摘要
目的:探讨建立血液供需预测模型在优化血液管理及保证血液临床供应中的应用效果。方法:分别应用灰色预测和BP神经网络模型对2007-2019年潮州市中心血站采集量和临床供血量进行统计分析,利用平均绝对百分比误差法(MAPE)对单一预测结果进行权重分配,建立组合预测模型(BGM混合模型)。最后利用BGM混合模型,对2020年、2021年和2022年的采集量和临床供血量进行短期预测。结果:BGM混合模型在血液采集和供给量的相对误差分别为1.62%、2.69%,均优于单一的灰色预测模型和BP神经网络模型。结论:血液供需预测模型在血站采集、制备、供给制定计划及强化库存科学管理方面,具有一定的指导意义。
Objective:To explore the application effect of establishing blood supply and demand prediction model in optimizing blood management and ensuring blood clinical supply.Methods:The gray prediction and BP neural network model were used to analyze the blood collection volume and clinical blood supply volume from 2007-2019.The mean absolute percentage error method(MAPE)was used to assign weight to the single prediction results,and a combined prediction model(BGM混合模型)was established.Finally,the combined forecasting model was used to make short-term forecasts of the collection volume and clinical blood supply volume in 2020,2021 and 2022.Results:The relative error of the combined prediction model in blood collection and supply was 1.62% and 2.69%,respectively,both were superior to a single gray prediction model and a BP neural network model.Conclusion:Blood supply and demand prediction model has certain guiding significance in blood station collection,preparation,supply planning and strengthening the scientific management of stock.
作者
陈婷
肖明星
周筠
Chen Ting;Xiao Mingxing;Zhou Yun(Chaozhou Blood Station,GuangDong Chao Zhou 521000)
出处
《中国社区医师》
2021年第22期86-87,共2页
Chinese Community Doctors
基金
潮州市卫生健康局科研项目(潮卫科研2019120号)。
关键词
灰色预测
BP神经网络
组合预测模型
血液供需
Gray prediction
BP neural network
Combined prediction model
Blood supply and demand
