THEOREM II, p. 869 — setting Sⱼ = Sⱼ₋₁ + Tⱼ gives the cost of the model with operations j, j+1 coupled
ProvedSimpsonInv.Vertex.theorem_IIConsider a model with operations, variables , constants and (), and cost
Fix with and suppose , i.e. the average inventory is zero. Form the coupled model with operations: inventory point is removed, operations and become one operation with processing time , so
and its variables are , with the same end constants . Then the original cost equals the coupled model's cost (the same formula with operations and data ):
The theorem says a vanishing inventory can be removed altogether without changing the cost; it is the reduction used for the second kind of face in the proof of LEMMA II.
Formalization Note Here the original model has variables (Fin (m + 1) → ℝ), and Lean's j : Fin (m + 1) is the paper's index = j + 1. Removing the variable is Fin.removeNth. No sign or domain hypothesis is needed (the page has none); the identity holds for every point satisfying .
import Mathlib import Definitions.Def_SimpsonInv_Vertex_Setting
namespace SimpsonInv.Vertex
theorem theorem_II {m : ℕ} (c S0 Sn : ℝ) (r T : ℕ → ℝ) (S : Fin (m + 1) → ℝ) (j : Fin (m + 1))
(hj : S j = extVec S0 Sn S j.val + T (j.val + 1)) :
cost c r T S0 Sn S =
cost c (coupledR r (j.val + 1)) (coupledT T (j.val + 1)) S0 Sn (Fin.removeNth j S) := by sorry
end SimpsonInv.Vertex
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.