AcceptedTheorem · Statement with omitted proof
Existence and stability of a one-dimensional weak solution
The weak Dirichlet problem has a unique solution and an energy bound in terms of the forcing.
Larsson & Thomée ↗
Theorem 2.4 · printed p. 22
Formal statement. The assumptions and conclusion are explicit; the theorem proof is omitted (sorry).
01 / MATHEMATICAL STATEMENT
Meaning & conditions
Mathematical summary
The weak Dirichlet problem has a unique solution and an energy bound in terms of the forcing.
Assumptions & context
- Square-integrable forcing and the coefficient coercivity condition in equation (2.12), encoded explicitly in the full Lean file.
Source
Larsson & Thomée. Partial differential equations with numerical methods.
Theorem 2.4 · printed p. 22
02 / LEAN REPRESENTATION
Formal statement
Statement with omitted proof. Expand the complete file below to read imports, local definitions and surrounding context.
theorem thm_2_4 (a b c : ℝ → ℝ) (a₀ : ℝ) (h212 : Assumption212 a b c a₀)
(f : Euclidean 1 → ℝ) (hf : InL2 Ω1 f) :
∃ u : Euclidean 1 → ℝ, SolvesVariational a b c f u ∧
(∀ v : Euclidean 1 → ℝ, SolvesVariational a b c f v →
v =ᵐ[domainMeasure Ω1] u) ∧
(∀ G : Fin 1 → Euclidean 1 → ℝ, HasH1Gradient Ω1 u G →
H1Norm u (G 0) ≤ (2 / a₀) * L2Norm f) := by
sorry
Complete Lean file and local definitions
import PDETextbook.Base
open MeasureTheory
open scoped BigOperators ENNReal
noncomputable section
namespace PDETextbook.LarssonThomee.Thm24
/-- The interval Ω = (0,1) from (2.11), realized inside one-dimensional Euclidean space. -/
def Ω1 : Set (Euclidean 1) := {x | 0 < x 0 ∧ x 0 < 1}
/-- Assumption (2.12), together with the standing Section 2.3 assumption that the
coefficients a, b, c are smooth. Smoothness is taken on Ω̄ = [0,1]. The conditions are
`a(x) ≥ a₀ > 0` and `c(x) - b'(x)/2 ≥ 0` for `x ∈ Ω̄ = [0,1]`. -/
def Assumption212 (a b c : ℝ → ℝ) (a₀ : ℝ) : Prop :=
ContDiffOn ℝ (⊤ : ℕ∞) a (Set.Icc 0 1) ∧
ContDiffOn ℝ (⊤ : ℕ∞) b (Set.Icc 0 1) ∧
ContDiffOn ℝ (⊤ : ℕ∞) c (Set.Icc 0 1) ∧
0 < a₀ ∧
∀ x ∈ Set.Icc (0 : ℝ) 1, a₀ ≤ a x ∧ 0 ≤ c x - derivWithin b (Set.Icc 0 1) x / 2
/-- Bilinear form of the variational formulation of (2.11). Multiplying
`-(a u')' + b u' + c u = f` by a test function and integrating by parts gives
`a(v,w) = ∫₀¹ (a v' w' + b v' w + c v w) dx`, where `v'` and `w'` are weak derivatives. -/
def bform (a b c : ℝ → ℝ) (v v' w w' : Euclidean 1 → ℝ) : ℝ :=
∫ x in Ω1, (a (x 0) * v' x * w' x + b (x 0) * v' x * w x + c (x 0) * v x * w x)
/-- `L(w) = (f,w) = ∫₀¹ f w dx`. -/
def Lform (f w : Euclidean 1 → ℝ) : ℝ :=
∫ x in Ω1, f x * w x
/-- `u ∈ H₀¹` is a solution of (2.16): `a(u,φ) = L(φ)` for all `φ ∈ H₀¹`.
H₀¹ is the closure of C₀¹, matching the source's density remark. -/
def SolvesVariational (a b c : ℝ → ℝ) (f u : Euclidean 1 → ℝ) : Prop :=
InH10Closure Ω1 u ∧
∀ G : Fin 1 → Euclidean 1 → ℝ, HasH1Gradient Ω1 u G →
∀ φ : Euclidean 1 → ℝ, InH10Closure Ω1 φ →
∀ H : Fin 1 → Euclidean 1 → ℝ, HasH1Gradient Ω1 φ H →
bform a b c u (G 0) φ (H 0) = Lform f φ
/-- `‖v‖₁ = (‖v‖² + ‖v'‖²)^{1/2}`, where `v'` is the weak derivative. -/
def H1Norm (v v' : Euclidean 1 → ℝ) : ℝ :=
Real.sqrt ((∫ x in Ω1, v x ^ 2) + ∫ x in Ω1, v' x ^ 2)
/-- `‖f‖` in `L₂(Ω)`. -/
def L2Norm (f : Euclidean 1 → ℝ) : ℝ :=
Real.sqrt (∫ x in Ω1, f x ^ 2)
/-- Theorem 2.4 (Larsson–Thomée). Assume (2.12) and `f ∈ L₂`. Then (2.16) has a unique
solution `u ∈ H₀¹`, unique as an a.e. class. This solution satisfies (2.20):
`‖u‖₁ ≤ C‖f‖` with `C = 2/a₀`. -/
theorem thm_2_4 (a b c : ℝ → ℝ) (a₀ : ℝ) (h212 : Assumption212 a b c a₀)
(f : Euclidean 1 → ℝ) (hf : InL2 Ω1 f) :
∃ u : Euclidean 1 → ℝ, SolvesVariational a b c f u ∧
(∀ v : Euclidean 1 → ℝ, SolvesVariational a b c f v →
v =ᵐ[domainMeasure Ω1] u) ∧
(∀ G : Fin 1 → Euclidean 1 → ℝ, HasH1Gradient Ω1 u G →
H1Norm u (G 0) ≤ (2 / a₀) * L2Norm f) := by
sorry
end PDETextbook.LarssonThomee.Thm24
How to read the formalization
- Uniqueness is expressed as almost-everywhere equality, consistent with the Sobolev-space setting.