船舶安全熵概念的提出及应用

船舶航行安全很大程度上取决于操作者对船舶航行过程中各种信息的把握,信息掌握的多少可作为衡量船舶安全的一个标准。把船舶航行链分为5种信息处理过程,利用信息熵的原理,提出船舶安全熵的概念,并建立了数学模型。根据船舶航行信息过程的具体因素,建立船舶航行安全评价指标体系;利用熵权法确定5种信息单元及其安全因素权重,对船舶航行的安全熵进行计算,可根据评价模型对船舶安全进行有效评价。利用渤海湾船舶M进行验证,其结果发现船员素质安全熵最大。船舶安全熵方法是对船舶安全评估的一种新探索。     

浅析国内外重大危险源辨识中危险物质临界量的差异

通过分析国内外重大危险源辨识标准中危险物质临界量的差异,以汽油和液化石油气为例,分别选择池火灾和沸腾液体扩展蒸气云爆炸模型计算事故后果,并结合我国企业的安全管理状况,说明了修改我国《重大危险源辨识》中危险物质临界量的必要性。     

重大危险源动态监管与应急救援平台建设研究

由于缺乏相关的国家及行业标准规范,各地重大危险源监管与应急救援平台建设中主要存在如下问题:其一,只重视硬件建设,而忽视软件和信息建设;其二,重大危险源安全监管数据采集标准不统一,采用的相关软件和硬件技术不规范,难以在各级平台间实现数据共享。笔者参照重大危险源4级监管与应急救援体系,将国家科技攻关课题的相关研究成果同安全生产工作的实际需要有机地结合,提出重大危险源动态监管和事故应急救援平台建设方案。分析我国重大危险源监管及应急救援系统的建设现状及存在的主要问题;提出平台的设计原则、实现数据共享的主要思路;设计软件系统总体规划及各部分的功能。     

工艺过程危险有害因素辨识的研究

鉴于化工企业工艺过程的特殊性,笔者建议其危险、有害因素的辨识可以在直观经验分析法和系统安全分析法的基础上,结合危险和可操作性研究(HAZOP)的思想来进行辨识。笔者尝试性地给出了相关术语的说明、介绍辨识方法和操作程序,并分别从生产过程(包括化学反应、化工操作单元和物料输送)和工艺设备、装置角度,对其中的所包含的具体内容进行剖析,最后以电解过程举例说明,该方法可用于化工企业进行危险性因素辨识。     

基于模糊聚类的马氏链模型在交通事故预测中的应用

运用模糊聚类分析方法,借助于MATLAB软件,对中国31个地区交通事故的危害程度进行动态分类和综合评价。根据中国1999—2005年31个地区的交通事故4个指标和5个社会经济因素数据,把31个地区分为3类,即轻灾区、较重灾区、重灾区;其中浙江、福建与宁夏的交通事故危害严重,上海、北京与天津的交通状况正在好转,这与客观实际符合。根据模糊聚类的结果,求得交通事故状态转移概率矩阵,利用马氏链模型可预测未来31个地区的交通事故发展趋势。     

煤矿危险源辨识与多层评价分级研究

根据煤炭系统生产的复杂特点,提出了采用系统科学的方法,对煤炭生产进行系统的单元划分,采用安全系统科学辨识的方法对危险源进行辨识.在此基础上,提出利用层次分析法和模糊动态评判的方法对危险源进行初次分级,根据危险源的耦合特性,提出二次动态分级的方法,对初次分级进行调整,确定危险源科学、合理的级别,并提出了进一步现实性评价的建议.     

HUMAN BEHAVIOR AND RESPONSE TOWARDS STORM HAZARD IN WEST VANCOUVER,BRITISH COLUMBIA1

ABSTRACT: Human behavior and response towards storm hazard are examined in this study. One-hundred and twenty flood victims in West Vancouver, British Columbia, were administered a sentence completion test to ascertain how these victims behaved and responded to the flash flood of July 12, 1972. The sentence completion test consisted of 11 sentence stems. Seven of these dealt with the respondent's behavior toward the storm experience, and four with the issue of internal vs. external locus of control. Completions of the seven sentence stems at three time situations showed that West Vancouverites expressed the normal reactions of fear and anxiety before the storm; showed negative emotions during the storm; and displayed a feeling of mutual cooperation after the storm. Completions of the remaining four sentence stems revealed that West Vancouverites emphasized the importance of internal factors, viz., autonomy, drive and hard work, in controlling their lives. They recognized God as a benevolent - protective and almighty - powerful figure. Forty percent of the respondents did not believe in luck, while 8 percent regarded it as important in their lives. One might conclude that West Vancouverites could cope with natural hazardous threats better than those who rely on external factors in directing their lives.     

IMPACT OF FLOOD HAZARD ON RESIDENTIAL PROPERTY VALUES IN GALT (CAMBRIDGE), ONTARIO1

ABSTRACT: This study is an examination of the relationship between flooding and property values for an urban community in southwestern Ontario which has a lengthy history of flooding. As well, peoples’perceptions of the effect of flooding on property values is discussed. Analysis of sales prices and assessment data demonstrated no statistically significant differences in values of property for residences located in high- and low-risk areas either before or after a major flood in 1974. With both the high- and low-risk areas, sales prices after the 1974 flood were significantly higher than sales prices before the flood. The perceived property values followed a similar pattern. We concluded that differences in flood risk and flood experience did not adversely affect actual or perceived long-term property values.     

STOCHASTIC MODELING OF MONTHLY AND DAILY RAINFALL SEQUENCES1

ABSTRACT: A monthly model and two daily models (I and II) are presented for the purpose of generating monthly and daily rainfall sequences in the Quae Yai river basin in Thailand. Performance of the models are evaluated by comparing the statistical parameters of the generated sequences with those from historical data. For monthly generation, Thomas-Fiering model worked satisfactorily in spite of the monthly correlations being weak, if any. Daily Model I, which assumes no persistence between daily rainfall amounts within the wet spells, could not preserve some important parameters regardless of the simplicity in model construction. Application of multi-state transition probability matrix model gave good results, although the user has to modify some parameters looking at the performance of the model for each historical record.