基于免疫算法及蚁群算法的应急物资储备库选址研究

发布时间：2018-01-11 06:29

本文关键词：基于免疫算法及蚁群算法的应急物资储备库选址研究　出处：《吉林大学》2017年硕士论文　论文类型：学位论文

【摘要】：近年来,世界范围内气候条件变化明显,导致各类突发事件频发,给各国的经济和社会带来了严重灾难,使人们蒙受生命和财产的双重损失。此外,随着经济社会的不断发展,人们的生产、生活方式发生了巨大的改变,导致人口集聚化、产业集中化等社会现象的出现,并由此加剧了突发事件的危害性。突发灾害事件给经济和社会带来的严重影响,对应急救援资源的管理与配送提出了严峻挑战,而科学合理的应急物资选址是适时、快速地提出应急救援方案的前提,也是应急物资储备库救援资源效益最大化的关键环节。论文通过对研究背景的分析,发现我国在应急救援领域存在的问题,并在总结国内外应急救援研究现状的基础上,明确本文的研究方向;同时,通过对应急物资储备库选址的相关理论及常用选址模型的梳理,为后文的研究奠定理论基础。根据选址理论研究现状和应急救援管理存在的问题,在结合应急物资储备库选址特征的基础上,本文提出受灾点需求权重问题、选址问题以及储备库分级问题,并分别构建模型,给出相应的求解方法。首先,本文对受灾点需求权重预测进行研究,利用主成分分析法,使用SPSS求解需求权重影响因子的数值,结合需求权重特征,构建需求权重预测模型,并以吉林省为实例,通过编程求解,得到各地市(区)的需求权重。其次,结合应急救援呈现独特的救灾时效性需求,以及应急救援过程中涉及的动态救援物资配送问题,从需求覆盖和最优路径两个方面建立目标函数,分别构建基于需求距离最小模型和基于最优路径模型。在模型求解过程中,运用免疫算法解决应急物资储备库选址问题中的需求覆盖问题,得到初级选址方案;接着,运用蚁群算法解决最优路径问题,在初级选址方案的基础上,进行储备库分级,确定最终选址方案。最后,以吉林省应急物资储备库选址为算例,通过Matlab编程,对上述选址模型进行了有效性和实用性验证,并为吉林省应急储备库分级建设工作提供了理论依据。
[Abstract]:In recent years, the change of climate conditions in the world is obvious, which leads to frequent occurrence of all kinds of unexpected events, brings serious disasters to the economy and society of various countries, and makes people suffer double losses of life and property. With the development of economy and society, people's production and life style have changed greatly, resulting in the emergence of social phenomena such as population agglomeration, industrial concentration and so on. As a result, it intensifies the harmfulness of emergency events, which brings serious impact to economy and society, and poses a severe challenge to the management and distribution of emergency rescue resources. And the scientific and reasonable location of emergency materials is a timely, rapid emergency rescue program premise, but also the key link to maximize the efficiency of emergency material storage and rescue resources. The paper through the analysis of the research background. Find out the existing problems in the field of emergency rescue in China, and on the basis of summarizing the current situation of emergency rescue research at home and abroad, make clear the research direction of this paper; At the same time, by combing the relevant theories and common location models of emergency material storage, the paper lays a theoretical foundation for the later research. According to the current situation of location theory and the problems of emergency rescue management. Based on the characteristics of emergency material storage location, this paper puts forward the problem of the weight of disaster point demand, the problem of location selection and the classification of the reserve, and builds the model respectively, and gives the corresponding solution method. In this paper, the demand weight prediction of the disaster point is studied. Using principal component analysis, SPSS is used to solve the value of demand weight influence factor, and combining the demand weight characteristics, the demand weight prediction model is constructed. And take Jilin Province as an example, through programming solution, get the weight of the demand of each prefecture (district). Secondly, combined with emergency rescue presents a unique demand for the timeliness of disaster relief. As well as the dynamic relief material delivery problem involved in the emergency rescue process, the objective function is established from two aspects: demand coverage and optimal path. In the process of solving the model, the immune algorithm is used to solve the demand coverage problem in the emergency material reserve location problem, and the primary location scheme is obtained. Then, the ant colony algorithm is used to solve the optimal path problem. On the basis of the primary site selection scheme, the reserve bank classification is carried out to determine the final location scheme. Finally, the emergency material reserve location in Jilin Province is taken as an example. Through Matlab programming, the validity and practicability of the above model are verified, and the theoretical basis is provided for the graded construction of emergency reserve in Jilin Province.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：D63;TP18

【相似文献】