基于混沌时间序列模型的危险化学品泄漏事故预测

Prediction of Hazardous Chemicals Leakage Accidents Based on Chaotic Time Series Model

在对2010—2015年我国危险化学品(简称危化品)泄漏事故数量进行统计的基础上,建立了危化品泄漏事故混沌时间序列模型,预测了未来3年我国发生危化品泄漏事故的数量,并以2016年我国发生的危化品泄漏事故数量的实际值对模型的预测精度进行了验证。结果表明:2016年我国发生危化品泄漏事故数量的预测值与实际值的拟合度指标达到97.52%,拟合效果较好,说明所建立的危化品泄漏事故混沌时间序列模型可行;2017年和2018年我国发生危化品泄漏事故数量的预测值分别为932起和924起,相对2016年会有小幅度上升,其原因与中小型化工企业数量在我国化工企业总数中占比较高有关,这为研究我国危化品泄漏事故的发展趋势与制定预防措施提供了依据。

Based on the statistical data of hazardous chemicals leakage accidents in China during 2010—2015,this paper establishes a chaotic time series model for hazardous chemicals leakage to predict the number of accidents in the next three years.Then the paper verifies the accuracy of the predicted value by using the statistics of the number of hazardous chemicals leakage accidents in 2016.The result shows that the simulated and real numbers of China＇s hazardous chemicals leakage accidents in 2016 fit well,the goodness of fit index reaches 97.52%.That means the prediction model of hazardous chemicals leakage accidents is feasible.The prediction shows that the number of China＇s hazardous chemicals leakage accidents is 932 and 924 in 2017 and 2018 respectively,exhibiting a slightly increase compared to 2016,which may relate to the situation that small and medium-sized chemical enterprises account a big proportion of China＇s chemical enterprises.The paper provides some reference to the study of hazardous chemicals leakage accidents in China,and the formulation of preventive measures.