A PSO-based group decision making model with multiplicative reciprocal matrices under flexibility
A PSO-based group decision making model with multiplicative reciprocal matrices under flexibility
复制标题
基于PSO的柔性乘法倒数矩阵群决策模型
DOI：
10.1007/s00500-018-3641-8
复制
发表时间：
2019
期刊：
影响因子：
4.1
通讯作者：
Yu Qin
中科院分区：
文献类型：
--
作者：
Liu Fang;Wu Yuhao;Zhang Jiawei;Yu Qin
使用层次分析法（AHP）的方法，邀请多位专家对一组有限的备选方案进行两两比较，并在群体决策中找到解决方案。在层次分析法框架下，允许给定的乘性互惠偏好关系与一致矩阵有一定程度的偏离。在本研究中，假设专家在群体决策问题中的初始位置由一组替代方案上的乘法倒数矩阵来表征。为每位专家提供灵活性程度，以便她/他的职位可以适时修改。信息颗粒是根据两种不同的方法构建的。为了在群体决策中建立共识，建立了一种新颖的目标函数来衡量多位专家内部的共识水平并量化群体一致性水平。基于改进的粒子群优化（PSO）算法，通过最小化目标函数来建模群体决策中的动态迭代共识过程。报告的数值结果是为了说明群体决策中提出的共识模型。通过比较现有的基于 PSO 方法的群体决策模型，揭示了其新颖性。所提出的模型可用于处理社交网络信息的大规模群体决策问题。
Using the method of the analytic hierarchy process (AHP), multiple experts are invited to pairwise compare a finite set of alternatives and find a solution in group decision making. Under the AHP framework, the given multiplicative reciprocal preference relations are allowed to be deviated from the consistent matrices to a certain degree. In this study, it is assumed that the initial positions of the experts in a group decision making problem are characterized by multiplicative reciprocal matrices over a set of alternatives. A flexibility degree is offered to each expert so that her/his position can be modified opportunely. The information granules are constructed according to two different approaches. In order to build consensus in group decision making, a novel objective function is established to measure the consensus level within the multiple experts and to quantify the level of group consistency. Based on a modified particle swarm optimization (PSO) algorithm, the dynamic and iterative consensus process in group decision making is modeled by minimizing the objective function. Numerical results are reported to illustrate the proposed consensus model in group decision making. The novelty is revealed by comparing the existing group decision making model based on the PSO method. The proposed model could be used to deal with the large-scale group decision making problem with social network information.