§3 equation (3) — the two-state Bellman equation
ProvedRossQC.Sufficient.eq_3bellman-equationdynamic-programmingp2o-batch-pfp2ap2o-gran-per-chapterp2o-plan-paperp2o-v1quality-control
Let , , and . For every bad-state belief , the finite-horizon values defined by equation (4) converge to a bounded discounted value , and
where .
This equation characterizes the action costs used by all the remaining statements in this mission.
Formalization Note is the limit of value iteration starting at zero, rather than an arbitrary function assumed to solve the equation. Boundedness and pointwise convergence are explicit.
Preamble
import Definitions.Def_RossQC_Sufficient_Model open Filter
Formal statement
namespace RossQC.Sufficient
/-- Ross, *Quality Control under Markovian Deterioration*, §3, equation (3),
p. 590 (unnumbered claim around the displayed equation). The finite-horizon
values from (4) converge to a bounded `V_β` satisfying (3) on the belief
interval. -/
theorem eq_3 (M : Model)
(hβ0 : 0 < M.β) (hβ1 : M.β < 1)
(hπ0 : 0 ≤ M.π) (hπ1 : M.π ≤ 1)
(hC0 : 0 < M.C) (hCI : M.C < M.I) (hIR : M.I < M.R) :
(∀ P ∈ Set.Icc (0 : ℝ) 1,
Tendsto (fun n : ℕ => M.valueIter n P) atTop (nhds (M.value P))) ∧
(∃ B : ℝ, ∀ P ∈ Set.Icc (0 : ℝ) 1, |M.value P| ≤ B) ∧
(∀ P ∈ Set.Icc (0 : ℝ) 1,
M.value P = min (M.rhs3 M.value P .produce)
(min (M.rhs3 M.value P .inspect) (M.rhs3 M.value P .revise))) := by sorry
end RossQC.Sufficient
Source
Ross, Quality Control under Markovian Deterioration, Management Science 17(9):587–596 (1971), DOI 10.1287/mnsc.17.9.587, p. 590, §3, equations (3)–(4)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.