Proof of Theorem 4.10 (p. 641) — the grid increments of a Markov wear process satisfy (4.3)–(4.5)
ProvedShockWear.WearProcess.proof410_incrementsLet satisfy (4.8), (4.9) and (4.10): almost every path starts at and is nondecreasing, the process is Markov for its natural filtration, and a version of is decreasing in both and for , , . Fix and put
Then are nonnegative random variables satisfying the conditions (4.3), (4.4) and (4.5) of Lemma 4.1b, for some choice of versions of the conditional laws of given .
This is the reduction of the continuous-time problem to the discrete dependent-damage model: the Markov property (4.9) gives (4.3), and the two monotonicities in (4.10) give (4.4) and (4.5).
Formalization Note The conclusion is ∃ κ', IsDamageSeq pr X κ'; nonnegativity in IsDamageSeq is almost sure, which is what (4.8) provides. Lean's wearIncr Z Δ i is the paper's .
import Mathlib import Definitions.Def_ShockWear_WearProcess_Model open MeasureTheory ProbabilityTheory Filter Topology open scoped ENNReal NNReal
namespace ShockWear.WearProcess
theorem proof410_increments {Ω : Type*} [MeasurableSpace Ω] (pr : Measure Ω)
[IsProbabilityMeasure pr] (Z : ℝ≥0 → Ω → ℝ) (κ : ℝ≥0 → ℝ≥0 → Kernel ℝ ℝ)
(hZ : IsWearProcess pr Z κ) (Δ : ℝ≥0) (hΔ : 0 < Δ) :
∃ κ' : ℕ → Kernel ℝ ℝ, IsDamageSeq pr (wearIncr Z Δ) κ' := by sorry
end ShockWear.WearProcess
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.