为了研究"清剿火患"行动对安徽省火灾形势的抑制作用,采用ARIMA干预模型与火灾数据对比的方法对2007—2012年安徽省火灾统计数据进行了分析。2种ARIMA干预模型都能检测到"清剿火患"行动对季节调整后火灾起数时间序列的抑制效果。根据干预模型,抑制效果在"清剿火患"行动开展后第3个月时出现,火灾立即减少约300起,随后抑制效果逐渐衰减。在"清剿火患"行动开展14个月后,抑制效果几乎衰减为0。通过对比"清剿火患"期间与2009-10—2010-02、2010-10—2011-02的火灾统计数据,发现"清剿火患"行动期间的火灾起数较前两年同期的火灾起数大幅度降低,平均降幅达64%。在此基础上又分析了不同火灾原因、火灾场所和火灾起数的降低情况。对比发现,"清剿火患"期间火灾死亡人数与死亡人数不低于3人的火灾起数均降低为0。
The purpose of this study is to explore the suppressive effects of extinguishing fire hazard( EFH) measures in Anhui.As a matter of fact,we have collected and analyzed the statistic data of the fire disasters from 2007 to 2012 in the province. The said fire-suppressive effects of EFH have been investigated by using an intervention mode of Average Auto-Regressive Integrated Moving( AARIM) model and the fire history statistics comparison. More specifically speaking,the two ARIMA intervention models tend to bring about the suppressive effect of EFH on the seasonally adjusted fire-accident series. In fact,the intervention models tend to offer the suppressive effect in the third month after the beginning of EFH,which tends to be of an exponential decay with a sudden "pulse"decrement( about 300 fires) in the third month after the start of the action,which would be followed by a gradual deceleration back to the original pre-intervention level with no permanent effect. And,then,14 months later,the intervention effect of the suppressive effect tends to cease with the use of the ARIMA intervention model. The so-called ARIMA intervention model can also help to depict the fluctuation taking place in the process of suppressive effect decay,which behaves more realistically than the previous models. In addition,the comparison of the fire statistics results indicates that the number of fires that took place in the period of EFH can significantly be reduced in comparison with that in the corresponding period of the past two years by an average decrease rate of 64%. However,when we analyzed the causes for the fire probability reduction,we have found that the highest decrease rate can be as high as that of88%. What is more,a significant reduction frequency has taken place in the higher-hazard-prone areas,too. Moreover,it would be possible even for the places where fire accidents most likely occur,such as the 3-in-1 sites and the nursing homes,have been managed to reduce their fire accident rate to 0 in some periods