基于Gologit模型的地铁事故严重程度影响因素研究  

Research on the Factors Affecting of Subway AccidentsSeverity Based on Gologit Model

作　　者：王甜甜[1] 李伟[1] WANG Tiantian;LI Wei(School of Management and Engineering,Capital University of Economics and Business,Beijing 100070,China)

机构地区：[1]首都经济贸易大学管理工程学院,北京100070

出　　处：《安全》2021年第9期58-63,70,共7页Safety & Security

基　　金：北京市教育委员会科研计划项目(KM202110038002);首都经济贸易大学北京市属高校基本科研业务费专项资金资助项目。

摘　　要：为明晰地铁事故严重程度的影响因素,保障地铁运营安全,本文基于某市2017-2018年间的1180起地铁事故数据,选取人员、人员行为、设备及环境4个影响要素中的16个因素为自变量,地铁事故严重程度为因变量,构建Gologit模型,进行参数估计、模型有效性检验及地铁事故模型边际效用分析,并从人员特性、人员行为特性、设备特性、环境特性4方面提出建议措施。研究表明:上下班高峰期、老幼群体、紧急制动、电梯故障、闸机故障及列车故障导致地铁发生重伤事故概率增加均在30%以上;模型有效性检验结果表明,Gologit模型可以放宽自变量的比例优势假设,从而使得模型精度更高。In order to clarify the influencing factors on subway accidents severity,based on the data of 1,180 subway accidents in a city from 2017 to 2018,in this paper,16 of the four influencing factors of personnel,personnel behavior,equipment and environment were selected as independent variables and subway accident severity as dependent variable to construct Gologit model and carry out parameter estimation,model validity test and marginal utility analysis of subway accident model.Then,measures were proposed from four aspects of personnel characteristics,personnel behavior characteristics,equipment characteristics,and environmental characteristics.The results showed that the probability of serious injury caused by rush hour,the elderly and the young groups,emergency braking,elevator failure,brake failure and train failure increased by more than 30%and the generalized ordered Logit model could relax the proportional advantage assumption and get higher precision.

关 键 词：地铁事故 严重程度 影响因素 Gologit模型 

分 类 号：X948[环境科学与工程—安全科学]