基于案例推理的城市洪涝灾害应急方案生成方法研究
发布时间：2018-08-21 12:00
【摘要】：由于全球气候异常、生态环境破坏、人类活动等原因,洪涝灾害成为人类面临的重大威胁。随着我国城市化进程不断加深,城市洪涝致灾模式已经并将继续发生变化,同时,现代城市面对暴雨洪水显得更为脆弱。因此,在城市洪涝灾害发生之初如何选择适当的应对方案并及时加以实施,是有效减轻灾害损失的重要途径。对城市洪涝灾害应急方案生成方法的相关问题进行研究是城市安全领域的热点问题,对我国以后的城市洪涝灾害应急决策具有很强的实践指导作用。缺乏系统完善的应急领域知识、决策时间短、决策信息缺失等往往是城市洪涝灾害应急决策者常面临的难题。应急预案一直被视为是应对城市洪涝灾害的有力武器,但应急预案具有获取途径困难、实际操作性欠佳和修改周期较长等问题。因此本文将案例推理技术引入城市洪涝灾害的应急决策领域,用以往的应急案例实施方案进行修正后代替应急预案,快速生成方案。案例推理方法不仅可以实现救援措施的具体化、解决应急决策信息部分缺失问题,而且该案例推理系统能不断更新加入新案例,删除无用旧案例,更新完善进度比应急预案快速有效许多。本文以案例推理的四个环节为主线,首先在对城市洪灾形成原理进行深入剖析,概括洪涝灾害影响因素的基础上,选取框架表示法来对应急案例的问题域、解决方案域和应急效果域三大方面内容进行表示,并构建城市洪涝灾害的案例属性体系;然后针对城市洪涝灾害存在的属性值缺失、属性值不确定和属性间复杂关系对计算案例相似度所造成的影响问题,给出一种新的基于归纳索引法和RIMER方法的综合索引技术,并详细阐述了其在城市洪涝灾害中的应用求解过程,进一步地,在案例检索过程中还考虑了应急方案的实施效果,运用证据推理方法进行两次证据合成来进行效果评估,使检索结果更具合理性和准确性;接着研究了案例推理方法中案例修正及案例学习等在应急方案生成过程中的实现方式。最后给出了一种基于CBR的城市洪涝灾害应急方案生成方法基本框架。
[Abstract]:As a result of global climate anomaly, ecological environment damage, human activities and other reasons, flood disaster has become a major threat to mankind. With the deepening of urbanization in China, urban flood disaster pattern has changed and will continue to change. Meanwhile, modern cities are more vulnerable to torrential rain and flood. Therefore, it is an important way to reduce the loss of urban flood disaster by choosing the appropriate plan and implementing it in time at the beginning of the flood disaster. It is a hot issue in the field of urban safety to study the generation method of urban flood disaster emergency plan, which has a strong practical guidance role in the future urban flood disaster emergency decision-making in China. Lack of systematic knowledge of emergency response field, short decision-making time and lack of decision-making information are often difficult problems faced by urban flood emergency decision makers. The emergency plan has been regarded as a powerful weapon to deal with urban flood disaster, but the emergency plan has many problems, such as difficult way to obtain, poor practical operation and long period of modification. In this paper, the case-based reasoning technology is introduced into the emergency decision-making field of urban flood disaster, and the former emergency case implementation scheme is modified to replace the emergency plan, and the scheme is quickly generated. Case-based reasoning (CBR) can not only realize the embodiment of rescue measures and solve the problem of missing part of emergency decision information, but also can update and add new cases and delete useless old cases. The update and improvement schedule is much faster and more effective than the emergency plan. In this paper, the four links of case-based reasoning are taken as the main line. Firstly, on the basis of deeply analyzing the formation principle of urban flood disaster and summarizing the influencing factors of flood disaster, the frame representation method is selected to solve the problem domain of emergency cases. The solution domain and emergency effect area are represented, and the case attribute system of urban flood disaster is constructed, and then the attribute value of urban flood disaster is missing. This paper presents a new comprehensive indexing technique based on inductive index method and RIMER method. Furthermore, in the process of case retrieval, the effect of emergency plan is considered, and the effect is evaluated by using the method of evidence reasoning twice. The retrieval results are more reasonable and accurate. Then, the implementation of case correction and case learning in the process of emergency scheme generation is studied. Finally, a basic framework of urban flood emergency scheme generation based on CBR is presented.
【学位授予单位】：武汉理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU998.4
[Abstract]:As a result of global climate anomaly, ecological environment damage, human activities and other reasons, flood disaster has become a major threat to mankind. With the deepening of urbanization in China, urban flood disaster pattern has changed and will continue to change. Meanwhile, modern cities are more vulnerable to torrential rain and flood. Therefore, it is an important way to reduce the loss of urban flood disaster by choosing the appropriate plan and implementing it in time at the beginning of the flood disaster. It is a hot issue in the field of urban safety to study the generation method of urban flood disaster emergency plan, which has a strong practical guidance role in the future urban flood disaster emergency decision-making in China. Lack of systematic knowledge of emergency response field, short decision-making time and lack of decision-making information are often difficult problems faced by urban flood emergency decision makers. The emergency plan has been regarded as a powerful weapon to deal with urban flood disaster, but the emergency plan has many problems, such as difficult way to obtain, poor practical operation and long period of modification. In this paper, the case-based reasoning technology is introduced into the emergency decision-making field of urban flood disaster, and the former emergency case implementation scheme is modified to replace the emergency plan, and the scheme is quickly generated. Case-based reasoning (CBR) can not only realize the embodiment of rescue measures and solve the problem of missing part of emergency decision information, but also can update and add new cases and delete useless old cases. The update and improvement schedule is much faster and more effective than the emergency plan. In this paper, the four links of case-based reasoning are taken as the main line. Firstly, on the basis of deeply analyzing the formation principle of urban flood disaster and summarizing the influencing factors of flood disaster, the frame representation method is selected to solve the problem domain of emergency cases. The solution domain and emergency effect area are represented, and the case attribute system of urban flood disaster is constructed, and then the attribute value of urban flood disaster is missing. This paper presents a new comprehensive indexing technique based on inductive index method and RIMER method. Furthermore, in the process of case retrieval, the effect of emergency plan is considered, and the effect is evaluated by using the method of evidence reasoning twice. The retrieval results are more reasonable and accurate. Then, the implementation of case correction and case learning in the process of emergency scheme generation is studied. Finally, a basic framework of urban flood emergency scheme generation based on CBR is presented.
【学位授予单位】：武汉理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU998.4
【参考文献】
相关期刊论文 前10条
1 郑晶;王应明;叶歆;;复杂数据情形下的应急案例相似度测算方法[J];中国安全科学学报;2014年07期
2 李永海;樊治平;袁媛;;考虑应急方案实施效果的突发事件应急方案生成方法[J];控制与决策;2014年02期
3 王U，
本文编号：2195651
本文链接：https://www.wllwen.com/jingjilunwen/jianzhujingjilunwen/2195651.html
