【作者】 陈鹏；

【导师】 胡军华；

【作者基本信息】 中南大学 ， 管理科学与工程， 2013， 硕士

【摘要】 面对日益激烈的市场竞争环境,各个企业及组织所面临的决策问题越来越复杂。决策问题的复杂性加之决策个体认知的局限性,导致决策者无法进行精确地判断与决策。利用模糊数来表示决策信息是一种有效的解决办法。犹豫模糊数是模糊数的最新扩展形式,其主要特点是它的隶属度不是一个确定值或服从某种分布,而是几个可能的值,这更加贴近实际情境,且在处理群决策信息时有独特优势——可以有效避免集结算子导致的信息丢失,因此受到学者的广泛关注。犹豫模糊语言集是犹豫模糊集研究领域的一个分支,目前这方面研究较少,本文在犹豫模糊集与模糊语言集的研究基础上,对犹豫模糊语言及相应多准则(群)决策方法进行了研究,主要成果如下：(1)鉴于二元语义处理语言的优势,定义了二元语义犹豫模糊语言集,给出了相关定理与性质,基于此改进了产生语言评价值的Context-free grammar方法与转换函数EGH,以支持评价语言转化为二元语义犹豫模糊语言。同时提出了二元语义犹豫模糊语言集的记分函数与距离公式,并改进了PROMETHEEΠ方法。在此基础上,针对权重已知与未知分别提出相应的犹豫模糊语言多准则决策方法。(2)针对区间梯形二型模糊数能更准确模拟语言值的特性,定义了区间梯形二型犹豫模糊集(数),并从其几何角度出发,提出基于面积法的区间梯形二型犹豫模糊数可能度定义与差异度算法,同时提出了融合决策者偏好信息的权重确定模型。在此基础上,针对准则权重已知与未知两种情况分别提出基于区间梯形二型模糊数的犹豫模糊语言多准则决策方法。(3)将犹豫模糊语言拓展到多准则群决策环境中,针对之前学者在处理过程中忽略专家权重影响且对重复值只取其一的情况,利用证据推理方法对其进行了改进；同时考虑不同专家所熟知的领域不同,提出在不同准则下各专家主客观权重的确定方法。在此基础上,针对准则权重已知与未知两种情形分别提出相应的犹豫模糊语言多准则决策方法。图1幅,表18个,参考文献91篇更多还原

【Abstract】 With the rapid development of economy, economic situation changes ever fast and the competition between companies becomes more and more fierce, thus leading to increasingly complex decision making problems. The complexity of the decision making problems and the cognitive limitations of individuals hinder decision makers (DMs) making accurate judgments in the decision-making process. Employing fuzzy number, which was proposed by Zadeh, to represent the decision-making information is an effective solution. Hesitant fuzzy number is the latest extension form of fuzzy number, featuring in its membership, which is not a certain value or obeys a certain distribution, but several possible values. Hesitant fuzzy number is much closer to the real situation, and has unique advantage in dealing with group decision making problems, which can avoid loss of information caused by aggregation operators effectively, so it appeals to many scholars’attention.Hesitant fuzzy linguistic term set (HFLTS) is a branch of research on hesitant fuzzy set and currently there are few researches on it. In this paper, HFLTS and the corresponding fuzzy multi-criteria decision-making (MCDM) methods are studied based on hesitant fuzzy set and fuzzy linguistic term set. The main achievements are as follows:(1) Given the advantages of2-tuple in processing linguistic information, hesitant fuzzy2-tuple linguistic term set (HF2TLS) and its relevant theorems and properties are defined, based on which, Context-free grammar methods and conversion functions EGH are improved to help convert the linguistic evaluation information into hesitant fuzzy2-tuple linguistic term set. Meanwhile, the score function and distance formula of HF2TLS are put forward and PROMETHEEΠ method is improved. On the basis of these, we propose hesitant fuzzy linguistic MCDM methods, in which the criteria are both known and unknown.(2) Considering that trapezoidal interval type-2fuzzy number can describe the linguistic evaluation information more accurately, trapezoidal interval type-2hesitant fuzzy number is defined to handle hesitant fuzzy linguistic information. Then from the prospective of geometrical property, its possibility degree and distance formula are proposed, besides, a new model for determining criteria weights is defined, involving DMs’ preferences. On the basis of these, we propose hesitant fuzzy linguistic MCDM methods, in which the criteria are both known and unknown.(3) The HFLTS MCDM methods are proposed in group decision making environment based on evidence reasoning, which conquer the problems of ignoring the impact of DMs’ weights and repeating values. Meanwhile, a new model for determining DMs’ weights is defined, considering fully DMs’ knowledge and experience. On the basis of these, we propose hesitant fuzzy linguistic MCDM methods, in which the criteria are both known and unknown.1figure,18tables and91references更多还原