Pith. sign in

REVIEW 1 major objections 3 minor 35 references

Insensitizing controls for stochastic parabolic equations with dynamic boundary conditions

T0 review · 1 major / 3 minor · reviewed 2026-08-12 · deepseek-v4-flash

Pith's one-line read The paper proves that, for a forward stochastic heat equation equipped with dynamic boundary conditions, a triple of controls can be chosen so that the energy functional measuring the solution over an observation region is insensitive to…

desk verdict The intended result is new and the surrounding framework is sound, but the observability inequality relies on a false pointwise bound at t=0, so the main theorem is unproved as written. read the letter →

arxiv 2411.18501 v1 pith:K6CNQZCI submitted 2024-11-27 math.OC math.AP

classification math.OCmath.AP MSC 93B0593C2060H15
keywords insensitizingcontrolsstochasticparabolicequationsdynamicboundaryconditionsCarlemanestimatesobservabilityinequalitynullcontrollabilityforward-backwardsystemdualitymethod
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to prove that a forward stochastic heat equation with dynamic boundary conditions admits insensitizing controls: a triple of controls $(u,v_1,v_2)$ that makes the energy functional $\Phi$, measuring the solution over an observation region, locally independent of small unknown perturbations of the initial data. If true, this means an observer's energy measurements of such a system can be protected from initial-condition uncertainty, which matters for control and inverse problems on domains where the boundary has its own dynamics. The proof reduces insensitivity to a null controllability question for a coupled forward-backward stochastic parabolic system, establishes a global Carleman estimate for that system, and then uses a duality argument to build the controls. The main theorem asserts existence together with a control-cost bound in terms of a weighted norm of the source terms, under the geometric condition that the control region $G_0$ and observation region $O$ intersect and under zero initial data.

What carries the argument

The central object is the adjoint coupled system (3.10)-(3.11), consisting of a backward stochastic parabolic equation for $(p,p_\Gamma,P,\hat P)$ and a forward one for $(q,q_\Gamma)$, linked through observation terms $\chi_O q$ and $\chi_{O_\Gamma}q_\Gamma$. The load-bearing tool is the Carleman estimate (4.5), a weighted integral inequality with weight $\theta^2\gamma^3$ (where $\theta=e^{\lambda\alpha}$, $\gamma=1/(t(T-t))$) that controls global $L^2$ norms of both components by local data on $G_0$ and by the noise variables $P,\hat P$; from it the paper derives the observability inequality (4.17) with the exponential weight $\exp(-Mt-1)$. This inequality is the bridge to duality: it makes the linear functional in Section 5 bounded on the space of adjoint solutions, which is exactly what turns into insensitizing controls via Proposition 3.1.

What would settle it

Evaluate the inequality (4.23) at a fixed $x\in G$ and let $t\to 0^+$: the left-hand factor $\exp(-Mt-1)$ approaches $e^{-1}$, while the right-hand Carleman weight $\theta^2\gamma^3 = \exp(2\lambda e^{\mu\psi(x)}-2\lambda e^{2\mu|\psi|_\infty})/(t^3(T-t)^3)$ approaches $0$ because $\psi(x)<|\psi|_\infty$; the claimed bound therefore cannot hold for any finite $C$, and the derivation of the observability inequality (4.17) fails as written.

Watch

Extended reading notes

Core claim

The paper's central claim is Theorem 1.1: whenever $G_0\cap O\neq\emptyset$ and the initial state is $(0,0)$, there are constants $M,C>0$ such that for every source pair $(\xi_1,\xi_2)$ satisfying the weighted integrability condition (1.4), one can find $(u,v_1,v_2)$ insensitizing $\Phi$ in the sense of Definition 1.1. The insensitivity conditions are the vanishing of the two partial derivatives (1.3) at $\tau_1=\tau_2=0$, and the cost estimate (1.5) bounds the controls by the same weighted source norm. The mechanism is to observe that insensitivity is equivalent to $z(0,\cdot)=0$, $z_\Gamma(0,\cdot)=0$ for the backward component of the cascade (Proposition 3.1), to prove the Carleman estimate (4.5) and then the observability inequality (4.17) for the adjoint system, and finally to apply Hahn-Banach and Riesz representation to extract the controls. The paper also records that this works for every initial data if the observation functional starts after a positive time $t_0>0$ and the sources vanish in $(0,t_0)$ (Remark 1.1).

Load-bearing premise

The load-bearing premise is that the exponential comparison in (4.23) holds on $(0,T/2)$, but near $t=0$ the left side approaches $e^{-1}$ while the Carleman weight tends to $0$, so as written the comparison fails and the observability bridge to duality is not established.

Editorial extensions

If this is right

  • For zero initial data and sources with finite weighted norm (1.4), the theorem guarantees an insensitizing control triple whose cost is bounded by that weighted norm; the paper presents this as the first such result for stochastic parabolic equations with dynamic boundary conditions.
  • Insensitivity of $\Phi$ is completely equivalent to null controllability of the cascade (3.1)-(3.2), so any future null-controllability result for that cascade automatically yields insensitizing controls.
  • If the sources vanish on an initial time interval, the condition on the initial data disappears: every initial state can be insensitized for the modified functional $\Phi_{t_0}$ (Remark 1.1).
  • The paper identifies the disjoint case $G_0\cap O=\emptyset$ as an open problem, so the theorem's geometric assumption is not known to be necessary.
  • Insensitizing controls exist in the three-control form $(u,v_1,v_2)$; the paper explicitly asks whether a single control $u$ (without the extra noise controls) can suffice, pointing to a spectral approach for that question.

Reading between the lines

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

  • Beyond the paper: if the faulty comparison (4.23) near $t=0$ can be repaired by choosing a weight that does not collapse at the initial time, the overall structure of the proof would still deliver Theorem 1.1; the flaw is in the derivation, not necessarily in the statement.
  • Beyond the paper: the same duality setup should transfer to the $\varepsilon$-insensitizing problem described in Section 6, turning the observability inequality into an approximate controllability statement and giving a quantitative version of the paper's open problem (3).
  • Beyond the paper: because the Carleman estimate controls both bulk and boundary components uniformly, the method may extend to semilinear equations with globally Lipschitz nonlinearities by a standard fixed-point iteration, matching the paper's proposed open problem (2).
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

1 major / 3 minor

Summary. This paper studies insensitizing controls for a forward linear stochastic heat equation with dynamic boundary conditions on a bounded domain. The main result, Theorem 1.1, claims that if G0∩O is nonempty and the initial data are zero, then for sources satisfying the weighted integrability condition (1.4) there exists a control triple (u,v1,v2) that insensitizes the energy functional Φ defined in (1.2). The proof follows a standard three-step architecture: Proposition 3.1 reduces insensitivity to a partial null controllability property for a cascade of forward-backward stochastic parabolic equations; Theorem 4.1 establishes a global Carleman estimate for the coupled adjoint system (3.10)-(3.11); Proposition 4.1 derives an observability inequality; and Section 5 uses duality, Hahn-Banach, and Riesz representation to construct the controls. The paper closes with open problems in Section 6.

Significance. If the main result were fully proved, this would be a natural and meaningful extension of insensitizing-control results to stochastic parabolic equations with dynamic boundary conditions, a setting for which the literature is still sparse. The reduction in Proposition 3.1 is clearly formulated, and the strategy of combining published Carleman estimates from reference [2] with a new coupling argument is appropriate. The paper is also transparent about the role of the auxiliary controls v1 and v2 and about the zero-initial-data restriction. However, as discussed below, the observability inequality in Proposition 4.1, which is the load-bearing bridge to the duality construction in Section 5, is not proved as written because the crucial pointwise weight comparison near t=0 is false.

major comments (1)
  1. [§4.2, Eq. (4.23)] The pointwise comparison asserted immediately before (4.23) is false near t=0. With α(t,x)=(e^{μψ(x)}-e^{2μ|ψ|∞})/(t(T-t)), the numerator is strictly negative and bounded away from zero, so for every x∈G one has θ²γ³ = exp(2λα)γ³ ≤ C e^{-c/t}/t³ → 0 as t↓0. On the other hand, exp(-Mt-1) tends to e^{-1}>0 as t↓0 for every fixed M. Hence the inequality exp(-Mt-1) ≤ Cθ²γ³ cannot hold on (0,T/2)×G for any finite M,C. Since (4.23) is the only step that transfers the Carleman estimate (4.5) to the unweighted p-integral over (0,T/2), inequality (4.24) is not established. The energy estimates (4.20) and (4.22) cover only t∈(T/4,T) and do not control p on (0,T/4), so the observability inequality (4.17), on which the duality argument in Section 5 rests, is not proved as written.
minor comments (3)
  1. [Theorem 4.1, display (4.5)] The powers of λ on the P and \hat P terms in (4.5) appear to be 1, whereas the proof produces λ² in those terms, as in (4.6). Since λ is fixed large this does not affect the later argument, but the displayed statement should be made consistent with the proof.
  2. [References] References [7] and [9] list the same article (Bodart, González-Burgos, and Pérez-García, C. R. Math. 335 (2002) 677–682); these entries should be merged or corrected.
  3. [Introduction, Theorem 1.1] The notation exp(Mt−1) in (1.4) and (1.5) is easy to misread; using explicit parentheses, for example exp(Mt−1), would improve clarity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is a standard black-box use of earlier Carleman estimates, and the self-citations do not define the target property into existence.

full rationale

The paper's derivation chain is not circular. Theorem 1.1 is obtained by reducing the insensitizing condition (1.3) to the null controllability condition z(0)=z_Gamma(0)=0 (Proposition 3.1), then proving an observability inequality (4.17) via a Carleman estimate (4.5), and finally applying Hahn-Banach and Riesz representation. Each step uses the preceding step as an input; no equation is defined in terms of its conclusion, and no fitted parameter is renamed as a prediction. The heavy reliance on Lemmas 4.2 and 4.3 and Theorems 2.1 and 2.2 from reference [2] is self-citation, but those cited results are prior published statements about well-posedness and Carleman estimates for single forward/backward parabolic equations; their assumptions do not include the insensitizing property, so they are not equivalent to the target result by construction. The apparent failure of the comparison (4.23) near t=0 is a serious correctness gap in the proof of Proposition 4.1, but it is a mathematical error in the argument, not a circular reduction: the claimed inequality is false, not merely a restatement of an input. Consequently, under the rule that self-citation becomes circularity only when the load-bearing argument reduces to the self-citation itself, this paper receives a circularity score of 0.

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

This is a purely mathematical existence proof with no empirically fitted parameters or new postulated physical entities. The constants and large parameters in the Carleman argument are technical proof devices, not free parameters of the model. The main external inputs are well-posedness and Carleman estimates from the authors' own prior work [2].

assumptions (5)
  • domain assumption Well-posedness of forward and backward stochastic parabolic equations with dynamic boundary conditions (Theorems 2.1 and 2.2).
    Invoked in Section 2 and throughout to define solutions of systems (3.1)-(3.2) and (3.10)-(3.11). The results are cited from reference [2], a paper with overlapping authorship.
  • domain assumption Carleman estimates for a single forward stochastic parabolic equation and a single backward stochastic parabolic equation with dynamic boundary conditions (Lemmas 4.2 and 4.3).
    These are the main technical inputs used to prove the new coupled Carleman estimate (4.5) in Theorem 4.1. They are cited from [2], not proved in this paper.
  • standard math Existence of a weight function psi with psi=0 on the boundary and |grad psi|>0 outside a small open set (Lemma 4.1).
    Standard weight construction from Fursikov and Imanuvilov [15], used to build the Carleman weights theta and gamma.
  • standard math Ito formula for products of solutions of forward and backward stochastic parabolic equations.
    Used in Proposition 3.1, Theorem 4.1, and the proof of Theorem 1.1 to transfer integrals between coupled systems.
  • standard math Hahn-Banach and Riesz representation theorems in adapted L2 spaces.
    Used in Section 5 to extend the functional L from the observability subspace H to the full control space and to represent it as controls (u,v1,v2).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Insensitizing controls for stochastic parabolic equations with dynamic boundary conditions." pith.science (2026). https://pith.science/paper/K6CNQZCI

@misc{pith2026241118501,
  author       = {Pith},
  title        = {Pith review of: Insensitizing controls for stochastic parabolic equations with dynamic boundary conditions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/K6CNQZCI}},
  note         = {Machine review of arXiv:2411.18501}
}
read the original abstract

In this paper, we continue the study of some controllability issues for the forward stochastic heat equation with dynamic boundary conditions. The main novelty in the present paper consists of considering only one control without extra forces in the noise parts. Under a strong measurability condition, and using a spectral inequality, we first establish an appropriate observability inequality for the corresponding adjoint system. Then, by the classical duality approach, the null and approximate controllability results are established.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 34 canonical work pages

  1. [2]

    Baroun, S

    M. Baroun, S. Boulite, A. Elgrou, and L. Maniar, Null cont rollability for stochastic parabolic equations with dynamic boundary conditions, Journal of Dynamical and Control Systems , 29(2023), 1727–1756. 1, 1.3, 2, 2, 4.1, 4.1

  2. [1]

    E. M. Ait Ben Hassi, S. E. Chorfi, L. Maniar, and O. Oukdach, Lipschitz stability for an inverse source problem in anisotropic parabolic equations with dynamic boundary c onditions, Evolution Equations and Control Theory , 10(2021), 837–859. 1

  3. [3]

    Baroun, S

    M. Baroun, S. Boulite, A. Elgrou, and L. Maniar, Null cont rollability for backward stochastic parabolic convection-diffusion equations with dynamic boundary cond itions, Mathematical Control and Related Fields , (2024). 1

  4. [4]

    Baroun, S

    M. Baroun, S. Boulite, A. Elgrou, and L. Maniar, Null cont rollability for one-dimensional stochastic heat equations with mixed Dirichlet-dynamic boundary conditio ns. To appear. 1

  5. [5]

    Bodart and C

    O. Bodart and C. Fabre, Controls insensitizing the norm o f the solution of a semilinear heat-equation, Journal of Mathematical Analysis and Applications , 195(1995), 658–683. 1, 3

  6. [7]

    Bodart, M

    O. Bodart, M. Gonz´ alez-Burgos, and R. P´ erez-Garc ´ ıa,Existence of insensitizing controls for a semilinear heat equation with a superlinear nonlinearity, Communications in Partial Differential Equations , 29(2004), 1017–

  7. [8]

    Bodart, M

    O. Bodart, M. Gonz´ alez-Burgos, and R. P´ erez-Garc ´ ıa,Insensitizing controls for a heat equation with a nonlinear term involving the state and the gradient, Nonlinear Analysis: Theory, Methods & Applications , 57(2004), 687–

  8. [9]

    Bodart, M

    O. Bodart, M. Gonz´ alez-Burgos, and R. P´ erez-Garc ´ ıa,Insensitizing controls for a semilinear heat equation with a superlinear nonlinearity, Comptes Rendus Mathematique , 335(2002), 677–682. 1

Show all 35 references
  1. [10]

    Boulite, A

    S. Boulite, A. Elgrou, L. Maniar, and O. Oukdach, Contro llability for forward stochastic parabolic equations with dynamic boundary conditions without extra forces, arXiv preprint arXiv:2312.15537 , (2023). 1.3

  2. [11]

    Boutaayamou, S

    I. Boutaayamou, S. E. Chorfi, L. Maniar, and O. Oukdach, T he cost of approximate controllability of heat equation with general dynamical boundary conditions, Portugaliae Mathematica , 78(2021), 65–99. 1

  3. [12]

    Boutaayamou, L

    I. Boutaayamou, L. Maniar, and O. Oukdach, Stackelberg -Nash null controllability of heat equation with general dynamic boundary conditions, Evolution Equations and Control Theory , 11(2022), 1285–1307. 1

  4. [13]

    Elgrou, L

    A. Elgrou, L. Maniar, and O. Oukdach, Inverse initial pr oblem under Nash strategy for stochastic reaction- diffusion equations with dynamic boundary conditions, arXiv preprint arXiv:2410.10007, (2024). 1

  5. [14]

    Ervedoza, P

    S. Ervedoza, P. Lissy, and Y. Privat, Desensitizing con trol for the heat equation with respect to domain variations , Journal de l’ ´Ecole polytechnique—Math´ ematiques, 9(2022), 1397–1429. 1.2

  6. [15]

    A. V. Fursikov and O. Yu. Imanuvilov, Controllability o f evolution equations, Lecture Note Series 34, Research Institute of Mathematics, Seoul National University, (1996). 4.1

  7. [16]

    Guerrero, Null controllability of some systems of tw o parabolic equations with one control force, SIAM journal on control and optimization , 46(2007), 379–394

    S. Guerrero, Null controllability of some systems of tw o parabolic equations with one control force, SIAM journal on control and optimization , 46(2007), 379–394. 1, 4

  8. [17]

    Khoutaibi, L

    A. Khoutaibi, L. Maniar, and O. Oukdach, Null controlla bility for semilinear heat equation with dynamic boundary conditions, Discrete Continuous Dynamical Systems-Series S , 15(2022). 1

  9. [18]

    Kirane, A

    M. Kirane, A. Lopushansky, and H. Lopushanska, Determi nation of two unknown functions of different variables in a time-fractional differential equation. Mathematical M ethods in the Applied Sciences. 1

  10. [19]

    J.-L. Lions, Remarques pr´ eliminaires sur le contrˆ ole des systemesa donn´ ees incompletes, In Actas del Congreso de Ecuaciones Diferenciales y Aplicaciones (CEDYA), Unive rsidad de Malaga , 43–54. 1 18 M. BAROUN, S. BOULITE, A. ELGROU, AND O. OUKDACH

  11. [20]

    Liu, Global Carleman estimate for stochastic parabo lic equations, and its application, ESAIM: Control, Optimisation and Calculus of Variations , 20(2014), 823–839

    X. Liu, Global Carleman estimate for stochastic parabo lic equations, and its application, ESAIM: Control, Optimisation and Calculus of Variations , 20(2014), 823–839. 1

  12. [21]

    Liu and Y

    X. Liu and Y. Yu, Carleman estimates of some stochastic d egenerate parabolic equations and application, SIAM Journal on Control and Optimization , 57(2019), 3527–3552. 1

  13. [22]

    L¨ u, Some results on the controllability of forward s tochastic heat equations with control on the drift, Journal of Functional Analysis , 260(2011), 832–851

    Q. L¨ u, Some results on the controllability of forward s tochastic heat equations with control on the drift, Journal of Functional Analysis , 260(2011), 832–851. 1.3

  14. [23]

    L¨ u and X

    Q. L¨ u and X. Zhang, Mathematical control theory for sto chastic partial differential equations, Springer, (2021). 1.3

  15. [24]

    Maniar, M

    L. Maniar, M. Meyries, and R. Schnaubelt, Null controll ability for parabolic equations with dynamic boundary conditions, Evolution Equations and Control Theory , 6(2017), 381–407. 1

  16. [25]

    Maniar, O

    L. Maniar, O. Oukdach, and W. Zouhair, Lebeau–Robbiano inequality for heat equation with dynamic boundary conditions and optimal null controllability, Differential Equations and Dynamical Systems , (2023), 1–17. 1

  17. [26]

    Oukdach, I

    O. Oukdach, I. Boutaayamou, and L. Maniar, Hierarchica l control problem for the heat equation with dynamic boundary conditions, IMA Journal of Mathematical Control and Information , (2024). 1

  18. [27]

    Oukdach, S

    O. Oukdach, S. Boulite, A. Elgrou, and L. Maniar, Multi- objective control for stochastic parabolic equations with dynamic boundary conditions, arXiv preprint arXiv:2405.15730 , (2024). 1

  19. [28]

    M. C. Santos, N. Carre˜ no, and R. Morales, An Insensitiz ing control problem involving tangential gradient terms for a reaction-diffusion equation with dynamic boundary con ditions, arXiv preprint arXiv:2407.09882 , (2024). 1, 4

  20. [29]

    Tang and X

    S. Tang and X. Zhang, Null controllability for forward a nd backward stochastic parabolic equations, SIAM Journal on Control and Optimization , 48(2009), 2191–2216. 1.3

  21. [30]

    L. D. Teresa, Insensitizing controls for a semilinear h eat equation: semilinear heat equation, Communications in Partial Differential Equations , 25(2000), 39–72. 1, 1.1, 2

  22. [31]

    L. D. Teresa, Controls insensitizing the norm of the sol ution of a semilinear heat equation in unbounded domains, ESAIM: Control, Optimisation and Calculus of Variations , 2(1997), 125–149. 1

  23. [32]

    L. D. Teresa and E. Zuazua, Identification of the class of initial data for the insensitizing control of the heat equation, Communication on pure and applied analysis , 8(2009). 1

  24. [33]

    Yan and F

    Y. Yan and F. Sun, Insensitizing controls for a forward s tochastic heat equation. Journal of mathematical analysis and applications , 384(2011), 138–150. 1

  25. [34]

    Zhang, J

    M. Zhang, J. Yin, and H. Gao, Insensitizing controls for the parabolic equations with dynamic boundary conditions. Journal of Mathematical Analysis and Applications , 475(2019), 861–873. 1, 2

  26. [35]

    Yamamoto, Carleman estimates for parabolic equatio ns and applications, Inverse problems , 25(2009), 123013

    M. Yamamoto, Carleman estimates for parabolic equatio ns and applications, Inverse problems , 25(2009), 123013. 1

  27. [36]

    Yang and J

    D. Yang and J. Zhong, Observability inequality of backw ard stochastic heat equations for measurable sets and its applications, SIAM Journal on Control and Optimization, 54(2016), 1157–1175. 1.3

Pith tools

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