{"id":"7ddd1598-762e-4bc9-871f-af7f5c4b8fdc","arxiv_id":"2412.04499","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Defines Stokes-Lagrange structures for N-dimensional port-Hamiltonian systems, showing that Stokes-Dirac and Stokes-Lagrange representations are equivalent and useful for control design.","lead":"This paper extends the port-Hamiltonian framework by defining Stokes-Lagrange structures on N-dimensional domains, which let energy be defined implicitly through differential operators and boundary ports. The framework unifies several known representations of wave, plate, and Maxwell systems and offers new ways to control them through energy ports.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 6's co-isotropy rests on Assumption 4, which is neither verified for the paper's examples nor proved without a domain gap in the wave-like case (Lemma 19).","rationale":"I read Theorem 6 as a conditional algebraic statement, and its two-line proof is essentially correct: isotropy follows from Assumption 3, and co-isotropy follows from Assumption 4 plus density of the boundary range. The paper's own Remark 16 acknowledges the Assumption 2 restriction, so that is a scope limitation rather than a hidden flaw. The genuinely load-bearing point is that the theorem transfers to the examples only if Assumption 4 holds there. The current text gives no verification for the plate or Maxwell cases, and the one general verification (Lemma 19) has a domain-theoretic gap. The typos catalogued by the reader (Reissner-Mindlin signs, Maxwell 1/epsilon0, dx/ds) are real and justify the conditional verdict, but they are separate from this structural concern. I would keep the verdict CONDITIONAL: the framework is promising and likely correct, but the maximality condition must be checked in the examples and the domain gap in A.2 closed before unconditional acceptance.","tokens_in":32502,"tokens_out":21665,"duration_ms":196947,"concrete_test":"For the wave-like class in A.2, prove or disprove that D((K†ηK|_{Z0})*) = D(K†ηK) when Assumptions 2-5 hold, using the definition of K† as the adjoint of K on ker(γ). If the domains differ, exhibit an element z̃ in the adjoint domain but not in D(K†ηK); if they coincide, redo Lemma 19 without the premature adjoint steps. For the examples, compute Z0=ker(γ)∩ker(β) and check the inclusion ker([S0^*, -P0^*]) ⊂ Im([P;S]) for the Reissner-Mindlin and Maxwell operators.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is conditional: L=L^{⊥_-} holds only if Assumptions 2-5 are satisfied. The co-isotropy inclusion is exactly Assumption 4 (maximality), so anything that casts doubt on Assumption 4 for the advertised systems weakens the paper's main contribution. In the plate and Maxwell examples, gamma and beta are never defined and Assumption 4 is not checked; the paper merely asserts the Stokes-Lagrange operators and power balances. For the general wave-like class, Appendix A.2 (Lemma 19) attempts to prove maximality. The proof uses the steps <K†(ηK)z0, z̃> = <ηK z0, K z̃> and <K z0, ηK z̃> = <z0, K†(ηK)z̃> before z̃ has been shown to belong to D(K) or D(K†ηK); these identities are only valid on the respective adjoint domains. If D((K†ηK|_{Z0})*) is strictly larger than D(K†ηK), the conclusion e1=K†(ηK)z̃ does not follow. This is a gap in the proof that the canonical construction is a Stokes-Lagrange structure, hence in the equivalence Theorem 1. It is probably fixable by a standard adjoint-domain argument, but it should be closed explicitly before the framework can be considered fully rigorous.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes an N-dimensional extension of the 1D Stokes-Lagrange subspaces introduced in prior work, defining a Stokes-Lagrange structure as the image of operators (P,S,γ,β) satisfying symmetry, maximality, and density assumptions, and proving in Theorem 6 that such a subspace is a Lagrange structure with respect to the minus pairing. It also establishes an operator-transposition equivalence between Stokes-Dirac and Stokes-Lagrange representations, and illustrates the framework on Reissner-Mindlin and Kirchhoff-Love plates, Maxwell's equations, and the Dzektser equation, with applications to passivity-based control via energy ports. The central theorem is plausible and the proof structure in Appendix C.1 is coherent, but the manuscript contains concrete sign and coefficient errors in the examples and an unclosed domain issue in the maximality proof for the canonical wave-like construction.","tokens_in":32712,"tokens_out":11282,"duration_ms":94741,"significance":"If Theorem 6 and the representation equivalence Theorem 1 are made fully rigorous, the paper provides a useful operator-theoretic framework for implicit Hamiltonian definitions and boundary energy ports in N-dimensional port-Hamiltonian systems, complementing Stokes-Dirac structures and enabling new interconnection schemes. The examples are well chosen and the control section gives a clear motivation for energy ports. However, the significance is conditional: the examples do not verify the assumptions of Theorem 6, and the proof that the canonical construction satisfies the maximality assumption has a gap. These are load-bearing issues, though they appear repairable. The paper is not accompanied by code or machine-checked proofs; its value is conceptual and methodological.","major_comments":[{"comment":"The variational derivative δφH_SL_RM is computed with the wrong sign. Since H_SL_RM contains the term (1/2)∫ kGh|grad w − φ|², the contribution to δφ is −kGh(grad w − φ), not +kGh(grad w − φ). The correct expression is δφH_SL_RM = −kGh(grad w − φ) − Div(D Grad φ). Consequently, in the matrix S^SL_RM the (3,1) entry should be −kGh grad(·), not +kGh grad(·). With the signs as printed, the constitutive relation e^SL = S^SL_RM α^SL does not reproduce the Reissner-Mindlin dynamics (17) through (20). This is load-bearing because the claimed Stokes-Lagrange representation of the plate model is one of the central examples.","section":"Section 3.1.2, Lemma 9 and S^SL_RM"},{"comment":"The displayed operator S^SL is inconsistent with the Hamiltonian H_SL_EM. Since e^SL = (δ_D H_SL_EM, δ_A H_SL_EM) = ((1/ε0)D, curl((1/µ0)curl A)), the operator S^SL must be diag((1/ε0)I3, curl((1/µ0)curl·)). The paper displays S^SL = diag(I3, curl((1/µ0)curl·)), omitting the 1/ε0 factor. In addition, P^SL is written as [[I3,0],[0,1 I3]], which is not the identity operator used in the state equation. As printed, the constitutive relation for the electric effort does not match the Hamiltonian.","section":"Section 3.2.2, S^SL and P^SL"},{"comment":"The examples in the plate and Maxwell sections are presented as Stokes-Lagrange structures without verifying the assumptions of Theorem 6. In both sections the latent space Z1, the boundary operators γ and β, and the maximality and density conditions are not specified. For the Reissner-Mindlin plate, the boundary energy ports in Theorem 10 suggest γ(w,φ) = (w|∂Ω, φ|∂Ω) and β(w,φ) = (kGh(grad w − φ)·n, D Grad φ n), but this identification is not stated and Assumption 4 is not checked. Without such checks, the power-balance computations alone do not establish that the displayed subspaces are Stokes-Lagrange structures in the sense of Definition 5.","section":"Sections 3.1.2 and 3.2.2 (Assumptions 3-5)"},{"comment":"The maximality proof contains an adjoint-domain gap. Before z̃ is known to lie in D(K) or D(K†ηK), the proof uses the identities ⟨K†(ηK)z0, z̃⟩ = ⟨ηK z0, K z̃⟩ and ⟨K z0, ηK z̃⟩ = ⟨z0, K†(ηK)z̃⟩. These identities are only justified when the relevant vectors belong to the corresponding adjoint domains; at that point z̃ is only known to lie in D(S0*). The conclusion z̃1 ∈ D(K†ηK) and e1 = K†(ηK)z̃1 therefore does not follow as written. This is load-bearing because Lemma 19 is the verification of Assumption 4 for the canonical construction behind the wave-like equivalence in Theorem 1. The gap is probably fixable by a standard adjoint-domain argument, but it must be closed explicitly.","section":"Appendix A.2, Lemma 19"}],"minor_comments":[{"comment":"The second integral in the power balance is written over ∂Ω, but it should be over Ω: the term is the distributed power ∫Ω J·E dx, not a boundary integral.","section":"Theorem 14"},{"comment":"The claim that ∂t div(D) = 0 follows from div curl = 0 omits the current term. From ∂t D = curl H − J one obtains ∂t div D = −div J, so preserving div D = 0 requires div J = 0.","section":"Remark 27"},{"comment":"There is a typo in the definition P0 := S|Z0; it should read P0 := P|Z0.","section":"Assumption 4 and Lemma 23"},{"comment":"The displayed total Hamiltonian is missing the squares on p and ε. It should read H̃ = (1/2)∫ (1/ρ)p² + kε² + (1/ε0)D² + (1/µ0)B² + 2qDε dx.","section":"Example 4"},{"comment":"After imposing the Kirchhoff-Love constraints, the reduced Hamiltonian is still denoted H_SL_RM; it should be H_SL_KL. The same typo appears in Theorem 11.","section":"Section 3.1.2, Kirchhoff-Love reduction"},{"comment":"There are unresolved citation placeholders, for example '[ ?]' in Section 1 and in Section 3.3 (Dzektser equation). These should be completed before publication.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"To the editor: the paper is within the journal's scope and the conceptual contribution is potentially solid. My main concern is that the examples are asserted to be Stokes-Lagrange structures without verifying the assumptions of Theorem 6, and that the maximality proof for the canonical example has a real domain gap. Both issues are repairable, but a revision should either verify Assumption 4 for the examples or explicitly present them as formal computations pending functional-analytic justification. The sign and coefficient errors in the plate and Maxwell examples should also be corrected; independent verification of the variational derivatives is advisable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper deserves a serious look. It does something genuinely new: it takes Maschke–van der Schaft's 1D Stokes-Lagrange subspaces and defines an N-dimensional Stokes-Lagrange structure via four operators (P, S, γ, β) satisfying symmetry, maximality, and density assumptions, and proves in Theorem 6 that the image is a Lagrange subspace. The equivalence theorem between Stokes-Dirac and Stokes-Lagrange representations (Theorem 1) is clean and useful, and the examples—Reissner-Mindlin, Kirchhoff-Love, Maxwell, Dzektser—show the intended scope. The energy-port control section (Section 4) is a nice addition; it generalizes the recent interconnection schemes and gives a physical interpretation of the integral of a passive output.\n\nThe central theorem is plausible and the proof in C.1 is coherent under the stated Assumptions. But the paper has real soft spots.\n\nThe main one is in Appendix A.2. Lemma 19 is supposed to prove maximality for the wave-like canonical construction. The proof uses identities like ⟨K†(ηK)z1_0, z̃1⟩ = ⟨ηK z1_0, K z̃1⟩ and then ⟨Kz1_0, ηK z̃1⟩ = ⟨z1_0, K†(ηK)z̃1⟩ before showing that z̃1 belongs to D(K) or D(K†ηK). Those steps are only valid on the appropriate adjoint domains. As written, the proof has a domain gap. It is probably fixable by a standard density argument, but it needs to be closed explicitly. Since co-isotropy is exactly Assumption 4, this gap affects the main theorem for the abstract class.\n\nSecond, the examples in Section 3 are asserted to be Stokes-Lagrange structures but the paper never defines γ and β for the plate and Maxwell cases or checks Assumptions 3-5. The power-balance calculations are suggestive, but they don't substitute for verifying the definition. For Dzektser they do give γ and β but again no maximality check.\n\nThird, there are concrete errors: the S^SL_RM matrix has sign errors relative to Lemma 9 (the φ terms should be +div(kGh φ) and +kGh φ, not −); the Maxwell S^SL is missing the 1/ε0 factor in the D-block; Theorems 10 and 14 have boundary integrals with dx instead of ds; and there are two missing reference placeholders. These are typos, but in a paper whose contribution is a mathematical structure they should be corrected before publication.\n\nAssumption 2 is a real limitation—only one of P,S can be unbounded—but the paper states this honestly in Remark 16.\n\nWho is this for: anyone working on distributed port-Hamiltonian systems, especially implicit constitutive relations and boundary energy control. It deserves peer review, but the referee should ask for a fix of Lemma 19, verification of the examples, and a correction pass. I'd accept it for review, not for publication as-is.","headline":"Genuinely new N-D Stokes-Lagrange framework with a fixable domain gap in the main proof and several example typos; worth reviewing but needs revision.","tokens_in":33293,"tokens_out":5071,"would_cite":true,"duration_ms":43262,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["93C20","37K06","74K20","35Q61"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper defines a Stokes-Lagrange structure on $N$-dimensional domains and proves it is a Lagrange structure, giving port-Hamiltonian systems implicit Hamiltonians and boundary energy ports.","keywords":["Port-Hamiltonian systems","Stokes-Dirac structure","Stokes-Lagrange structure","implicit constitutive relations","boundary energy ports","plate models","Maxwell's equations","passivity-based control"],"falsifier":"Choose $P=-\\Delta$ and $S=-\\Delta$ on a bounded $N$-dimensional domain, with $\\gamma$ the Dirichlet trace and $\\beta$ the Neumann trace, and compute whether the image subspace $L=\\{(Pz,\\gamma z,Sz,\\beta z)\\}$ satisfies $L=L^{\\perp_-}$; if it fails, the boundedness assumption is essential and the framework does not cover such fourth-order-like systems.","tokens_in":32254,"feed_emoji":"⚡","tokens_out":9560,"duration_ms":94310,"temperature":0.7,"pith_summary":"This paper extends the port-Hamiltonian framework to $N$-dimensional spatial domains by introducing a Stokes-Lagrange structure: a subspace built from four operators that satisfy an integration-by-parts identity, playing for the Hamiltonian the role that a Stokes-Dirac structure plays for the dynamics. The central claim is that this subspace is a Lagrange structure under natural symmetry, maximality, and density assumptions, so it can define a Hamiltonian implicitly even when constitutive relations are nonlocal or involve unbounded operators. The paper shows that standard distributed systems, including the wave equation, Reissner-Mindlin and Kirchhoff-Love plates, Maxwell's equations, and the Dzektser seepage equation, admit both Stokes-Dirac and Stokes-Lagrange representations related by a transposition operator. If the claim is right, modelers can choose where differential operators live and can use boundary energy ports for passivity-based control.","feed_headline":"Stokes-Lagrange structure gives port-Hamiltonian systems energy ports","feed_subtitle":"Implicit Hamiltonians from differential operators let boundary energy, not just power, enter the model in any dimension.","key_machinery":"The load-bearing object is the Stokes-Lagrange structure itself, defined as the image of the operator block $\\begin{bmatrix}P&\\gamma&S&\\beta\\end{bmatrix}^\\top$ applied to a latent space $\\mathcal{Z}_1$, where $P$ maps latent states to state variables, $S$ maps them to effort variables, and $\\gamma$ and $\\beta$ record boundary observation and boundary control. Assumption 3 encodes an integration-by-parts identity (formal symmetry), Assumption 4 encodes maximal reciprocity, and Assumption 5 ensures density of boundary data; together they make the minus-pairing orthogonal of $L$ equal to $L$. The Hamiltonian is then $H(z)=\\frac12\\langle Sz,Pz\\rangle+\\frac12\\langle\\beta z,\\gamma z\\rangle$, which is explicit in the latent variable but implicit in the physical state. A secondary machinery is the transposition operator $G$ that converts a Stokes-Dirac representation into a Stokes-Lagrange one, moving differential operators from the structure matrix into the Hamiltonian.","core_discovery":"At the paper's core is Theorem 6: given operators $P$ (state map) and $S$ (effort map) from a latent space $\\mathcal{Z}_1$ to state and effort spaces, together with boundary operators $\\gamma$ and $\\beta$, and given symmetry, maximality, and density assumptions, the image subspace $L=\\{(Pz,\\gamma z,Sz,\\beta z)\\mid z\\in\\mathcal{Z}_1\\}$ equals its orthogonal complement with respect to the minus pairing $\\langle\\langle(x_1,\\chi_1,e_1,\\varepsilon_1),(x_2,\\chi_2,e_2,\\varepsilon_2)\\rangle\\rangle_- = \\langle e_2,x_1\\rangle-\\langle e_1,x_2\\rangle+\\langle\\varepsilon_2,\\chi_1\\rangle-\\langle\\varepsilon_1,\\chi_2\\rangle$. In other words, a Stokes-Lagrange structure is a Lagrange structure. A companion result, Theorem 1, gives an equivalence between Stokes-Dirac and Stokes-Lagrange representations through a transposition operator $G$ and its formal adjoint $G^\\dagger$, so the same physical system can be written with differential operators in the structure matrix or in the Hamiltonian. The examples then show how implicit constitutive relations, including Eringen-type nonlocal elasticity, plate models, and the vector-potential form of Maxwell's equations, fit the new structure and yield boundary energy ports rather than only boundary power ports.","pith_inferences":["A natural step the paper leaves open is a structure-preserving discretization of the minus-pairing orthogonality; a mixed finite-element method preserving $L=L^{\\perp_-}$ would likely keep nonlocal models sparse and could be tested on the Dzektser equation.","The boundedness assumption, at most one of $P$ and $S$ unbounded, is the main restriction; extending the proof to both-unbounded cases would require working on $D(P)\\cap D(S)$ and would open up fourth-order plate and beam models.","The energy-port viewpoint suggests a general passivity-based design rule: take as interconnection variable the boundary coordinate whose time derivative is the passive output, then shape the total Hamiltonian by adding a potential of those coordinates; the paper illustrates this on waves and piezoelectric coupling but does not state such a general recipe."],"forward_implications":["Any system that admits a Stokes-Dirac representation in the wave-like form (12) also admits an equivalent Stokes-Lagrange representation, so a modeler can choose whether differential operators sit in the structure matrix or in the Hamiltonian.","Implicit or nonlocal constitutive relations, such as Eringen-type elasticity or the Dzektser equation, get a well-defined Hamiltonian without expanding the flow and effort spaces, which enables sparse numerical formulations for nonlocal problems.","Boundary energy ports, whose coordinates are traces of state variables such as boundary displacement, become available as controls; interconnecting two such systems through a potential yields passivity with respect to a non-separable total Hamiltonian.","The framework covers 2D and 3D examples including Reissner-Mindlin and Kirchhoff-Love plates and Maxwell's equations in both classical and vector-potential forms, preserving the power-balance interpretation across representations."],"supporting_citations":[{"why":"Supplies the 1D Stokes-Lagrange subspace construction with implicitly defined energy that this paper extends to $N$ dimensions.","marker":"[33]"},{"why":"Provides the colligation-style definition of Dirac structures on Hilbert spaces and the symmetry and maximality template used for the Stokes-Lagrange structure.","marker":"[30]"},{"why":"Gives the original boundary-energy-flow port-Hamiltonian formulation of distributed systems and the wave-equation example used throughout.","marker":"[52]"},{"why":"This is the authors' earlier 1D comparison of Stokes-Lagrange and Stokes-Dirac representations, whose extension is the announced goal of the paper.","marker":"[5]"},{"why":"Motivates the energy-port interconnection schemes by showing how integrated or differentiated passive outputs can be used for control.","marker":"[7]"},{"why":"Introduces energy and power ports for port-Hamiltonian systems, used here to justify the energy control ports in the Stokes-Lagrange setting.","marker":"[29]"},{"why":"Provides the Stokes-Dirac formulation and power balance for the Reissner-Mindlin plate used as a central example.","marker":"[8]"},{"why":"Introduces the Eringen nonlocal constitutive relation that motivates implicit Hamiltonians and is used for the nanorod and nonlocal examples.","marker":"[17]"},{"why":"Supplies the Dzektser seepage equation, the nonlocal dissipative example that closes the examples section.","marker":"[15]"}],"fun_headline_variants":["Stokes-Lagrange structure brings boundary energy ports to pH systems","Implicit Hamiltonian via Stokes-Lagrange: new control tools","Stokes-Lagrange vs Stokes-Dirac: equivalent but more flexible","Stokes-Lagrange structure: energy ports for higher-dimensional pH systems"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is Assumption 2, which requires at least one of the operators $P$ and $S$ to be bounded; if both are unbounded, the construction of the Stokes-Lagrange structure and Theorem 6 do not apply.","fun_headline_variants_meta":{"raw":{"variants":["Stokes-Lagrange structure brings boundary energy ports to pH systems","Implicit Hamiltonian via Stokes-Lagrange: new control tools","Stokes-Lagrange vs Stokes-Dirac: equivalent but more flexible","Stokes-Lagrange structure: energy ports for higher-dimensional pH systems"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000293,"raw_usage":{"total_tokens":1720,"prompt_tokens":968,"completion_tokens":752,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":584,"completion_tokens_details":{"reasoning_tokens":675}},"tokens_in":584,"tokens_out":752,"duration_ms":7164,"temperature":1.0,"reasoning_tokens":675,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T04:54:24.496766+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Choose $P=-\\Delta$ and $S=-\\Delta$ on a bounded $N$-dimensional domain, with $\\gamma$ the Dirichlet trace and $\\beta$ the Neumann trace, and compute whether the image subspace $L=\\{(Pz,\\gamma z,Sz,\\beta z)\\}$ satisfies $L=L^{\\perp_-}$; if it fails, the boundedness assumption is essential and the framework does not cover such fourth-order-like systems.","supporting_citations":[{"cited_title":"Kurula, H","cited_arxiv_id":null,"evidence_quote":"Provides the colligation-style definition of Dirac structures on Hilbert spaces and the symmetry and maximality template used for the Stokes-Lagrange structure."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the original boundary-energy-flow port-Hamiltonian formulation of distributed systems and the wave-equation example used throughout."},{"cited_title":"Bendimerad-Hohl, D","cited_arxiv_id":null,"evidence_quote":"This is the authors' earlier 1D comparison of Stokes-Lagrange and Stokes-Dirac representations, whose extension is the announced goal of the paper."},{"cited_title":"Borja, J","cited_arxiv_id":null,"evidence_quote":"Motivates the energy-port interconnection schemes by showing how integrated or differentiated passive outputs can be used for control."},{"cited_title":"Krha ˇc, B","cited_arxiv_id":null,"evidence_quote":"Introduces energy and power ports for port-Hamiltonian systems, used here to justify the energy control ports in the Stokes-Lagrange setting."},{"cited_title":"Brugnoli, D","cited_arxiv_id":null,"evidence_quote":"Provides the Stokes-Dirac formulation and power balance for the Reissner-Mindlin plate used as a central example."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the Eringen nonlocal constitutive relation that motivates implicit Hamiltonians and is used for the nanorod and nonlocal examples."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Dzektser seepage equation, the nonlocal dissipative example that closes the examples section."}],"review_version":1}