大数据挖掘在协议库存物资需求预测的研究和应用

Research and Application of Big Data Mining in Forecasting Stock Material Demand Forecast

国网福建电力公司开展协议库存物资需求预测,着重解决当前因协议库存物资需求计划不准确而造成物资结构性缺货、协议合同长期未执行等难题。通过整合ERP系统项目模块和物资模块的海量数据,利用Python算法库挖掘数据的内在联系,先对项目名称进行关键词选取及项目分类,再对项目类型、投资金额、地市、城乡等开展皮尔森相关分析,选取关键影响因子,最后采用随机森林构建大数据模型进行配网物资需求预测,大幅提升预测准确性。

State Grid Fujian Electric Power Co.,Ltd.carried out the agreement on the forecast of inventory demand,focusing on solving the problems of the current material shortage due to the inaccurate plan of the inventory inventory demand,and the long-term failure to implement the contract.By integrating the massive data of the ERP system project module and material module,the Python algorithm library is used to mine the internal relationship of the data,first select the keyword name and project classification of the project name,and then carry out the Pierre for the project type,investment amount,city,urban and rural areas.Sen related analysis,select the key impact factors,and finally use the random forest to build a big data model to forecast the distribution network material demand,and greatly improve the prediction accuracy.