, Hamidreza Modares, and Frank L Lewis. “Optimal and autonomous control using reinforcement learn- ing: A survey”. In: IEEE Transactions on Neural Networks and Learning Systems 29.6 (2017)

    2017

  30. [38]

    Kernel-based approximation of the Koopman gen- erator and Schr¨ odinger operator

    Stefan Klus, Feliks N¨ uske, and Boumediene Hamzi. “Kernel-based approximation of the Koopman gen- erator and Schr¨ odinger operator”. In:Entropy 22.7 (2020). 12

    2020

  31. [39]

    Data-driven approximation of the Koop- man generator: Model reduction, system identifica- tion, and control

    Stefan Klus, Feliks N¨ uske, Sebastian Peitz, Jan- Hendrik Niemann, Cecilia Clementi, and Christof Sch¨ utte. “Data-driven approximation of the Koop- man generator: Model reduction, system identifica- tion, and control”. In: Physica D: Nonlinear Phe- nomena 406.132416 (2020)

    2020

  32. [40]

    L∞-error bounds for approximations of the Koop- man operator by kernel extended dynamic mode decomposition

    Frederik K¨ ohne, Friedrich M Philipp, Manuel Schaller, Anton Schiela, and Karl Worthmann. “L∞-error bounds for approximations of the Koop- man operator by kernel extended dynamic mode decomposition”. In: SIAM Journal on Applied Dy- namical Systems (2024). To appear (arXiv pre...

    2024 arXiv

  33. [41]

    Hamiltonian systems and transformation in Hilbert space

    Bernard O Koopman. “Hamiltonian systems and transformation in Hilbert space”. In: Proceedings of the National Academy of Sciences 17.5 (1931)

    1931

  34. [42]

    Dynamical systems of continuous spectra

    Bernard O Koopman and John von Neumann. “Dynamical systems of continuous spectra”. In: Proceedings of the National Academy of Sciences 18.3 (1932)

    1932

  35. [43]

    Linear predictors for nonlinear dynamical systems: Koopman operator meets model predictive control

    Milan Korda and Igor Mezi´ c. “Linear predictors for nonlinear dynamical systems: Koopman operator meets model predictive control”. In: Automatica 93 (2018)

    2018

  36. [44]

    On convergence of extended dynamic mode decomposition to the Koopman operator

    Milan Korda and Igor Mezi´ c. “On convergence of extended dynamic mode decomposition to the Koopman operator”. In: Journal of Nonlinear Sci- ence 28.2 (2018)

    2018

  37. [45]

    Optimal construc- tion of Koopman eigenfunctions for prediction and control

    Milan Korda and Igor Mezi´ c. “Optimal construc- tion of Koopman eigenfunctions for prediction and control”. In: IEEE Transactions on Automatic Control 65.12 (2020)

    2020

  38. [46]

    Sampled-data fun- nel control and its use for safe continual learning

    Lukas Lanza, Dario Dennst¨ adt, Karl Worthmann, Philipp Schmitz, G¨ ok¸ cen Devlet S ¸en, Stephan Trenn, and Manuel Schaller. “Sampled-data fun- nel control and its use for safe continual learning”. In: Systems & Control Letters 192.105892 (2024)

    2024

  39. [47]

    Event-based control of nonlinear systems with partial state and output feedback

    Tengfei Liu and Zhong-Ping Jiang. “Event-based control of nonlinear systems with partial state and output feedback”. In: Automatica 53 (2015)

    2015

  40. [48]

    A state-feedback approach to event-based control

    Jan Lunze and Daniel Lehmann. “A state-feedback approach to event-based control”. In: Automatica 46.1 (2010)

    2010

  41. [49]

    Data-Driven System Analysis of Nonlinear Systems Using Poly- nomial Approximation

    Tim Martin and Frank Allg¨ ower. “Data-Driven System Analysis of Nonlinear Systems Using Poly- nomial Approximation”. In: IEEE Transactions on Automatic Control 69.7 (2024)

    2024

  42. [50]

    Koopman-based lifting techniques for nonlinear systems identification

    Alexandre Mauroy and Jorge Goncalves. “Koopman-based lifting techniques for nonlinear systems identification”. In: IEEE Transactions on Automatic Control 65.6 (2019)

    2019

  43. [51]

    Linear identification of nonlinear systems: A lifting tech- nique based on the Koopman operator

    Alexandre Mauroy and Jorge Goncalves. “Linear identification of nonlinear systems: A lifting tech- nique based on the Koopman operator”. In: 55th IEEE Conference on Decision and Control (CDC). 2016

    2016

  44. [52]

    Koopman operator in systems and control

    Alexandre Mauroy, Y Susuki, and Igor Mezic. Koopman operator in systems and control . Springer, 2020

    2020

  45. [53]

    Analysis of fluid flows via spectral properties of the Koopman operator

    Igor Mezi´ c. “Analysis of fluid flows via spectral properties of the Koopman operator”. In: Annual review of fluid mechanics 45 (2013)

    2013

  46. [54]

    On numerical approximations of the Koopman operator

    Igor Mezi´ c. “On numerical approximations of the Koopman operator”. In: Mathematics 10.7 (2022)

    2022

  47. [55]

    Spectral properties of dynamical sys- tems, model reduction and decompositions

    Igor Mezi´ c. “Spectral properties of dynamical sys- tems, model reduction and decompositions”. In: Nonlinear Dynamics 41.1-3 (2005)

    2005

  48. [56]

    Data-driven feedback stabilisation of nonlinear systems: Koopman-based model pre- dictive control

    Abhinav Narasingam, Sang Hwan Son, and Joseph Sang-Il Kwon. “Data-driven feedback stabilisation of nonlinear systems: Koopman-based model pre- dictive control”. In: International Journal of Con- trol 96.3 (2023)

    2023

  49. [57]

    Finite- data error bounds for Koopman-based prediction and control

    Feliks N¨ uske, Sebastian Peitz, Friedrich Philipp, Manuel Schaller, and Karl Worthmann. “Finite- data error bounds for Koopman-based prediction and control”. In: Journal of Nonlinear Science 33.14 (2023)

    2023

  50. [58]

    A review of safe online learning for nonlinear control systems

    Matthew Osborne, Hyo-Sang Shin, and Antonios Tsourdos. “A review of safe online learning for nonlinear control systems”. In: International Con- ference on Unmanned Aircraft Systems (ICUAS) . 2021

    2021

  51. [59]

    Data-driven model predictive con- trol using interpolated Koopman generators

    Sebastian Peitz, Samuel E Otto, and Clarence W Rowley. “Data-driven model predictive con- trol using interpolated Koopman generators”. In: SIAM Journal on Applied Dynamical Systems 19.3 (2020)

    2020

  52. [60]

    Lya- punov design for safe reinforcement learning

    Theodore J Perkins and Andrew G Barto. “Lya- punov design for safe reinforcement learning”. In: Journal of Machine Learning Research 3 (2002)

    2002

  53. [61]

    Variance representations and convergence rates for data-driven approxima- tions of Koopman operators

    Friedrich M Philipp, Manuel Schaller, Septi- mus Boshoff, Sebastian Peitz, Feliks N¨ uske, and Karl Worthmann. “Variance representations and convergence rates for data-driven approxima- tions of Koopman operators”. In: ArXiv preprint arXiv:2402.02494 (2024)

    2024 arXiv

  54. [62]

    Er- ror bounds for kernel-based approximations of the Koopman operator

    Friedrich M Philipp, Manuel Schaller, Karl Worth- mann, Sebastian Peitz, and Feliks N¨ uske. “Er- ror bounds for kernel-based approximations of the Koopman operator”. In: Applied and Computa- tional Harmonic Analysis 71.101657 (2024). 13

    2024

  55. [63]

    Dynamic mode decomposition with con- trol

    Joshua L Proctor, Steven L Brunton, and J Nathan Kutz. “Dynamic mode decomposition with con- trol”. In: SIAM Journal on Applied Dynamical Sys- tems 15.1 (2016)

    2016

  56. [64]

    Stochastic stability of event- triggered anytime control

    Daniel E Quevedo, Vijay Gupta, Wann-Jiun Ma, and Serdar Y¨ uksel. “Stochastic stability of event- triggered anytime control”. In: IEEE Transactions on Automatic Control 59.12 (2014)

    2014

  57. [65]

    Data-Driven Predictive Control of Nonholo- nomic Robots Based on a Bilinear Koopman Re- alization: Data Does Not Replace Geometry

    Mario Rosenfelder, Lea Bold, Hannes Eschmann, Peter Eberhard, Karl Worthmann, and Henrik Ebel. “Data-Driven Predictive Control of Nonholo- nomic Robots Based on a Bilinear Koopman Re- alization: Data Does Not Replace Geometry”. In: arXiv preprint arXiv:2411.07192 (2024)

    2024

  58. [66]

    A two-degree-of-freedom PI controller based on events

    J S´ anchez, Antonio Visioli, and S Dormido. “A two-degree-of-freedom PI controller based on events”. In: Journal of Process Control 21.4 (2011)

    2011

  59. [67]

    A continuous-time fundamental lemma and its application in data- driven optimal control

    Philipp Schmitz, Timm Faulwasser, Paolo Rapis- arda, and Karl Worthmann. “A continuous-time fundamental lemma and its application in data- driven optimal control”. In:Systems & Control Let- ters 194.105950 (2024)

    2024

  60. [68]

    Safe data-driven reference tracking with prescribed performance

    Philipp Schmitz, Lukas Lanza, and Karl Worth- mann. “Safe data-driven reference tracking with prescribed performance”. In: 27th International Conference on System Theory, Control and Com- puting (ICSTCC) . IEEE. 2023

    2023

  61. [69]

    Multivari- able feedback control: analysis and design

    Sigurd Skogestad and Ian Postlethwaite. Multivari- able feedback control: analysis and design. John Wi- ley & Sons, 2005

    2005

  62. [70]

    A nonlinear MPC scheme for output tracking without terminal ingredients

    Raffaele Soloperto, Johannes K¨ ohler, and Frank Allg¨ ower. “A nonlinear MPC scheme for output tracking without terminal ingredients”. In: IEEE Transactions on Automatic Control 68.4 (2022)

    2022

  63. [71]

    Koopman- based feedback design with stability guarantees

    Robin Str¨ asser, Manuel Schaller, Karl Worthmann, Julian Berberich, and Frank Allg¨ ower. “Koopman- based feedback design with stability guarantees”. In: IEEE Transactions on Automatic Control (2024)

    2024

  64. [72]

    Koopman operator based observer synthesis for control-affine nonlinear systems

    Amit Surana. “Koopman operator based observer synthesis for control-affine nonlinear systems”. In: 55th IEEE Conference on Decision and Control (CDC). 2016

    2016

  65. [73]

    Fundamental lemma for data-driven analysis of linear parameter-varying systems

    Chris Verhoek, Roland T´ oth, Sofie Haesaert, and Anne Koch. “Fundamental lemma for data-driven analysis of linear parameter-varying systems”. In: 60th IEEE conference on decision and control (CDC). 2021

    2021

  66. [74]

    A pre- dictive safety filter for learning-based control of constrained nonlinear dynamical systems

    Kim P Wabersich and Melanie N Zeilinger. “A pre- dictive safety filter for learning-based control of constrained nonlinear dynamical systems”. In: Au- tomatica 129.109597 (2021)

    2021

  67. [75]

    Pre- dictive control barrier functions: Enhanced safety mechanisms for learning-based control

    Kim P Wabersich and Melanie N Zeilinger. “Pre- dictive control barrier functions: Enhanced safety mechanisms for learning-based control”. In: IEEE Transactions on Automatic Control 68.5 (2022)

    2022

  68. [76]

    Scattered data approximation

    Holger Wendland. Scattered data approximation . Vol. 17. Cambridge university press, 2004

    2004

  69. [77]

    A note on persistency of excitation

    Jan C Willems, Paolo Rapisarda, Ivan Markovsky, and Bart LM De Moor. “A note on persistency of excitation”. In: Systems & Control Letters 54.4 (2005)

    2005

  70. [78]

    A Data-Driven Approximation of the Koopman Operator: Extending Dynamic Mode Decomposition

    Matthew Williams, Ioannis Kevrekidis, and Clarence Rowley. “A Data-Driven Approximation of the Koopman Operator: Extending Dynamic Mode Decomposition”. In: Journal of Nonlinear Science 25.6 (2015)

    2015

  71. [79]

    A kernel-based method for data-driven koopman spectral analysis

    Matthew O Williams, Clarence W Rowley, and Ioannis G Kevrekidis. “A kernel-based method for data-driven koopman spectral analysis”. In: Jour- nal of Computational Dynamics 2.2 (2016)

    2016

  72. [80]

    A quan- titative analysis of Koopman operator methods for system identification and predictions

    Christophe Zhang and Enrique Zuazua. “A quan- titative analysis of Koopman operator methods for system identification and predictions”. In:Comptes Rendus. M´ ecanique351.S1 (2023). 14

    2023

Pith tools

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