基于不确定信息的救灾物资分配决策模型研究

发布时间：2018-02-22 02:38

本文关键词： 需求分级证据理论救灾物资分配　出处：《东北大学》2014年硕士论文　论文类型：学位论文

【摘要】：大型自然灾害发生后,会伴随一段“灰箱”时间,这时候受灾点受到破坏的程度、道路交通情况对救灾物资的需求信息都是不明确的。决策者需要在不完全信息的基础上对有限的救灾物资就行分配,以达到救灾的目的。我国是个自然灾害频发的国家,需要对灾后救灾物资分配这个问题进行的深入地研究,以便减少人员伤亡和财产损失,达到救灾减灾的目的。本文综合运用应急管理理论、证据理论来解决不完全信息下救灾物资分配的问题。本文首先根据救灾物资的用途将救灾物资分为四类：救生类物资、生活类物资、取暖御寒类物资以及医药类物资。利用层次分析法,采用专家调查问卷的形式对同一准则下不同种类的物资进行重要性比较,得出判断矩阵进行一致性检验,从而求得四种救灾物资的权重。然后结合自然灾害发生后的“灰箱”时间,利用不完全信息,通过决策者对受灾点关于救灾物资需求等级的语言值描述；利用D-S证据理论,先求各个受灾点关于优先等级的基本信任函数,再利用证据理论的合成规则,对各个受灾点的基本信任函数进行合成,从而得出信函数和似然函数的区间,然后比较其优先等级,确定各个受灾点关于救灾物资的优先度,利用模糊信息处理的方法将不确定的受灾点的效用系数变成确定的数值。以汶川地震为例,进行算例仿真分析。最后考虑受灾点关于救灾物资需求优先性的情况,对受灾点的优先性进行数值量化得到受灾点的效用系数；建立双层模型,以应急时间最短和受灾点的救灾物资效用最大建立救灾物资分配函数；使用遗传算法进行求解。以汶川地震为例,进行算例仿真分析。
[Abstract]:After a large-scale natural disaster occurs, it will be accompanied by a period of "gray box" time, which is the extent to which the disaster site is damaged. The road traffic situation has unclear information on the demand for disaster relief materials. Decision makers need to distribute the limited relief materials on the basis of incomplete information in order to achieve the purpose of disaster relief. China is a country with frequent natural disasters. In order to reduce casualties and property losses and achieve the purpose of disaster relief and disaster reduction, this paper comprehensively applies the theory of emergency management. The theory of evidence is used to solve the problem of the distribution of disaster relief materials under incomplete information. Firstly, according to the use of disaster relief materials, this paper divides the relief materials into four categories: life-saving materials, life-saving materials, living materials, By means of analytic hierarchy process (AHP), the importance of different kinds of materials under the same criterion was compared by using the questionnaire of experts, and the consistency of the judgment matrix was verified. The weight of the four kinds of relief materials is obtained. Then combining the "grey box" time after the natural disaster, using incomplete information, the decision makers describe the linguistic value of the disaster relief material demand level, and use the D-S evidence theory. First, the basic trust function of each disaster point about priority level is obtained, and then the basic trust function of each disaster point is synthesized by using the composite rule of evidence theory, and the interval of information function and likelihood function is obtained. Then compare their priority levels, determine the priority of disaster relief materials in each disaster site, use the method of fuzzy information processing to convert the utility coefficient of the uncertain disaster site into a definite value. Take the Wenchuan earthquake as an example. Finally, considering the priority of disaster relief material demand, the utility coefficient of disaster point is obtained by numerical quantification of the priority of disaster point, and the two-layer model is established. Based on the shortest emergency time and the maximum utility of disaster relief materials, the distribution function of disaster relief materials is established, and the genetic algorithm is used to solve the problem. The Wenchuan earthquake is taken as an example to carry out the simulation analysis.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：D632.5;F259.21

【相似文献】