基于Apriori算法的煤矿安全事故分析

为分析煤矿事故报告中的危险致因因素,统计分析了2018—2022年全国煤矿事故报告数据,采用Apriori关联规则算法,并利用Gephi进行关联规则可视化,探究各个致因之间的复杂关系。首先对数据进行预处理,计算词频-逆向文件频率(Term Frequency-Inverse Document Frequency, TF-IDF),提取了78个煤矿事故致因因素,其中人因层包括31个因素,设备层包括9个因素,管理层包括31个因素,环境层包括7个因素;然后,经过关联规则挖掘算法,得到了585条关联规则,绘制了其支持度、置信度和提升度的散点图;最后,根据Gephi生成的事故致因复杂网络图,分别分析了高支持度、高置信度和高提升度关联规则致因因素。结果表明:基于Apriori算法的煤矿事故致因分析,得到了人因层、管理层、设备层和环境层4个方面的关键致因因素;对煤矿关键致因因素进行直观、多视图的展现,有助于提高煤矿安全管理水平。

采用STAMP-24Model的多组织事故分析

安全生产事故往往由多组织交互、多因素耦合造成,事故原因涉及多个组织。为预防和遏制多组织生产安全事故的发生,基于系统理论事故建模与过程模型(Systems-Theory Accident Modeling and Process,STAMP)、24Model,构建一种用于多组织事故分析的方法,并以青岛石油爆炸事故为例进行事故原因分析。结果显示:STAMP-24Model可以分组织,分层次且有效、全面、详细地分析涉及多个组织的事故原因,探究多组织之间的交互关系;对事故进行动态演化分析,可得到各组织不安全动作耦合关系与形成的事故失效链及管控失效路径,进而为预防多组织事故提供思路和参考。

高校实验室PAR事故致因模型构建及应用

高校实验室是集聚和培养优秀专业人才、开展高水平科学研究的重要基地,但目前仍面临较多安全管理难题。现以“中科院某研究所高压反应釜爆炸事故”为例,采用PAR事故致因模型开展高校实验室安全管理研究,以安全感知—安全分析—安全应答(PAR)为节点,对实验室建设与管理处、中心实验室、实验室操作主体3个层面进行分析。研究表明,PAR模型能充分发挥安全信息的交流作用,从宏观、中观、微观3个层面进行事故致因分析,为完善实验室事故分析体系等提供有效支撑。本研究对高校实验室安全管理的安全信息分析和风险防控提供参考,对实验室安全管理有一定理论指导意义。

基于HFACS的道路运输事故致因分析与分类系统研究

为了深入研究道路运输事故的致因因素,挖掘出关键致因规则,在人因分析和分类系统(HFACS)的基础上,构建道路运输事故致因分析与分类系统(RTAACS),并根据该系统框架对收集到的事故进行分析。首先将RTAACS框架分为监管部门、道路运输企业、驾驶人3个层级,将事故致因因素纳入该系统框架进行频率统计,通过R语言改进卡方检验、让步比分析得出致因因素之间的关联规则,最后根据ANP网络层次分析法计算各项致因因素的权重,并确定关键致因规则。结果表明:工作流程与不充分的企业监督、工具技术与习惯性违规、资源管理与工具技术、资源管理与习惯性违规、未纠正的已知错误与习惯性违规为道路运输事故的关键致因规则。研究构建的系统框架以及提出的分析方法为道路运输业的安全预防提供一定的借鉴意义。

Comparative study of HFACS and the 24Model accident causation models

A comparative study is conducted to compare the theory and application effect of two accident causation models, the human factors analysis and classification system（HFACS） and the accident causation ＂2-4＂ model（24 Model）, as well as to provide a reference for safety researchers and accident investigators to select an appropriate accident analysis method. The two models are compared in terms of their theoretical foundations, cause classifications, accident analysis processes, application ranges, and accident prevention strategies. A coal and gas outburst accident is then analyzed using both models, and the application results are compared. This study shows that both the 24 Model and HFACS have strong theoretical foundations, and they can each be applied in various domains. In addition, the cause classification in HFACS is more practical, and its accident analysis process is more convenient. On the other hand, the 24 Model includes external factors, which makes the cause analysis more systematic and comprehensive. Moreover, the 24 Model puts forward more corresponding measures to prevent accidents.

基于AISM的高校燃烧与爆炸实验室事故关键致因分析

高校燃烧与爆炸实验室涉及燃烧、爆炸、高温、高压等危险因素,具有较高的事故风险。结合三类危险源理论辨识出高校燃烧与爆炸实验室爆炸、火灾、中毒、烫伤4种主要事故类型的25个事故致因因素;进而运用对抗解释结构模型(AISM)对因素间相关关系进行分析,得到高校燃烧与爆炸实验室事故关键致因。以某高校燃烧与爆炸实验室为例,通过对比AISM模型的UP型和DOWN型有向拓扑层级图,确定出该校燃烧与爆炸实验室事故的5个根源性关键致因因素、5个敏感性关键致因因素、3组互为因果关系的关键致因因素。研究结果为制定更精准的预防事故措施提供了依据,为提高现代高校燃烧与爆炸实验室安全管理效率提供了一种新的思路。

Analysis of a Lightning Strike Accident of a Rural House and Defensive Measures

A lightning strike accident of a residential house in the rural areas of Neijiang is analyzed through on-site investigation,and the causes are discussed. Defensive measures of lightning disasters are proposed to provide reference for the prevention of lightning disasters in the construction of new countryside.

基于DEMATEL-TAISM的钢厂风险因素分析与管理策略优化

钢铁制造企业具有复杂化、系统化的特点,在这种发展趋势下潜在风险因素日益增多,钢铁制造企业高质量发展和可持续发展的道路充满挑战。由于风险因素之间的耦合关系,导致日常生产中面临难以抓主次、厘清因果关系的问题,实践中缺少系统的指导,安全管理过程中难免发生疏漏。基于此,研究采用一种新的决策实验室法和综合对抗解释结构模型(Decision-Making Trial and Evaluation Laboratory-Total Adversarial Interpretive Structural Model)相结合的方式,建立了钢厂安全影响因素体系和模型。结果表明,与AISM相比,新模型加强了中心度与层级拓扑图的联系,可以更加直观地展示钢厂事故致因之间的作用路径,分析系统中因素间的影响路径、程度和逻辑关系。该研究可对钢厂日常安全管理水平提高提供理论参考。

Cause Analysis of Wuhan Tianheng Building Pile Accident

The geological condition and the original structure feature and foundation design of Wuhan Tianheng building are described. The accident appearance of pile foundation in the construction execution of work is illustrated. The generating source of this pile foundation accident is analyzed in great details.

基于扎根理论和ISM-MICMAC的石油化工事故致因因素研究

准确识别石油化工事故致因因素,分析各因素间的内在逻辑关系,有助于遏制石油化工事故的发生。通过研究2017—2022年发生的41起石油化工事故,使用扎根理论总结出16个石油化工事故致因因素,使用ISM模型法和MICMAC分析法对导致石油化工事故发生的16个致因因素进行了研究,得到“石油化工事故致因因素”解释结构模型。此模型将16个事故致因因素分为6个层级和4个象限,该模型揭示了化工事故的根本原因,提供了一种新型化工事故的致因分析方法,可适用于大多数化工生产事故的分析。

ANALYSISOFOPERATIONALCAUSESOFAIRPLANETAKEOFFACCIDENTS

分析了飞机起飞阶段由于操作失误／偏差所引起的飞行事故原因，分别从离地前、离地附近和开始爬升三个阶段。

基于SNA的危化品道路运输事故致因研究

为明确危化品道路运输事故致因因素及其相互关系,首先,基于事故致因“2-4”模型分析2019—2023年间发生的155起危化品道路运输事故案例,然后,利用社会网络分析(SNA)方法构建危化品道路运输事故致因网络,并进行网络中心性、核心—边缘结构及凝聚子群分析。结果表明:在得到的30项事故致因因素中,14项致因因素位于致因网络的核心区域,其中,操作不当、特殊路段、超速、注意力不集中、安全意识淡薄等5项致因因素对事故发生有着较强的影响力;致因网络中共形成5个凝聚子群,各凝聚子群间以及凝聚子群内各致因因素间关系紧密。研究结果可为提高危化品道路运输安全管理水平提供理论依据。

基于STAMP-DEMATEL的既有建筑改造安全事故致因研究

城市更新背景下既有建筑改造将是大势所趋,而既有建筑改造安全事故逐年增多,危害到国家、企业和个人生命财产安全。厘清既有建筑改造安全事故的致因并针对性防范,对降低安全事故的发生频率具有重要现实意义。基于STAMP模型,分析并构建既有建筑改造项目安全事故的分层控制结构,并以此为框架识别和筛选事故致因因素;运用DEMATEL算法分析安全事故系统各层级致因因素之间的关联关系,辨识关键性致因因素;以泉州XJ酒店改造工程坍塌事故开展案例研究,验证STAMP-DEMATEL模型的适用性。结果表明:既有建筑改造项目分层控制结构包括物理层、现场基础层、运营层、政府及社会层等4个递进层级;政府社会层如五方责任主体监督监管失察、既有建筑改造施工标准体系不健全、企业行为规范缺失等原因因素属于导致既有建筑改造事故发生的高效力因素;物理层、基层如钢柱板件严重变形、应急预案缺失、改造技术方案安全性缺失、施工机械选择不当等结果因素属于诱发事故的直接因素;基于STAMP和DEMATEL的分析结果与案例事故报告高度贴合。所构建的安全事故致因模型可用于指导既有建筑改造施工安全管理。

行为追溯安全培训法(ACT)理论构建与应用

为减少作业人员不安全行为,提高现场作业安全绩效,以案例教学和事故致因“2-4”模型(24Model)为指导,结合策划、实施、检查、处理(PDCA)循环理论,构建以事故-原因-培训(ACT)为关系链的行为追溯安全培训法,亦称ACT安全培训法。该方法将事故案例贯穿培训全程,发挥案例教学优势,利用事故致因24Model追溯导致事故的组织层面和个人层面的不安全因素,基于分析结果制定针对性的培训方案,并以有限空间作业课程培训体系设计实施为例,验证ACT安全培训法的实际应用效果。结果表明:在有限空间作业培训课程体系实施过程中,参训人员数量及考核成绩均呈整体上升趋势,受训人员参与积极性及学习效果都得到有效提高,验证了ACT安全培训法具有良好的实用效果。

Cause analysis of expressway traffic accidents based on accident statistics of Hunan provincial expressway in 2012-2018

Based on the 2012–2018 accident statistics on the Hunan Provincial Expressway in China,this paper analyses the proportion to which various factors contribute to accidents,as well as the trends over time.From 2012 to 2018,the total number of general procedure handling accidents(i.e.major accidents)increased first and then decreased,while the number of accidents and deaths per 1000 kilometres decreased year by year.The total number of simple procedure handling accidents(i.e.minor accidents)increased year by year,but the number of accidents per 1000 kilometres remained relatively stable.Speeding,fatigued driving,improper driving operation,failure to maintain a safe distance from the preceding vehicle and pedestrians entering the expressway are the main causes of general procedure handling accidents.Failure to maintain a safe distance from the preceding vehicle as required and improper driving operation accounted for approximately 2/3 of the simple procedure handling accidents.In view of the accidents caused by these factors,how to use law enforcement,education,science and technology and facilities construction to further reduce the occurrence of expressway traffic accidents is discussed.

Fatality from Road Traffic Accident in Guinea: A Retrospective Descriptive Analysis

Objectives: Causes and risk factors that result in fatal road traffic accident have not been described at the national level in Guinea yet. The goal of this study is to explore the causes and risk factors related to fatal road traffic accident, identified most vulnerable road users, and inform the road traffic prevention policy in Guinea. Methods: We made a retrospective descriptive analysis based on national fatal road traffic accident data from the Department of Health Information at the Guinean Ministry of Health for year 2011. Results: In 2011, road traffic accident was responsible for an aggregate number of 1655 deaths with an overall death rate of 15.3 per 100,000 population. Male experienced more than twice the risk of death from road traffic accidents (21.9 deaths per 100,000 population) compared with female (9.0 deaths per 100,000 population). While taking the population as a whole, the highest death rate was found among the middle aged in 35 - 49 age group accounting for (29.7 deaths per 100,000 population), followed successively by young adults age group 25 - 34 years (24.6 deaths per 100,000 population), and the middle aged in 50 - 64 age group (22.9 deaths per 100,000 population). Principally, occupants, motorcyclists and pedestrians sustained considerable burden of deaths respectively (9.2;2.9;2.2 per 100,000 population). In re-gional setting, the highest death rate was found in Upper Guinea (19.5 per 100,000 population), followed by Forest Guinea (18.7 per 100,000 population) and Middle Guinea (16.8 per 100,000 population). A large proportion of male was killed as motorcyclist than female while high per-centage of female died as occupant than male for all age group. The regional distribution showed that when a remarkable number of occupant death were observed in Upper and Forest Guinea, more people died as pedestrian and pedal cyclist in Conakry. Conclusions: This study demonstrated that most of the deaths were among occupants, motorcyclists and pedestrians, and the productive workforce aged 25 - 49 years. It was found that majority of the deaths happened in Upper Guinea followed by Forest Guinea. Improvement of roads design, strict enforcement of road safety laws and raising the awareness of general public about the causes and risks factors of road traffic accident through various channels are highly required which will promote economic growth in the local communities and then help people escape the poverty trap.

基于改进Apriori算法的塔吊事故致因动态关联网络分析

为研究塔吊事故致因及其相互作用,采用人因分析和分类系统(HFACS)模型、关联规则挖掘和复杂网络理论相结合的方法,对塔吊事故致因进行定性分类和定量分析。首先,收集241份塔吊事故调查报告,并根据塔吊事故特点调整HFACS模型框架,以此确定46个影响塔吊事故发生的因素,再运用改进的Apriori算法挖掘事故致因之间的关联性;然后,结合关联规则挖掘结果构建塔吊事故致因动态关联网络,并综合探究塔吊事故的关键影响因素。结果表明:人员岗位与能力不匹配、监管人员未按要求进行安全检查、部分岗位人员配备不齐等是造成塔吊事故发生的关键因素。改进后的方法实现了人为因素的量化分析,可为预防塔吊事故的发生提供决策依据。

基于改进Apriori算法的铁路交通作业事故致因关联规则研究

铁路运输系统复杂,事故致因因素众多且有一定的关联关系,分析事故致因及关联特性,对于事故预防具有重要的意义。按照人因层、设备层、环境层、管理层4个层级对事故因素进行分层分析,整合相同或相似的条件,编写Python程序对近期某3年所有事故调查报告进行致因因素抓取,首先将致因因素拆解为相关的致因关键字,并利用正则表达式在指定段落中进行抓取,然后通过关键字抓取获得所有事故的致因因素,最终提取了90个铁路事故致因因素。在此基础上,基于压缩和哈希表技术对传统Apriori算法进行改进,以此对铁路交通作业事故的致因进行关联规则挖掘,找到159条关联规则,通过分析事故致因高支持度、高置信度等特征,发现关键因素、关键致因关联、事故致因规律,有助于提升事故预防和安全管理水平。