Intelligent Prediction-Inventory-Scheduling Closed-Loop Nearshore Supply Chain Decision System

Abstract

This study proposes an intelligent prediction-inventory-scheduling closed-loop decision system for near-shore supply chain operations. By integrating three core modules-LSTM/Transformer demand forecasting, reinforcement learning inventory replenishment, and VRP path planning-the system achieves end-to-end collaborative optimization. An innovative "public health emergency" scenario generator is designed to quantitatively evaluate the system's robustness under extreme risks and its cost-inventory balance capability. Through heterogeneous model fusion, multi-objective dynamic optimization, and closed-loop feedback mechanisms, a spatiotemporal coupled decision framework is established. The system effectively mitigates prediction error propagation, optimizes inventory-path coordination, and demonstrates significant resilience enhancement during simulated emergencies.