基于等维递补GM(1,1)模型的新疆法定传染病疫情预测

Analysis on the Epidemic Situation of Notifiable Diseases in Xinjiang Base on Model of Equidimensional Filling Vacancies

目的:探索建立预测新疆法定传染病疫情传播预测模型,为法定传染病病防治提供科学依据。方法:对新疆维吾尔自治区2004～2010年法定传染病发病率进行研究,建立静态GM(1,1)模型及动态等维递补GM(1,1)模型,并对两类模型拟合效果进行比较。结果:静态GM(1,1)模型以2006～2010年历史数据建立5维和2005～2010年历史数据建立的6维GM(1,1)模型精度较高,分别达到2级和3级.相应的递补模型精度即可达到1级,且预测收敛速度较原模型快。6维递补模型预测结果比较折中,5维递补模型预测结果收敛速度较快,预测结果较乐观。若采取折中方案预测,可以认为若采取有效地预防和治疗措施,新疆法定传染病发病率将在未来5年内逐年稳步下降,到2015年年发病率可降至每10万人不超过400人次。结论:针对新疆近期法定传染病疫情,6维递补灰预测模型拟合效果较好,可作为预测参考模型。

Objective：To establish a prediction model on the epidemic situation of notifiable diseases in Xinjiang,and provide a scientific basis for the prevention and control of infectious diseases.Methods：By studying of 2004～2010 years incidence of notifiable diseases in Xinjiang Uygur Autonomous Region,the static model and dynamic equivalent dimensions additional model were established,and the fitting effect on the two types of model are compared also.Results：5-dimensional and 6-dimensional static model base on historical data of 2006～2010 and 2005～2010 years have high precision,respectively,two and three.Accuracy of equivalent dimensions addition model can be achieved by one,and the predicted convergence rate is faster than the original models.The 6-dimensions additional model forecast results compare compromise and 5-dimensions additional predicted results faster convergence than optimistic forecast results.Take compromise forecast adopting effective prevention and treatment measures,incidence of Xinjiang notifiable diseases will has a steady decline in the next five years,which can be reduced to no more than 400 people per 100,000 population by 2015.Conclusion：In response to the epidemic situation of notifiable diseases in Xinjiang,6-dimensions additional gray forecasting model fit better as a prediction reference model.