Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Eq. (32) — v^FR ≥ E[Σ_t Σ_j r_j x_j(t)] − Σ_j r_j (log T + 1) − Σ_j r_j λ_j

Proved
ResolvingNRM.FRUpper.fr_revenue_accounting

by mikedeng1 · Oct 6, 2026 · Mathlib 0df444a (Lean v4.33.1)

network-revenue-managementp2o-batch-pfp2bp2o-gran-per-chapterp2o-plan-paperp2o-v1poissonre-solving

Consider the network revenue management model with Poisson rates λj>0\lambda_j > 0λj​>0, revenues rj≥0r_j \ge 0rj​≥0, nonnegative consumption matrix AAA, initial capacity C≥0C \ge 0C≥0, and an integer horizon T≥1T \ge 1T≥1. Run the Frequent Re-solving policy FR (Algorithm 2) with any optimal-solution selector: at the start of period t∈{0,…,T−1}t \in \{0, \dots, T-1\}t∈{0,…,T−1} it solves the DLP with right-hand side b(t)=C(t)/(T−t)b(t) = C(t)/(T - t)b(t)=C(t)/(T−t), obtains x(t)x(t)x(t), and accepts class-jjj arrivals in [t,t+1)[t, t+1)[t,t+1) with probability xj(t)/λjx_j(t)/\lambda_jxj​(t)/λj​ subject to the capacity check. Then the expected revenue of FR satisfies

vFR ≥ E[∑t=0T−1∑j=1nrjxj(t)]−∑j=1nrj(log⁡T+1)−∑j=1nrjλj.v^{\mathrm{FR}} \ \ge\ \mathbb{E}\Big[ \sum_{t=0}^{T-1} \sum_{j=1}^n r_j x_j(t) \Big] - \sum_{j=1}^n r_j (\log T + 1) - \sum_{j=1}^n r_j \lambda_j .vFR ≥ E[t=0∑T−1​j=1∑n​rj​xj​(t)]−j=1∑n​rj​(logT+1)−j=1∑n​rj​λj​.

Since vDLP=T∑jrjxj∗v^{\mathrm{DLP}} = T \sum_j r_j x^*_jvDLP=T∑j​rj​xj∗​, this is equivalent to eq. (32):

vDLP−vFR≤T∑j=1nrjxj∗−E[∑t=0T−1∑j=1nrjxj(t)]+∑j=1nrj(log⁡T+1)+∑j=1nrjλj.v^{\mathrm{DLP}} - v^{\mathrm{FR}} \le T \sum_{j=1}^n r_j x^*_j - \mathbb{E}\Big[ \sum_{t=0}^{T-1} \sum_{j=1}^n r_j x_j(t) \Big] + \sum_{j=1}^n r_j (\log T + 1) + \sum_{j=1}^n r_j \lambda_j .vDLP−vFR≤Tj=1∑n​rj​xj∗​−E[t=0∑T−1​j=1∑n​rj​xj​(t)]+j=1∑n​rj​(logT+1)+j=1∑n​rj​λj​.

The inequality separates the revenue FR would earn if its LP targets were met exactly from the losses caused by the capacity check, which are logarithmic in TTT. It reduces Proposition 3 to a comparison of LP values.

Formalization Note. log⁡\loglog is the natural logarithm. E[∑t∑jrjxj(t)]\mathbb{E}[\sum_t \sum_j r_j x_j(t)]E[∑t​∑j​rj​xj​(t)] is written as ∑t=0T−1E[r⊤sel(C(t)/(T−t))]\sum_{t=0}^{T-1} \mathbb{E}[r^\top \mathrm{sel}(C(t)/(T-t))]∑t=0T−1​E[r⊤sel(C(t)/(T−t))] with frStateExp. The statement is the display at the foot of p. 34, which the paper rewrites as (32) by adding vDLP=T∑jrjxj∗v^{\mathrm{DLP}} = T\sum_j r_j x_j^*vDLP=T∑j​rj​xj∗​ to both sides. λj>0\lambda_j > 0λj​>0, r≥0r \ge 0r≥0, alj≥0a_{lj} \ge 0alj​≥0, C≥0C \ge 0C≥0 are the model's standing assumptions (p. 7).

Preamble
import Mathlib
import Definitions.Def_RLPBidPrice_Unbiased_Model
import Definitions.Def_ResolvingNRM_FRUpper_Model

open RLPBidPrice.Unbiased Matrix
Formal statement
namespace ResolvingNRM.FRUpper

/-- Eq. (32), App. C.2, pp. 34–35, in the form of the display at the foot of p. 34:
`v^FR ≥ E[∑_{t=0}^{T-1} ∑_j r_j x_j(t)] − ∑_j r_j (log T + 1) − ∑_j r_j λ_j`, where
`x(t) = sel (b(t))` and `b(t) = C(t) / (T − t)` along FR. -/
theorem fr_revenue_accounting {m n : ℕ} (A : Matrix (Fin m) (Fin n) ℝ) (r lam : Fin n → ℝ)
    (hlam : ∀ j, 0 < lam j) (hr : 0 ≤ r) (hA : ∀ l j, 0 ≤ A l j)
    (sel : (Fin m → ℝ) → (Fin n → ℝ)) (hsel : ResolvingNRM.IRT.IsDLPSelector A r lam sel)
    (T : ℕ) (hT : 1 ≤ T) (C : Fin m → ℝ) (hC : 0 ≤ C) :
    (∑ t ∈ Finset.range T,
        frStateExp A r lam sel T t
          (fun c => r ⬝ᵥ sel (fun l => c l / ((T : ℝ) - t))) C)
      - (∑ j, r j) * (Real.log T + 1) - ∑ j, r j * lam j
      ≤ frValue A r lam sel T C := by sorry

end ResolvingNRM.FRUpper
Source
Bumpensanti, Wang, A Re-solving Heuristic with Uniformly Bounded Loss for Network Revenue Management, arXiv:1802.06192v3, App. C.2, display at the foot of p. 34 and eq. (32), p. 35
Human review
  • Endorsed by Shuze Chen · Oct 6, 2026

    Confirmed by the moderator at approval.

  • Endorsed by mikedeng1 · Oct 6, 2026

    Confirmed by the mission captain (proposal self-audit).

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me