THEOREM 4 (a > 0) — F has the form (14) and sup |F(t) − t| = |F(1/2) − 1/2| = rT/(8 + 4rT)
ProvedNHPPArrivals.LinearRate.linear_degree_posdegree-of-nonhomogeneitylinear-arrival-ratep2o-batch-p200ap2o-gran-per-chapterp2o-plan-paperp2o-v1
Consider the linear arrival rate on with , and , and let be its relative slope. Let be the conditional cdf. Then
and the degree of nonhomogeneity is attained at :
The formula shows that depends on the rate only through the product , and it is the building block for THEOREM 5, where it is applied to each subinterval.
Formalization Note is §3.3's standing assumption ("we will assume that the arrival rate function is increasing, i.e., ", p. 472); is implicit in "the interval ". The supremum is stated as attained (IsGreatest of the image of ) and also as the value of the mission's degree.
Preamble
import Mathlib import Definitions.Def_NHPPArrivals_LinearRate_ConditionalCdf
Formal statement
namespace NHPPArrivals.LinearRate
theorem linear_degree_pos (a b T : ℝ) (hT : 0 < T) (ha : 0 < a) (hb : 0 ≤ b) :
(∀ t ∈ Set.Icc (0:ℝ) 1, condCdf (linRate a b) T t =
(t * T + (b / a) * (t * T) ^ 2 / 2) / (T + (b / a) * T ^ 2 / 2)) ∧
IsGreatest ((fun t => |condCdf (linRate a b) T t - t|) '' Set.Icc (0:ℝ) 1)
|condCdf (linRate a b) T (1 / 2) - 1 / 2| ∧
degree (condCdf (linRate a b) T) = |condCdf (linRate a b) T (1 / 2) - 1 / 2| ∧
|condCdf (linRate a b) T (1 / 2) - 1 / 2| =
1 / 2 - (T / 2 + (b / a) * T ^ 2 / 8) / (T + (b / a) * T ^ 2 / 2) ∧
1 / 2 - (T / 2 + (b / a) * T ^ 2 / 8) / (T + (b / a) * T ^ 2 / 2) =
(b / a) * T / (8 + 4 * (b / a) * T) := by sorry
end NHPPArrivals.LinearRate
Source
Kim and Whitt, Are call center and hospital arrivals well modeled by nonhomogeneous Poisson processes?, Manufacturing Service Oper. Management 16(3), 2014, p. 472, THEOREM 4, (14), (16); standing assumption b ≥ 0, §3.3
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.