← 研踪 PaperTrailarXiv:2611.10999 · v1
Differentiable Agent-Based Modeling for Multi-Echelon Supply Chains: Gradient-Based Optimization of Heterogeneous Policies at Scale
首次提交 2026-11-08 06:48
版本更新 2026-11-08 06:48
雷达观测 2026-11-10 12:16
摘要
Global supply chains rely on the interactions of thousands of heterogeneous agents, posing significant challenges for optimization. While traditional optimization often assumes homogeneity to ensure tractability, Reinforcement Learning struggles with sample inefficiency when scaling to high-dimensional state spaces. We propose a Differentiable Agent-Based Model for multi-echelon inventory management. By implementing stochastic dynamics via differentiable tensor operations, we enable direct gradient propagation through the simulation. We validate the framework by recovering theoretical Newsvendor optima without supervision. Crucially, we demonstrate ``Mass Personalization'' by optimizing 10,000 distinct agent policies simultaneously in seconds, achieving a 53% cost reduction compared to scalar baselines. Furthermore, we utilize gradient sensitivity analysis to generate causal heatmaps, accurately recovering inherent lead-time delays from the computation graph. This approach bridges the gap between simulation and optimization, enabling the efficient training of self-correcting Digital Twins.
作者备注：13 pages, 3 figures, 2026 Annual Modeling and Simulation Conference
观测到的版本
相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。
v1 · 2026-11-08 06:48 · 最新观测版本
Differentiable Agent-Based Modeling for Multi-Echelon Supply Chains: Gradient-Based Optimization of Heterogeneous Policies at Scale
Global supply chains rely on the interactions of thousands of heterogeneous agents, posing significant challenges for optimization. While traditional optimization often assumes homogeneity to ensure tractability, Reinforcement Learning struggles with sample inefficiency when scaling to high-dimensional state spaces. We propose a Differentiable Agent-Based Model for multi-echelon inventory management. By implementing stochastic dynamics via differentiable tensor operations, we enable direct gradient propagation through the simulation. We validate the framework by recovering theoretical Newsvendor optima without supervision. Crucially, we demonstrate ``Mass Personalization'' by optimizing 10,000 distinct agent policies simultaneously in seconds, achieving a 53% cost reduction compared to scalar baselines. Furthermore, we utilize gradient sensitivity analysis to generate causal heatmaps, accurately recovering inherent lead-time delays from the computation graph. This approach bridges the gap between simulation and optimization, enabling the efficient training of self-correcting Digital Twins.
查看此版本解读 →查看此版本原文 ↗