Alphabeta Math
PropositionStatement: AI-adaptedProof: AI-generated
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Duhamel formula for a real-linear propagator

PDF p. 20: paragraph after (4.6), real-linearity qualification and Duhamel formulas

Statement

Assume the Axiom of Choice and hypotheses (H) and (W). In particular, Ω is the prescribed smooth connected exterior domain, T>0, λ∈R∖{0}, and the reference trajectory w belongs to C([0,T];HD2(Ω)∩W2,∞(Ω)). Let θ∈[−2,2], put E=HDθ(Ω), and let S(t,r) be the real-linear evolution for izt+ΔDz+V1z+V2z‾=0, where V1=2λ∣w∣2 and V2=λw2. Real-linear means linear over real scalars; complex linearity is not assumed. Choose R>0 with Θ⊂BR(0) and a smooth real cutoff φ:Rn→[0,1] which is zero for ∣x∣≤R, positive for ∣x∣>R, and one for ∣x∣≥R+1.

Use the following precise convention for a locally integrable spatial function q. At θ≥0, its regular scale realization Rθq exists when its almost-everywhere class belongs to HDθ⊂L2, and is that class. At θ=−s<0, it exists when qa‾∈L1(Ω) for every a∈HDs and there is a finite constant Cq such that ∣∫Ωqa‾ dx∣≤Cq∥a∥HDs for all such a; it is then the element of HD−s represented by this bounded conjugate-linear functional. This convention specifies the entire pivot functional at negative orders.

Let z0∈E and z∈C([0,T];E). Put Q=(0,T)×Ω and fix a jointly measurable representative w~ of the reference trajectory. For the nonlinear assertion, suppose there are jointly measurable complex-valued functions Z,h on Q with Z(t,⋅)∈Lloc3(Ω) and h(t,⋅)∈Lloc1(Ω) for almost every t, such that RθZ(t,⋅)=z(t) almost everywhere in time. Form the pointwise functions

qN=λ(2w~∣Z∣2+w~‾Z2+∣Z∣2Z),qg=φh.

Require their regular scale realizations N(t)=RθqN(t,⋅) and g(t)=Rθqg(t,⋅) to exist almost everywhere and to be strongly measurable E-valued functions with ∫0T(∥N(t)∥E+∥g(t)∥E) dt<∞. In the mild equation and formula below write F(w,z)=N and φh=g. Thus the nonlinear expression is constructed on the stated representative class, rather than on arbitrary negative-order elements of E.

Then z satisfies the free-flow mild formulation of izt+ΔDz+V1z+V2z‾+F(w,z)=φh, with initial datum z0, if and only if, for every t∈[0,T],

z(t)=S(t,0)z0+∫0tS(t,r)(iF(w,z)(r)) dr+∫0tS(t,r)(−iφh(r)) dr.

The integrals are Bochner integrals in the realification of E. For the linearized controlled equation the same formula holds with the F integral omitted. More generally this linear assertion holds for every z0∈E and every strongly measurable g∈L1(0,T;E) in place of φh, without a function-representative assumption on the state or forcing. A real-linear propagator need not commute with multiplication by i, so these factors are retained inside S(t,r).

Facts & Assumptions

[F1]

Under Choice, (H), and (W), E is a complex Banach space and W(t)=eitΔD is its strongly continuous unitary free group. The maps P(t)u=i(V1(t)u+V2(t)u‾) are bounded real-linear operators, norm-continuous in time; their evolution is the stated S(t,r) Dirichlet spectral Sobolev scale and duality Scale isometries and free Schrödinger evolution Time-dependent potential and conjugation bounds on the scale Forward and backward real-linear propagators The Axiom of Choice.

[F2]

For a strongly measurable f∈L1(0,T;E), the real-linear evolution result gives the unique continuous solution of u(t)=W(t)x+∫0tW(t−r)(P(r)u(r)+f(r)) dr as u(t)=S(t,0)x+∫0tS(t,r)f(r) dr. The integrals are Bochner integrals and the result applies to the realification of E Bounded real-linear perturbations preserve solution estimates.

[F3]

Bochner integration is additive for integrable functions with values in a real Banach space Linearity of the Bochner integral. The remainder equation and its signs are those of Quadratic-cubic remainder after subtracting a trajectory and Linearized equation and its real-duality adjoint.

[F4]

Strong measurability means almost-everywhere norm approximation by measurable simple functions, and Bochner integrability means approximation by integrable simple functions in the integral of the norm Strongly measurable Banach-valued function Bochner-integrable function. For a strongly measurable function into a real or complex Banach space, finite integral of the norm is equivalent to Bochner integrability Bochner integrability criterion.

[F5]

The hypotheses (H) and (W), including the measurable, essentially bounded function representative of w, are supplied by Exterior and truncated obstacle domains and Reference trajectory hypothesis W. The positive scale consists of L2 classes and the negative scale is the full pivot antidual specified in Dirichlet spectral Sobolev scale and duality. The cutoff and control convention are those of Control region and nested real cutoffs and Linear and nonlinear internally controlled exterior systems, and the pointwise polynomial and its local integrability under local cubic integrability are those of Quadratic-cubic remainder after subtracting a trajectory.

Proof

Given: The stated hypotheses, z0, z, the measurable representatives for the nonlinear assertion, and the two integrable realized forcings N,g.

Goal: Prove the equivalence and the linearized formula, preserving the positions of the factors i.

1.1F5given

For almost every time, local boundedness of w~ and local cubic integrability of Z make the polynomial qN locally integrable: on a compact set use ∣Z∣j≤1+∣Z∣3 for j=1,2 and the finite measure of that set. The bounded cutoff makes qg locally integrable as well. At nonnegative orders the required realizations are specified L2 classes in E. At negative orders the assumed absolute integrability and bound make a↦∫qa‾ a bounded conjugate-linear functional on all of HD−θ, hence a specified element of E by its antidual identification. All these assignments depend only on the almost-everywhere spatial function classes. Thus F(w,z)=N and φh=g have the precise scale meaning stipulated in the Statement.

2.1F1F3F4givenstep 1.1

Set f(r)=iN(r)−ig(r). A simple approximation for either forcing remains simple after multiplication by i, and its norm error is unchanged. Taking differences of the two simple approximations proves that f is strongly measurable in the realification of E; the norm bound ∥f(r)∥E≤∥N(r)∥E+∥g(r)∥E gives its integrability by [F4]. Multiplying the displayed equation formally by −i identifies its free-flow mild formulation as z(t)=W(t)z0+∫0tW(t−r)(P(r)z(r)+f(r)) dr. This identity is the mild meaning of the equation here; it does not assert a classical time derivative.

3.1F2step 2.1

Apply [F2] with x=z0 and this fixed forcing f. It gives equivalence of that free-flow identity with z(t)=S(t,0)z0+∫0tS(t,r)f(r) dr for all t. The argument applies to the particular f determined by z; it does not assume that F is linear in z.

4.1F2F3F4step 1.1step 2.1step 3.1

For each r, real-linearity gives S(t,r)f(r)=S(t,r)(iN(r))+S(t,r)(−ig(r)). The individual integrands are Bochner integrable by [F2], applied separately to the two forcings, which are integrable by [F4]. Additivity in [F3] yields the displayed two-integral formula with the notation of step 1.1. For the linearized equation apply exactly the same argument with N=0 and an arbitrary strongly measurable g∈L1(0,T;E). No product involving the state is needed in that argument, so no function representative is required; when g is the stated realization of φh, this gives the asserted controlled formula.

5.1step 4.1∎

Real-linearity alone does not imply complex-linearity: for example, on the realification of C, the evolution of u′=iu‾ is u(t)=cosh⁡(t)u(0)+isinh⁡(t)u(0)‾, as direct differentiation verifies. For t≠0 it sends i and i times the image of 1 to different values. Thus no interchange of i with a merely real-linear propagator is justified, and the formulas retain the factors inside. This completes the proof.

Source notes

The original article, frozen v2 PDF page 20 immediately after (4.6), supplies the attribution and the nonlinear and linearized formulas. Its image and corresponding extracted text were read for this attribution. The explicit regular-realization convention repairs the generated proposition's nonlinear domain; it is not an assertion that the authors defined a cubic operation on every negative-order distribution. The proof uses the named evolution, spectral-scale and Bochner suppliers, not the article as a proof premise. The conditions specify when the pointwise nonlinear forcing and the integrals are defined; no independent nonlinear existence or trace result is asserted here.

Depends on

Used by · 2

Dependency tree · 2 levels

Each result is drawn at its shortest distance from this item. Arrows run from prerequisites to the items that use them; this item has a heavier outline. External source nodes link directly to the cited text. The view shows up to two dependency steps; open a prerequisite to continue.

Sources