基于AI预测的国网协议库存智能补库研究

Research on Intelligent Replenishment of State Grid Agreement Inventory Based on AI Prediction

摘要: 围绕国家电网公司“绿色、数字、智能”现代供应链发展战略，上海电力公司积极响应加快数智供应链体系建设的要求，聚焦在实际运营中协议库存管理中存在的需求预测偏差大、库存周转效率低、数据协同能力弱等核心痛点，探索人工智能的协议库存智能补库研究。本文从多源数据融合、LSTM需求预测、动态库存优化等关键环节出发，构建集数据驱动、智能决策、业务适配为一体的协议库存智能补库模型，助力电网物资供应链敏捷性响应、运营效率与决策智能化水平有效提升。

Abstract: In alignment with State Grid Corporation’s “Green, Digital, Intelligent” modern supply chain development strategy, Shanghai Electric Power Company has actively responded to the call for accelerating digital and intelligent supply chain system construction. Focusing on core operational pain points in agreement inventory management—including significant demand forecasting deviations, low inventory turnover efficiency, and weak data collaboration capabilities—the company has explored AI-powered intelligent replenishment solutions for agreement inventory. This study develops an AI-driven model for intelligent replenishment of agreement inventory, integrating data-driven approaches, smart decision-making, and business adaptation. The model addresses key aspects such as multi-source data fusion, LSTM-based demand forecasting, and dynamic inventory optimization. These innovations enhance the agility of power grid material supply chains, improve operational efficiency, and elevate intelligent decision-making capabilities.