用时间序列模型预测Rh阴性红细胞库存需求

How to forecast the demand for Rh negative red blood cell inventory using time-series methods

目的探究上海市血液中心Rh阴性各型悬浮红细胞历年临床用量变化规律,应用时间序列法分析动态发展趋势,建立相关的预测模型。方法根据上海市血液中心提供的2012年1月-2018年6月Rh阴性各型悬浮红细胞临床用量来绘制时间序列图,并用趋势季节模型对其进行分析和定量预测,用平均相对误差来评价此方法的拟合度。结果1)每年的1、2月是全年用血最少的月份,3、7、8、11、12月的临床用量则相对较多。2)模型对2018年1-6月的需求量作出定量预测,趋势季节模型平均相对误差为0.175。结论1)各型Rh阴性红细胞临床用量逐年增加;可根据趋势季节模型,确定每月库存上限与下限。2)模型对2018年上半年的Rh阴性红细胞需求量作出定量预测,预测效果良好。

Objective To explore the change pattern of the clinical Rh negative red blood cell demand in Shanghai Blood Center over the years,analyze the dynamic development trend by time series method,and to build the prediction models.Methods Based on the data of clinical blood demand in Shanghai Blood Center from January 2012 to June 2018,we generated the time series maps and built seasonal trend model to analyze and predict it.Average relative error(ARE)was used to evaluate the degree of fitness of the mode1 s.Results 1)Demands in January and February were the least throughout the whole year,while the demands in March,July,August,November and December were relatively larger than the rest months.2)The model made a quantitative forecast for the demand from January to June in 2018.The average relative error of seasonal model was 0.175.Conclusion 1)The clinical blood demand of various types of Rh negative red blood cell increases steadily every year.The monthly upper and lower limit can be determined based on the seasonal trend model.2)The predictions of the clinical Rh negative red blood cell demand in the first half of 2018 generated by this model is good.