基于Copula理论的深圳市地面塌陷致灾因素的联合概率分析

Joint probability analysis of ground collapse disaster factors in Shenzhen City based on Copula theory

分析了深圳市近年来发生的地面塌陷事故数据,发现深圳市地面塌陷最主要的致灾因素是地下管网破损和地下施工不当,且管网破损地面塌陷数与施工不当地面塌陷数之间存在明显的相关性。选取了常见的几类边缘分布类型对管网破损地面塌陷数和施工不当地面塌陷数的分布规律进行拟合,其最优边缘分布类型分别为威布尔分布和截尾极值Ⅰ型分布。利用赤池信息准则(Akaike information criterion,AIC)和贝叶斯信息准则(Bayesian information criterion,BIC)确定了描述管网破损地面塌陷数和施工不当地面塌陷数之间相关性的最优Copula函数是Gaussian Copula函数,并以此建立了深圳市地面塌陷致灾因素的联合概率分布模型,进行了联合概率分析。研究表明:管网破损和施工不当这2个致灾因素同时出现得越频繁,造成地面塌陷的概率越大。

Based on the analysis of the ground collapse accidents in Shenzhen City in recent years,it is found that the ground collapse in Shenzhen City is mainly caused by the damage of underground pipe and the improper construction of underground engineering.There is an obvious correlation between the number of ground collapse caused by the damage of pipe and the number of ground collapse caused by improper construction.Several common types of marginal distribution are selected to fit the distribution laws of the number of ground collapse caused by the damage of pipe and the number of ground collapse caused by improper construction.The optimal marginal distribution types are Weibull distribution and truncated extreme value typeⅠdistribution,respectively.Based on the AIC criterion and BIC criterion,Gaussian copula function is determined as the optimal copula function to describe the correlation between the number of ground collapse caused by the damage of pipe and the number of ground collapse caused by improper construction.Based on this,a joint probability distribution model of disaster causing factors of ground collapse in Shenzhen City is established,and the joint probability analysis is carried out.The research shows that the more frequent the two disaster factors,pipeline damage and improper construction,appear at the same time,the greater the probability of ground collapse.