Quantitative Soft-to-Hard Terminal Constraint Convergence for the Heat Equation
Pith reviewed 2026-05-15 02:38 UTC · model grok-4.3
The pith
Penalized formulations of the heat-equation control problem converge to the exact hard terminal constraint at explicit rates O(alpha to the power minus theta).
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
We prove that the minimizers of the penalized optimal control problem for the heat equation converge to the solution of the hard-constrained problem as the penalty parameter alpha tends to infinity, with explicit rates O(alpha to the power minus theta) that become O(1/alpha) under stronger modal summability assumptions on the terminal mismatch; the underlying mechanism is illustrated by a finite-dimensional projection prototype.
What carries the argument
The quadratic penalty term (scaled by 1/alpha) added to the cost functional on the terminal-state deviation, whose minimizers converge to the hard-constrained optimum through the projection structure shown in the finite-dimensional prototype.
If this is right
- The penalized problems furnish computable approximations whose error is bounded a priori by the derived rates.
- The sharp O(1/alpha) rate is available precisely when the terminal mismatch decays sufficiently fast in the eigenfunction basis.
- The finite-dimensional prototype isolates the projection mechanism responsible for the convergence.
- The companion numerical study confirms that the predicted rates are observed in practice.
Where Pith is reading between the lines
- The same penalty analysis may extend directly to other linear parabolic equations with similar modal expansions.
- The dependence of the rate on modal summability suggests that smoother terminal targets produce faster convergence.
- These explicit rates can be used to select penalty parameters that balance computational cost against approximation accuracy in applications.
Load-bearing premise
The terminal mismatch must satisfy modal summability conditions whose strength determines the exact convergence rate.
What would settle it
A numerical computation of the penalized-control error for successively larger alpha that fails to decay at the rate O(alpha to the power minus theta) predicted by the modal summability of the chosen terminal mismatch.
read the original abstract
We study an optimal control problem for the heat equation with a prescribed terminal state. To circumvent the difficulty of enforcing a hard terminal constraint, we analyze a penalized formulation and prove that the corresponding optimal controls and terminal states converge to the exact constrained solution as the penalty parameter \(\alpha \to \infty\). We establish explicit quantitative convergence estimates of order \(O(\alpha^{-\theta})\), including the sharp \(O(1/\alpha)\) rate under stronger modal summability assumptions on the terminal mismatch. A finite-dimensional prototype is used to illustrate the underlying projection structure, while numerical illustrations are reported in a companion study.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies an optimal control problem for the heat equation subject to a prescribed terminal state. It replaces the hard terminal constraint with a quadratic penalty term scaled by α and proves that the optimal controls and terminal states of the penalized problem converge to those of the exact constrained problem as α → ∞. Explicit quantitative rates are derived: a general rate O(α^{-θ}) for some θ > 0, together with the sharp rate O(1/α) under additional modal summability assumptions on the terminal mismatch. A finite-dimensional prototype illustrates the underlying projection mechanism, while the main analysis exploits the spectral decomposition of the heat semigroup.
Significance. If the stated convergence rates hold, the manuscript supplies the first explicit quantitative error estimates for soft-to-hard terminal constraint approximation in an infinite-dimensional parabolic control setting. The distinction between the generic rate and the sharp rate under verifiable summability conditions is useful for both theoretical analysis and the design of numerical penalty methods. The spectral approach and the finite-dimensional illustration are cleanly executed and directly support the central claims.
major comments (2)
- [§4.2, Theorem 4.3] §4.2, Theorem 4.3: the proof of the general O(α^{-θ}) rate invokes a specific decay estimate on the modal coefficients of the terminal mismatch; the precise value of θ is not stated explicitly in the theorem statement, making it difficult to verify the claimed order without re-deriving the constant from the preceding lemmas.
- [§5, Assumption 5.1] §5, Assumption 5.1: the stronger modal summability condition required for the sharp O(1/α) rate is formulated in terms of an ℓ¹-type series on the Fourier coefficients; it would strengthen the result to include a brief remark on how this condition can be checked for typical target states (e.g., smooth or analytic data).
minor comments (2)
- [§3] The finite-dimensional prototype in §3 is helpful but its notation (projection onto the range of the control operator) is introduced without an explicit reference back to the infinite-dimensional setting; a short sentence linking the two would improve readability.
- [Eq. (2.7)] Equation (2.7) defines the penalized cost functional; the dependence of the optimal control u_α on α is not indicated in the notation, which occasionally leads to ambiguous statements in the convergence proofs.
Simulated Author's Rebuttal
We thank the referee for the careful review and constructive feedback. We address each major comment below and will make the suggested revisions to improve the clarity of the results.
read point-by-point responses
-
Referee: [§4.2, Theorem 4.3] §4.2, Theorem 4.3: the proof of the general O(α^{-θ}) rate invokes a specific decay estimate on the modal coefficients of the terminal mismatch; the precise value of θ is not stated explicitly in the theorem statement, making it difficult to verify the claimed order without re-deriving the constant from the preceding lemmas.
Authors: We agree with the referee that the explicit value of θ should be included in the statement of Theorem 4.3 for better verifiability. In the revised version of the manuscript, we will explicitly state the value of θ in Theorem 4.3, derived from the decay estimate on the modal coefficients as used in the proof. This will allow readers to confirm the order without needing to re-derive it from the lemmas. revision: yes
-
Referee: [§5, Assumption 5.1] §5, Assumption 5.1: the stronger modal summability condition required for the sharp O(1/α) rate is formulated in terms of an ℓ¹-type series on the Fourier coefficients; it would strengthen the result to include a brief remark on how this condition can be checked for typical target states (e.g., smooth or analytic data).
Authors: We appreciate this suggestion to strengthen the presentation. In the revised manuscript, we will add a brief remark after Assumption 5.1 explaining how the ℓ¹ summability condition on the Fourier coefficients can be verified for typical target states. For instance, we will note that for smooth or analytic terminal mismatches, standard Sobolev or Gevrey class estimates ensure the required decay of the coefficients. revision: yes
Circularity Check
No significant circularity: direct convergence proof from penalized to constrained problem
full rationale
The paper establishes quantitative convergence rates for the penalized terminal-constrained heat-equation optimal control problem as the penalty parameter α tends to infinity. The derivation relies on the spectral decomposition of the heat semigroup and explicit estimates on the modal coefficients of the terminal mismatch. No step reduces a claimed prediction to a fitted parameter by construction, nor does any load-bearing premise rest on a self-citation whose content is itself unverified within the paper. The finite-dimensional prototype is presented only as an illustrative analogy for the projection mechanism and is not used to derive the infinite-dimensional rates. The distinction between the general O(α^{-θ}) rate and the sharp O(1/α) rate under stronger summability assumptions is stated explicitly and follows from the analysis rather than from any definitional equivalence or imported uniqueness theorem. The argument is therefore self-contained against external benchmarks.
Axiom & Free-Parameter Ledger
axioms (2)
- standard math The heat equation is well-posed in appropriate Sobolev spaces with the given control operator
- domain assumption The terminal state is reachable under the hard constraint
Lean theorems connected to this paper
-
IndisputableMonolith/Cost/FunctionalEquation.leanwashburn_uniqueness_aczel unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
We establish explicit quantitative convergence estimates of order O(α^{-θ}), including the sharp O(1/α) rate under stronger modal summability assumptions on the terminal mismatch.
-
IndisputableMonolith/Foundation/AbsoluteFloorClosure.leanabsolute_floor_iff_bare_distinguishability unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
The admissible target set A_T(y0) is defined via summability of (2λ_n / (1−e^{-2λ_n T})) |d_n|^2 < ∞ with d_n = y_{T,n} − e^{-λ_n T} y_{0,n}.
What do these tags mean?
- matches
- The paper's claim is directly supported by a theorem in the formal canon.
- supports
- The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
- extends
- The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
- uses
- The paper appears to rely on the theorem as machinery.
- contradicts
- The paper's claim conflicts with a theorem or certificate in the canon.
- unclear
- Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.
Reference graph
Works this paper leans on
-
[1]
Bertsekas.Constrained Optimization and Lagrange Multiplier Meth- ods
Dimitri P. Bertsekas.Constrained Optimization and Lagrange Multiplier Meth- ods. Academic Press, New York, 1982
work page 1982
-
[2]
Eduardo Casas. Boundary control of semilinear elliptic equations with pointwise state constraints.SIAM Journal on Control and Optimization, 31(4):993–1006, 1993
work page 1993
-
[3]
Reachable states for the distributed control of the heat equation.Comptes Rendus
Mo Chen and Lionel Rosier. Reachable states for the distributed control of the heat equation.Comptes Rendus. Math´ ematique, 360(6):627–639, 2022
work page 2022
-
[4]
Evans.Partial Differential Equations, volume 19 ofGraduate Stud- ies in Mathematics
Lawrence C. Evans.Partial Differential Equations, volume 19 ofGraduate Stud- ies in Mathematics. American Mathematical Society, 2 edition, 2010
work page 2010
-
[5]
R. Glowinski and J.-L. Lions. Exact and approximate controllability for dis- tributed parameter systems.Acta Numerica, 3:269–378, 1994
work page 1994
-
[6]
Martin Gugat and Enrique Zuazua. Exact penalization of terminal constraints for optimal control problems.Optimal Control Applications and Methods, 37(5):909–924, 2016
work page 2016
-
[7]
Michael Hinze, Ren´ e Pinnau, Michael Ulbrich, and Stefan Ulbrich.Optimization with PDE Constraints. Springer, 2009
work page 2009
-
[8]
Kazufumi Ito and Karl Kunisch.Lagrange Multiplier Approach to Variational Problems and Applications. SIAM, 2008
work page 2008
-
[9]
S.-S. Kwon and M. Nunda. Numerical investigation of soft-to-hard terminal constraint convergence, 2026. Manuscript in preparation
work page 2026
-
[10]
Lions.Optimal Control of Systems Governed by Partial Differential Equa- tions
J.-L. Lions.Optimal Control of Systems Governed by Partial Differential Equa- tions. Springer-Verlag, 1971
work page 1971
-
[11]
Luenberger.Optimization by Vector Space Methods
David G. Luenberger.Optimization by Vector Space Methods. John Wiley & Sons, 1969
work page 1969
-
[12]
On optimal control problems with control and state constraints
Hans Maurer. On optimal control problems with control and state constraints. Optimal Control Applications and Methods, 29(1):43–62, 2008. 29
work page 2008
-
[13]
Daniel Meidner and Boris Vexler. A priori error estimates for space-time finite element discretization of parabolic optimal control problems. i. problems without control constraints.SIAM Journal on Control and Optimization, 47(3):1150– 1177, 2008
work page 2008
-
[14]
Daniel Meidner and Boris Vexler. A priori error estimates for space-time finite element discretization of parabolic optimal control problems. ii. problems with control constraints.SIAM Journal on Control and Optimization, 47(3):1301– 1329, 2008
work page 2008
-
[15]
Jorge Nocedal and Stephen J. Wright.Numerical Optimization. Springer, New York, 2 edition, 2006
work page 2006
-
[16]
Fredi Tr¨ oltzsch. Regular lagrange multipliers for control problems with mixed pointwise control-state constraints.SIAM Journal on Optimization, 15(2):616– 634, 2005
work page 2005
-
[17]
American Mathematical Society, 2010
Fredi Tr¨ oltzsch.Optimal Control of Partial Differential Equations: Theory, Methods and Applications, volume 112 ofGraduate Studies in Mathematics. American Mathematical Society, 2010. 30
work page 2010
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.