长沙市金属和类金属职业病危害因素检测超标率时间序列分析

发布时间：2018-03-22 21:10

本文选题：金属与类金属　切入点：职业病危害因素　出处：《中国卫生统计》2017年03期 　论文类型：期刊论文

【摘要】：目的了解长沙市金属和类金属职业病危害因素检测超标率变化趋势,建立季节性ARIMA模型进行短期预测,为控制工作场所金属和类金属危害提供措施建议。方法利用2008-2015年检测数据为训练集进行建模,利用2016年上半年数据为验证集检验,并进一步预测2016-2017年的超标率。结果金属和类金属超标率长期趋势为逐年下降,且具有一定的季节效应,其中二、三季度较高,一、四季度较低。建立ARIMA(0,1,1)×(0,1,1)_4模型,经检验残差为白噪声序列,预测的平均绝对误差(MAE)为4.54%。结论长沙市金属与类金属危害的控制近年来取得了一定的成效,超标率呈逐年下降趋势。2017年第二季度是金属与类金属危害控制的重点。季节性ARIMA模型在金属与类金属检测超标率预测方面有一定的可行性,今后可以和其他模型联合降低预测误差。
[Abstract]:Objective to investigate the trend of detection of metal and metal-like occupational disease hazards in Changsha and to establish a seasonal ARIMA model for short-term prediction. Methods using the 2008-2015 test data to model the training set, and using the first half of 2016 data as the validation set test, The results showed that the long-term trend of metal and similar metals exceeded the standard rate in 2016-2017 was decreasing year by year and had a seasonal effect, in which the second and third quarters were higher, and the first and fourth quarters were lower. After testing the residual error is white noise sequence, the predicted mean absolute error (MAE) is 4.54.Conclusion in recent years, some achievements have been made in the control of metal and metal-like hazards in Changsha. The rate of exceeding standard is decreasing year by year. In the second quarter of 2017, it is the key point of hazard control of metals and similar metals. The seasonal ARIMA model has some feasibility in predicting the rate of exceeding standard for metal and similar metals. The prediction error can be reduced in conjunction with other models in the future.
【作者单位】：长沙市疾病预防控制中心职业病防治所;中南大学湘雅公共卫生学院;
【基金】：2016年湖南省卫生计生委科研项目(B2016231)
【分类号】：R134

【相似文献】