导　　师： 许若宁

学科专业： 070103

授予学位： 硕士

作　　者： ;

机构地区： 广州大学

摘　　要： 在复杂系统和决策环境中,决策者的知识通常是不完美的,所以无法清楚地表达其决策值以及属性权重,即决策者给出的决策信息往往是不明确的,很难直接排序决策.在求解不明确多目标问题的关键是:建立能够有效表述不明确信息的优化模型,在求解过程中正确把握不明确中的确定规律和因素,使得最终优化设计结构最大程度上满足决策者的真实意图.本文在此思想的指导下,对几种不同类型的多属性决策问题的建模和求解方法做了进一步的研究和探讨,其主要内容如下: 1/)阐述了多属性决策的发展过程,重点概括了多属性决策属性权重以及决策值不明确情况的国内外发展现状. 2/)针对只有部分属性权重信息且方案的属性值以梯形模糊数给出的多属性决策的情形,分别从梯形模糊数的中心和相似度角度,建立优化模型得出相关权重,并根据贴近度以及梯形模糊数的可能度公式进行排序. 3/)研究梯形模糊数互补互反一致性判断矩阵和排序向量之间的类似特征向量的关系,并建立排序向量优化模型,求得排序向量并根据期望算子对方案进行排序. 4/)对于属性权重完全未知且有主客观偏好的梯形模糊数决策问题,提出了方案在主客观偏好的α截集下距离的加权偏差的方法,并对偏差进行α求积分,基于主客观偏好加权偏差最大化思想,建立非线性优化模型,求解得到属性权重,由等级均值积分对方案进行择优. 5/)对本文提出的几种模糊决策问题进行总结,并指出今后需要继续研究和改进的课题. Under the complex Making Decision /(MD/) environments,the MD’s knowledge is usually not perfect and the person could not express clearly for the decision values and attribute weights,so MD’s decision-making information is undefined and it is difficult to give ranking.The key factors for solving multi-optimization with undefined decision values and attribute weights are:to effectively express undefined information and correctly grasp the certain regular and factors for those undefined information,to achievethe system optimization results meeting MD’s requirement and expectation.Based on the above thoughts,the paper further develops the models and methods of several different types of multi-attribute making decision.The main content of this paper includes the following aspects: 1. Narrating the background of multi-attribute decision making and introducing the development status of the present study on undefined decision values and attribute weights. 2. Aimming the part of the multi-attribute weights information and in the case of attribute value of program is given by trapezoidal fuzzy number ,respectively,from the perspective of trapezoidal fuzzy center and similarity,established optimization model and reachde appropriate weight, and ranking by the formula of closeness and the possible degree of fuzzy number. 3. Studing the relationship between reciprocal and comparis consistency matrix and ranking vector similar to the feature vector ,and established ranking vector optimization model , obtaining ranking vector and ranking program by expectations operator. 4. For the attribute weights are completely unknown and have the preferences of objective and subjective trapezoidal fuzzy decision making problems,proposed the weighted deviation methods of program underαcut set distant of objective and subjective preferences and quadrature onαof the deviation.Based on the maximize thinking of objective and subjective weighted deviation,and established nonlinear programming model and obtained attribute weights,ranking program by mean points level. 5. Finally, presenting this thesis’s summary and the direction for further discussion.

关 键 词： 多属性决策 模糊中心 相似度 可能度

分 类 号： [C934]

领　　域： [经济管理]