基于一致性修正的多属性大群体决策方法

摘要

针对多属性大群体决策中群体一致性较差的问题,提出了一种新的群体一致性修正方法.建立了相似度和一致度双重控制变量下的目标规划模型并利用改进的花朵授粉算法求解专家权重;结合聚类思想中的属性权重调整过程,构建了群体一致性修正模型,通过评价函数的非线性迭代使群体一致性不断提高,最终达到决策者满意的阈值.通过实例验证了该方法的有效性和实用性.%According to a problem of poor group-consistency in large group-decision,this paper proposes a new method to amend group-consistency. An objective programming model is constructed based on similarity degree and consistent degree,then an improved flower pollination algorithm is used to solve the model and expert weight is determined. In order to amend group-consistency,a model is constructed combined with the thought of adjustment on attribute weight in clustering analysis. By nonlinear iterating on evaluation function,the group-consistency is constantly improved and finally reach decision-maker's satisfied threshold.A calculation example is used to show the effectiveness and practicality of the models and algorithm.