摘要
供应链环境中,产品需求预测对于制造业安排生产计划具有导向作用。对于大多数制造企业,通常生产的同一系列产品下有多种型号产品。由于受产能及市场容量限制,同系列下多型号产品之间对各自产品需求会产生相互影响作用。鉴于此,考虑历史需求数据在时间序列上的相邻关联性,研究在不同时间序列上各型号产品相互制约影响下产生的不同需求形态。同时考虑产品自身属性差异、供应链环境等影响因素,预测出某型号产品未来一段时间内的需求量。通过GRU-BP组合神经网络预测模型,对模型分析求解后证明预测结果的可行性。
In the supply chain environment,product demand forecasting plays an important role in the production planning of manufacturing industry.For most manufacturing enterprises,there are many types of products under the same series.Due to the limitation of production capacity and market capacity,multi-model products in the same series will interact with each other in terms of product demand.In view of this,considering the adjacent correlation of historical demand data in time series,the different demand patterns produced under the mutual restriction of different types of products in different time series were studied.At the same time,considering the factors such as product attributes difference and supply chain environment,the demand of a certain type of product in the future was predicted.Through GRU-BP combined neural network forecasting model,the feasibility of the forecasting results is proved after solving the model.
作者
李琼
陈佳
徐斌
Li Qiong;Chen Jia;Xu Bin(School of Maritime Economics and Management,Dalian Maritime University,Dalian 116026,Liaoning,China)
出处
《计算机应用与软件》
北大核心
2021年第3期59-69,共11页
Computer Applications and Software
