AHP决策方案集合的构造方法与初选方法研究

发布时间：2018-04-29 11:11

本文选题：层次分析 + 方案集合　；参考：《吉林大学》2006年硕士论文

【摘要】： 层次分析法是一种应用广泛的决策方法。现有AHP分析方法将目标的层次分析和备选方案的构造视为两个彼此不相关的独立过程,割裂了它们之间的系统联系。另外,AHP对方案的评价主要依据的是方案产生的效益,对方案的投入产出效率或相对效率并没有予以考虑。这样最终可能会影响方案排序的科学合理性,从而导致决策失误。为克服上述缺陷,论文利用功能结构树原理和正交试验技术给出了备选方案集合的构造方法。该方法能够确保备选方案分布的均衡性,有利于实现方案排序与方案选择的科学性。在此基础上,论文利用DEA模型对备选方案集合中的各个方案进行了相对效率评价,并基于爬山原理及循序渐进原则利用模拟退火方法对相对效率较差的方案进行了改进,之后将改进后的方案与原来的有效方案一起再次进行相对效率评价,并将该次评价得出的相对有效方案定义为AHP决策方案集合的优选子集。由于该优选子集不仅实现了对原有方案的合理扩充,而且其中所有的方案都是具有pareto最优性的相对有效方案,因此可以保证后续AHP分析的对象即待评价方案的数量和质量。最后,结合一个新产品开发的实例,对前文提出的方法体系进行了应用研究。应用研究结果表明,论文所提出的AHP备选方案集合构造方法和初选方法具有较好的实际应用可操作性。
[Abstract]:Analytic hierarchy process (AHP) is a widely used decision-making method. The existing AHP analysis methods treat the hierarchical analysis and alternative construction of the target as two independent processes which are independent of each other and separate the systematic relationship between them. In addition, the evaluation of the scheme by AHP is mainly based on the benefit of the scheme, and the input-output efficiency or relative efficiency of the scheme is not considered. This may ultimately affect the scientific reasonableness of program sequencing and lead to decision-making errors. In order to overcome the above defects, the method of constructing alternative scheme set is presented by using the principle of functional structure tree and orthogonal test technique. This method can ensure the equilibrium of the distribution of alternative schemes, and is beneficial to the realization of the scientific arrangement and selection of alternatives. On this basis, DEA model is used to evaluate the relative efficiency of each scheme in the set of alternative schemes, and based on the principle of mountain climbing and step by step principle, simulated annealing method is used to improve the scheme with low relative efficiency. Then the improved scheme and the original effective scheme are evaluated again, and the relative effective scheme is defined as the optimal subset of the AHP decision plan set. Since the optimal subset not only extends the original scheme reasonably, but also all of the schemes are relatively effective schemes with pareto optimality, it can guarantee the quantity and quality of the following AHP analysis object, that is, the scheme to be evaluated. Finally, an example of new product development is given to study the application of the proposed method system. The application results show that the set construction method and primary selection method of AHP alternatives proposed in this paper are feasible in practical application.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2006
【分类号】：C934;F224

【引证文献】