Adaptive Neutrosophic Goal Programming for SoH-Aware Tri-Objective Closed-Loop Supply Chain Optimisation under Battery Degradation Uncertainty

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:Provides a novel integration of battery degradation physics with neutrosophic optimization, offering a methodological template for uncertainty-aware CLSC models.

🏢実務担当者:Offers actionable SoH thresholds and a cost-emission ratio that can inform fleet management and carbon footprint estimation without complex infrastructure.

🏛政策担当者:Highlights the trade-off between operational efficiency and emissions, suggesting policy levers for battery replacement and recycling incentives.

<title>Abstract</title> <p> Background. Battery-powered logistics networks face a structurally complex challenge: progressive State-of-Health (SoH) degradation simultaneously contracts fleet capacity, inflates operational cost, and elevates parameter estimation uncertainty in ways that classical deterministic optimisation models cannot represent. This paper addresses the gap by embedding SoH directly into the feasible region of a tri-objective closed-loop supply chain (CLSC) linear programme in which all cost and emission parameters are encoded as adaptive Single-Valued Triangular Neutrosophic Numbers (SVTNN). <bold>Methods.</bold> The multi-echelon CLSC spans three suppliers, two manufacturing plants, three SoH-degradable distribution centres (capacity Capₑ(k) = 280 × SoH), four customer zones, two collection centres, and two remanufacturing centres over three planning periods. All 26 arc-level cost and emission parameters are initialised as SVTNN triplets (a, b, c; α, β, γ) and updated adaptively each period via three rules driven by SoH, State-of-Charge (SoC), and acoustic emission signal intensity. Three complementary solution methods are deployed: (i) a weighted-sum LP sensitivity sweep across seven SoH levels (1.00– 0.73) establishing the cost–emission–shortage baseline; (ii) an epsilon-constraint Pareto frontier at SoH = 0.92 over 29 emission budget levels; and (iii) Neutrosophic Goal Programming (NGP) with corrected constrained aspiration levels — G₁* = min Z₁ | Z₃ ≤ Z₃_WS, G₂* = min Z₂ | Z₃ ≤ Z₃_WS, G₃* = min Z₃ — and neutrosophic deviation weights wₜ = 0.60, w _i = 0.20, w_f = 0.20. The resulting NGP comprises 144 decision variables and 97 constraints per SoH level. <bold>Results.</bold> SoH degradation from 1.00 to 0.73 reduces Z₁ by 27.0% (₹9,780→₹7,139) and Z₂ by 27.0% (892→651 kg CO₂), while raising Z₃ from 975 to 1,052 units (83.6→ 90.2%). The invariant ratio Z₂/Z₁ = 0.0912 kg CO₂/₹ across all SoH levels enables carbon footprint estimation directly from cost accounts. The Pareto frontier reveals a near-constant exchange rate of 5.8 shortage units per 28 kg CO₂. The NGP yields LP-Optimal status at all seven SoH levels with d _i ⁺ = d _i ⁻ = 0 for all three objectives, confirming that the aspiration point is attainable with zero compromise. Additionally, Z₂_NGP lies 7.5–10.3 kg CO₂ below the weighted-sum LP emission at every degradation level — a routing-reallocation gain unavailable to the weighted-sum formulation. The adaptive SVTNN analysis identifies SoH = 0.80 as a diagnostic falsity-peak inflection warranting intensified battery management and SoH = 0.73 as the operational replacement threshold. <bold>Conclusions.</bold> The SVTNN–NGP framework is the first to jointly integrate battery degradation physics, adaptive neutrosophic uncertainty quantification, Neutrosophic Goal Programming with corrected aspiration levels, and an exact Pareto frontier into a multi-period closed-loop supply chain model. The derived SoH thresholds, emission–shortage exchange rate, cost–emission invariance rule, and NGP ideal-point attainability result are immediately deployable as fleet management and regulatory policy criteria without additional computational infrastructure. </p>

• Research Square https://doi.org/10.21203/rs.3.rs-9613524/v1first seen 2026-05-14 21:19:25 · last seen 2026-05-20 04:35:53