{"id":"f7337cb0-fd54-43a2-8468-39631b649fa0","arxiv_id":"2506.18634","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Cubic boundary feedback exponentially stabilizes 1-D quasilinear parabolic PDEs with finite-time blow-up and gives explicit estimates of the region of attraction, well-posedness, and positivity.","lead":"This paper designs cubic boundary feedback laws that prevent finite-time blow-up in a broad class of 1-D quasilinear parabolic equations, and it estimates how large the basin of stable initial states is. The result matters because it is among the first boundary-control guarantees for equations with state-dependent diffusion, convection, and destabilizing reaction.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Condition (14) is algebraically unsatisfiable for quadratic-only convection, so Theorem 1 excludes the Burgers-like case u^2 u_x advertised in the abstract.","rationale":"The reader's named weakest assumption is the one-dimensional Agmon inequality and the consequent absence of a higher-dimensional analogue. That is a real limitation, but it is explicitly acknowledged in the conclusion and does not contradict any claim made for the one-dimensional domain. The more load-bearing defect is internal to the one-dimensional theorem: hypothesis (14) cannot be met when quadratic convection is the only higher-order convective term, so Theorem 1 is vacuous for a subclass the abstract and introduction explicitly advertise. The reader's rationale does mention this defect and the inverse-formula typo, but their formal weakest-assumption field selects the dimensional limitation. My stress test therefore agrees partially: the final CONDITIONAL verdict is appropriate, but the condition (14) inconsistency is the concern that most directly threatens the central claim as stated. The inverse formula (17) is a genuine typo, but it does not bear on exponential stability or well-posedness, so it strengthens the need for revision without changing the verdict. A concrete analytic substitution settles the (14) issue immediately, and the proposed repaired-estimate check would determine whether the central theorem can be extended to quadratic convection or must be restricted.","tokens_in":19118,"tokens_out":8586,"duration_ms":87138,"concrete_test":"Set n = 2, γ_2 ≠ 0, γ_i = 0 for i ≥ 3, p = 2, and ε constant. Substitute into (14): the two sides are equal, |γ_2|/4, for every M, ω > 0, so no admissible ω exists. Then attempt the Lyapunov proof with Γ ≥ 0, using the exact boundary identity u(1)ε(u(1))v_1(u(1)) + (γ_2/4)u(1)^4 = (γ_2/4)(1 - ε(u(1))/ε)u(1)^4 - ε(u(1))λ_1 u(1)^2, to see whether a repaired statement can cover quadratic convection. If the repaired estimate closes, the fix is a condition change; if not, Theorem 1 must explicitly exclude u^2 u_x.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The stability claim of Theorem 1 rests on the hypothesis (14): whenever some γ_i with i ≥ 2 is nonzero, one must have Σ_{i=2}^n |γ_i|/(i+2) M^{i-2} > Σ_{i=2}^n |γ_i|/(i+2)(6ω)^{(i-2)/2}. If the only higher-order convection is quadratic, i.e., n = 2 and γ_2 ≠ 0 (with γ_1 arbitrary), then the i = 2 terms on both sides are identical: |γ_2|/4 · M^0 = |γ_2|/4 · (6ω)^0 = |γ_2|/4. No choice of M, ω, or the free gains can make the strict inequality true. Consequently, Theorem 1 gives no stabilization guarantee for the most natural quasilinear convective term u^2 u_x, the case motivated in the Introduction through Burgers' equation. The proof requires Γ(E) > 0 in (22)-(26) to ensure the boundary quartic terms are dissipative; for i = 2, Γ(E) = |γ_2|/4(1 - 1) = 0 identically, and the cubic boundary term cancels in (18), so the advertised region-of-attraction estimate does not close for this subclass. Simply replacing '>' by '≥' does not immediately repair the argument, because the proof then loses the positive boundary dissipation needed in (24)-(27); a separate estimate would be required. A second, less central defect is the inverse formula (17), where y/(2λ_l) should read y/(2μ), but that affects only the implementability of the Dirichlet form, not the stability conclusion.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a one-dimensional quasilinear parabolic equation (1) with state-dependent diffusion, polynomial convection, and a superlinear reaction term, and proposes cubic Neumann feedback laws (10)-(11). Theorem 1 claims exponential L2 and H1 stability of the origin with an explicit estimate of the region of attraction, exponential H2 and C1 convergence to zero, existence and uniqueness of complete classical solutions, positivity for nonnegative initial data, and invertibility of the boundary maps. The proof combines two Lyapunov functionals V and H, the one-dimensional Agmon inequality, a cutoff argument for well-posedness, and a parabolic comparison principle. The central Lyapunov chain in Lemmas 1-2 and the propagation argument for E(t)≤E(0) are coherent, and the well-posedness route via a truncated system is standard. However, a key hypothesis, condition (14), is unsatisfiable for the quadratic-convection case n=2, γ2≠0, and the explicit inverse formula (17) in Property 4 appears algebraically incorrect.","tokens_in":19422,"tokens_out":6210,"duration_ms":66707,"significance":"If the advertised results were fully valid, this would be a valuable contribution to boundary stabilization of quasilinear parabolic equations with blow-up, since the existing literature is largely limited to semilinear or special Burgers-type problems. The paper's strengths are its explicit and self-contained estimates, the clean two-functional Lyapunov construction, the standard cutoff/regularization route to well-posedness, and the use of the comparison principle for positivity. The paper does not fit data or rely on empirical predictions; all constants are explicit design parameters, which is a positive feature. The authors correctly acknowledge that the one-dimensional Agmon inequality is essential and that higher-dimensional extensions are open. Nevertheless, the advertised scope is currently too broad: condition (14) excludes the natural quadratic-convection subclass, and the explicit Cardano inversion formula contains a clear algebraic error. These issues are fixable but are load-bearing for the theorem as stated.","major_comments":[{"comment":"When n=2 and γ2≠0, which is exactly the quadratic-convection term u^2 u_x motivated in the Introduction through Burgers' equation, condition (14) reduces to |γ2|/4 > |γ2|/4, which is impossible for any choice of M, ω, or the free gains. Consequently, Theorem 1 gives no stabilization guarantee for this advertised subclass. In the proof, the corresponding term Γ(E) in (22) is identically zero for i=2, so the quartic boundary dissipation that the argument relies on is absent. The theorem should either be restricted explicitly to cases where at least one γ_i with i≥3 is nonzero, or a separate case covering i=2 must be proved, e.g., by observing that Γ(E)=0 still leaves a non-positive right-hand side in (22) as long as α(E) and Λ(E) are positive. As written, the mismatch between the advertised class and the hypotheses is a load-bearing gap.","section":"§3, Theorem 1, condition (14)"},{"comment":"The stated Cardano inversion formula does not solve the cubic equation (42). Setting y=u_x(l) and z=u(l), equation (42) is μ z^3 + λ_l z + (-1)^{l+1} y = 0, so the standard solution contains y/(2μ), not y/(2λ_l), and the sign must be adjusted according to l. Thus the formula in (17) is algebraically incorrect and does not establish Property 4 as stated. This is a local error, but it affects the claimed implementation of the controllers in Dirichlet or mixed form; the formula should be corrected and re-verified.","section":"§4.4, Property 4 and Eq. (17)"}],"minor_comments":[{"comment":"The sentence 'u≡0 is the unique complete complete solution to Σ̄' contains a duplicated word; please correct it.","section":"§4.2, after Eq. (39)"},{"comment":"The quantity ζ is used in the displayed formula for ζ̄ before it is defined later in the proof; please define ζ before Eq. (16) or restructure the notation so that the expression is self-contained.","section":"§3, Eq. (16)"},{"comment":"The conclusion correctly states that the approach does not extend to higher dimensions because Agmon's inequality is one-dimensional. It would be useful to place this limitation earlier, e.g., in the introduction or in the statement of Theorem 1, so that readers do not over-interpret the advertised scope.","section":"§5, Conclusion"},{"comment":"The manuscript contains several typographical spacing issues, such as 'L2 andH 1' and 'H ö lder', and author affiliation 'Departement' should be 'Département'. These should be corrected in a final pass.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The main stability proof is largely coherent, and the paper is potentially a good fit for Automatica. The two major issues are concrete and fixable: condition (14) must be repaired or qualified to cover quadratic convection, and Eq. (17) must be corrected. I would also encourage the authors to double-check all algebraic identities in the proof, since the presence of an incorrect explicit formula in a theorem statement suggests a need for a careful verification pass before resubmission."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something new: it gives a Lyapunov-based boundary feedback framework for 1-D quasilinear parabolic equations with state-dependent diffusion, convection, and superlinear reaction, where the open loop blows up. The L2/H1 coupled analysis, the cutoff/regularization route to well-posedness, and the positivity argument are all coherent and self-contained. No data are fitted; the gains are design parameters. For a broad class of convection coefficients, the stability theorem is plausible and the contribution is a genuine advance over prior semilinear or Burgers-specific results.\n\nThe soft spots are real, though not all equal in weight. The serious one is condition (14). When the only higher-order convection is quadratic (n=2, gamma_2 nonzero), the inequality becomes |gamma_2|/4 > |gamma_2|/4, which is impossible. The proof needs Gamma(E) > 0 to get boundary dissipation; with i=2 only, Gamma(E) is identically zero and the advertised region-of-attraction estimate does not close. So Theorem 1 as stated excludes u^2 u_x, a natural case the introduction motivates. A fix by replacing '>' with '>=' does not repair the proof because the boundary quartic term vanishes. This needs to be re-stated or handled separately.\n\nThe second defect is in the inverse formula (17). The cubic equation (42) should have ux(l)/mu, not ux(l)/lambda_l, and the Cardano terms in (17) need to follow suit. This is an implementation bug in Property 4, not a stability issue.\n\nThe third point is the acknowledged 1-D limitation: Agmon's inequality is load-bearing and has no higher-dimensional analogue. The authors admit this, so it is a limitation rather than an oversight.\n\nThe stress-test notes check out on reading. The paper deserves serious peer review, but not in this form. The hypotheses must be corrected to cover or explicitly exclude the quadratic-only convection case, and the inverse formula fixed. The core Lyapunov idea is sound, and with those repairs the result would be an important contribution to the PDE control literature.","headline":"A genuinely new Lyapunov framework for 1-D quasilinear blow-up PDEs with cubic boundary feedback, but Theorem 1 has a real hypothesis gap for u^2 u_x and the inverse formula is wrong.","tokens_in":20050,"tokens_out":4426,"would_cite":true,"duration_ms":42635,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35K55","35B44","93D15","93C20"],"pacs":[],"model":"deepseek-v4-flash","headline":"Cubic boundary feedback can exponentially stabilize quasilinear parabolic PDEs that blow up in open loop, with a computable region of attraction, full classical well-posedness, and positivity preservation.","keywords":["quasilinear parabolic PDEs","finite-time blow-up","boundary stabilization","cubic feedback","region of attraction","Lyapunov methods","well-posedness","positivity preservation"],"falsifier":"A concrete check: in the constant-diffusion case $\\varepsilon(u)\\equiv\\varepsilon$, $\\gamma_i=0$, $p=2$, choose $u_0$ with $\\kappa_0(u_0)\\le\\sqrt{2\\omega}$ and gains satisfying (12)-(14), and simulate the closed loop; if the sup norm ever exceeds $\\sqrt{6\\omega}$, or if $|u|_{L^2}$ does not lie below $\\|u_0\\|_{L^2}e^{-\\sigma t}$ with $\\sigma$ from (15), then the region-of-attraction estimate or the exponential decay claim fails.","tokens_in":18896,"feed_emoji":"🔥","tokens_out":10387,"duration_ms":93974,"temperature":0.7,"pith_summary":"Quasilinear parabolic equations of the form $u_t=\\varepsilon(u)u_{xx}+\\sum_{i=1}^n \\gamma_i u^i u_x+u^p$ on the unit interval, with zero boundary inputs, can blow up in finite time even for arbitrarily small initial data. This paper constructs boundary feedback laws that are cubic polynomials of the boundary values, $u_x(0)=\\lambda_0 u(0)+\\mu u(0)^3$ and $u_x(1)=-\\lambda_1 u(1)-\\mu u(1)^3$, and proves that, under three explicit algebraic inequalities on the gains and system parameters, the zero state is exponentially stable in $L^2$ and $H^1$ with a region of attraction estimated by $\\kappa_0(u_0)\\le\\sqrt{2\\omega}$. The same theorem establishes existence and uniqueness of complete classical solutions, convergence to zero in $H^2$ and $C^1$, and nonnegativity of solutions starting from nonnegative data. Because the controllers use only boundary measurements, the design covers state-dependent diffusion, convection, and destabilizing superlinear reaction, a regime where existing backstepping boundary controls could not be applied. This is the natural form of the result for this class, since global stabilization is known to be impossible in general.","feed_headline":"Cubic boundary feedback halts finite-time blow-up in quasilinear PDEs","feed_subtitle":"It gives explicit regions of attraction, well-posedness, and positivity, with basins that grow as diffusion increases.","key_machinery":"The carrying object is the Lyapunov functional $E=V+H$, where $V=\\frac12|u|_{L^2}^2$ and $H=\\frac12|u_x|_{L^2}^2+\\frac{\\lambda_1}{2}u(1)^2+\\frac{\\mu}{4}u(1)^4+\\frac{\\lambda_0}{2}u(0)^2+\\frac{\\mu}{4}u(0)^4$. Along solutions, the paper derives $\\dot E\\le -\\alpha(E)V-\\Lambda(E)|u_x|_{L^2}^2-\\varepsilon(u(1))\\Gamma(E)u(1)^4-\\varepsilon(u(0))\\Gamma(E)u(0)^4$, with the coefficient functions $\\alpha,\\Lambda,\\Gamma$ made positive by conditions (12)-(14). Agmon's inequality $|u|_\\infty^2\\le |u|_{L^2}^2+2|u|_{L^2}|u_x|_{L^2}$ converts the sup norm into a function of $E$, turning the reaction- and convection-dependent coefficients into negative dissipation terms. A cutoff version of the system is used to access a classical quasilinear well-posedness theorem, and a comparison principle gives positivity.","core_discovery":"Theorem 1 asserts that, for system (1) under Assumption 1, the cubic Neumann boundary laws $u_x(0)=\\lambda_0u(0)+\\mu u(0)^3$ and $u_x(1)=-\\lambda_1u(1)-\\mu u(1)^3$ make zero exponentially stable in $L^2$ and $H^1$ with region of attraction $\\kappa_0(u_0)\\le\\sqrt{2\\omega}$, provided $\\omega$ satisfies the explicit inequalities (12)-(14) and the gains $\\lambda_0,\\lambda_1,\\mu$ are selected as in (10)-(11). Every complete classical solution then has its $H^2$ and $C^1$ norms converge to zero, compatibility conditions give existence and uniqueness of a complete classical solution, and nonnegative initial data give nonnegative solutions. The theorem covers the quasilinear structure itself: the diffusion coefficient may depend on the state, the convection may be polynomial, and the reaction $u^p$ with $p>1$ is the destabilizing term responsible for open-loop blow-up. In many cases the guaranteed region of attraction and the decay rate grow unboundedly as the diffusion increases, so large diffusion acts as a resource for the boundary controller.","pith_inferences":["A testable extension the paper leaves implicit: in the constant-diffusion, purely reactive case ($\\varepsilon$ constant, $\\gamma_i=0$, $p=2$), the guaranteed basin radius should scale with $\\varepsilon$; numerical continuation on the one-dimensional equation could map the true basin and compare its asymptotics with the estimate.","Because Agmon's inequality is the only 1-D input, a multi-dimensional version would have to replace pointwise sup-norm control with boundary-trace estimates; the paper's conclusion flags exactly this obstruction, so the natural next conjecture is that the basin estimates survive with domain-dependent constants in two or three dimensions.","The controller uses only instantaneous boundary values, so it is compatible with observer-based or output-feedback architectures; the paper does not discuss estimation, but the boundary-measurement structure is presumably sufficient."],"forward_implications":["For the quasilinear class (1), the cubic boundary laws (10)-(11) certify a region of attraction $\\kappa_0(u_0)\\le\\sqrt{2\\omega}$ whenever the algebraic checks (12)-(14) hold, so designers can verify stabilizability without solving the PDE.","Inside this region, $L^2$ and $H^1$ norms decay exponentially and $H^2$ and $C^1$ norms converge to zero, so the controller delivers both transient stabilization and regularity recovery.","The same boundary laws admit Neumann, Dirichlet, and mixed-type implementations via the explicit inverse (17), making the design adaptable to actuation constraints in thermal or fluid systems.","Under a parity condition on the convection coefficients, one-sided boundary feedback alone (only at $x=0$ or only at $x=1$) achieves the same three guarantees, reducing actuator requirements.","Nonnegative initial data produce nonnegative solutions, which is essential in concentration or temperature models where negative states are unphysical."],"supporting_citations":[{"why":"Supplies Agmon's inequality used to bound the sup norm by the energy $E$, and the cubic boundary-feedback design that the paper adapts to quasilinear equations with destabilizing reaction.","marker":"[15]"},{"why":"Provides the classical existence/uniqueness theorem for quasilinear parabolic initial-boundary value problems, applied to a cutoff version of the closed-loop system to prove well-posedness.","marker":"[18]"},{"why":"Supplies the parabolic comparison principle used to prove that nonnegative initial data yield nonnegative solutions.","marker":"[25]"},{"why":"Documents finite-time blow-up for arbitrarily small initial conditions under null boundary inputs, the phenomenon the feedback is designed to prevent.","marker":"[26]"},{"why":"Shows that global stabilization is impossible for simple heat equations, justifying the estimate-of-region-of-attraction formulation of the result.","marker":"[8]"},{"why":"The prior backstepping boundary-feedback framework for semilinear Volterra-type parabolic PDEs that the paper identifies as unable to handle state-dependent diffusion, convection, or positivity.","marker":"[30,31]"}],"fun_headline_variants":["Cubic boundary feedback tames quasilinear blow-up, basin grows with diffusion","Quasilinear PDE blow-up halted by cubic boundary laws and explicit basin","Boundary cubic control gives exponential stability and positivity for quasilinear PDEs","State-dependent diffusion? Cubic boundary control still wins with attraction region"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"All of the stability estimates pass through the one-dimensional Agmon inequality $|u|_\\infty^2\\le |u|_{L^2}^2+2|u|_{L^2}|u_x|_{L^2}$; this bound has no analogue in higher-dimensional domains, so the Lyapunov argument closes only on $[0,1]$ and the advertised results do not extend to $d>1$ without a new idea.","fun_headline_variants_meta":{"raw":{"variants":["Cubic boundary feedback tames quasilinear blow-up, basin grows with diffusion","Quasilinear PDE blow-up halted by cubic boundary laws and explicit basin","Boundary cubic control gives exponential stability and positivity for quasilinear PDEs","State-dependent diffusion? Cubic boundary control still wins with attraction region"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000221,"raw_usage":{"total_tokens":1473,"prompt_tokens":990,"completion_tokens":483,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":606,"completion_tokens_details":{"reasoning_tokens":405}},"tokens_in":606,"tokens_out":483,"duration_ms":5109,"temperature":1.0,"reasoning_tokens":405,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T18:48:18.279029+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete check: in the constant-diffusion case $\\varepsilon(u)\\equiv\\varepsilon$, $\\gamma_i=0$, $p=2$, choose $u_0$ with $\\kappa_0(u_0)\\le\\sqrt{2\\omega}$ and gains satisfying (12)-(14), and simulate the closed loop; if the sup norm ever exceeds $\\sqrt{6\\omega}$, or if $|u|_{L^2}$ does not lie below $\\|u_0\\|_{L^2}e^{-\\sigma t}$ with $\\sigma$ from (15), then the region-of-attraction estimate or the exponential decay claim fails.","supporting_citations":[{"cited_title":"Krstic, On global stabilization of Burgers’ equation by boundary control, Systems & Control Letters, 37(3), 123-141, 1999","cited_arxiv_id":null,"evidence_quote":"Supplies Agmon's inequality used to bound the sup norm by the energy $E$, and the cubic boundary-feedback design that the paper adapts to quasilinear equations with destabilizing reaction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the classical existence/uniqueness theorem for quasilinear parabolic initial-boundary value problems, applied to a cutoff version of the closed-loop system to prove well-posedness."},{"cited_title":"Pao, Positive solutions of a nonlinear boundary-value problem of parabolic type, J","cited_arxiv_id":null,"evidence_quote":"Supplies the parabolic comparison principle used to prove that nonnegative initial data yield nonnegative solutions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents finite-time blow-up for arbitrarily small initial conditions under null boundary inputs, the phenomenon the feedback is designed to prevent."},{"cited_title":"Fern´ andez-Cara, and E","cited_arxiv_id":null,"evidence_quote":"Shows that global stabilization is impossible for simple heat equations, justifying the estimate-of-region-of-attraction formulation of the result."}],"review_version":2}