基于Volterra模型的电力物资需求动态预测方法  

Dynamic Forecasting Method of Power Material Demand Based on Volterra Model

作　　者：钟炯聪 ZHONG Jiongcong(China Southern Power Grid Supply Chain(Guangdong)Co.,Ltd.,Guangzhou 51ooo0,China)

机构地区：[1]南方电网供应链(广东)有限公司,广东广州510000

出　　处：《微型电脑应用》2024年第10期223-226,共4页Microcomputer Applications

摘　　要：针对电力物资需求量大、种类繁多、电力物资需求量预测误差较大等问题,设计电力物资需求预测系统,该系统由客户服务、数据中心和云通信3层结构组成,通过SQL Server数据库的电力物资数据信息访问,能够实现对电力物资历史数据和现有数据的分类;利用数学模型进行计算定量预测,实现电力物资需求预测的实时更新;采用STM32F103C8T6微控制器和NFC标签读写模块,实现结合电力物资需求量对电力物资库存数量的调整。通过实验,所研究电力物资需求预测系统的电力物资需求预测平均误差为11%,电力物资需求动态预测误差低。Aimed at the problems of large demand for power materials,wide variety,and larger error in forecasting power mate-rial demand,this research designs a power material demand forecasting system which is composed of three-layer structure of customer service,data center and cloud communication.The power material data information access of SQL Server database can realize the classification of historical data and existing data of power materials.The system uses mathematical models to calcu-late and quantitatively predict,and realizes real-time update of power material demand forecast.It uses STM32F103C8T6 mi-crocontroller and NFC tag to read the writing module to realize the adjustment of the inventory quantity of power materials in combination with the demand for power materials:Through experiments,the average error of power material demand forecast of the research power material demand forecasting system is 11%,and the dynamic forecasting error of power material demand is low.

关 键 词：电力物资 需求预测 库存管理 供应链 Volterra反馈预测算法 

分 类 号：TM206[一般工业技术—材料科学与工程]