一种基于云模型和惩罚函数的多属性评价方法

通过系统阐述多属性评价中存在的＂不确定性转换问题＂和＂属性组态控制问题＂,提出了一种基于云模型和惩罚函数的多属性评价方法。以云模型为理论基础,提出了定性属性、定量属性的云化处理方法,实现了属性值的不确定性度量;以惩罚函数为理论基础,根据不同属性的惩罚幅度、灵敏度需求,构建了3种连续型惩罚函数,实现了属性组态的有效控制;最后,提出了用云模型表达的属性值集结算法和排序选优方法。通过实例应用,验证了方法的可行性、有效性。

A Method of Multiple Attributive Evaluation Based on Cloud Model and Castigation Function

After the problem of uncertainty transforming and attributive configuration controlling existed in multiple attributive evaluation is systemically analyzed,a method for multiple attributive evaluation based on cloud model and castigation function is presented.Firstly,on the basis of cloud model theory,the algorithm of transforming of qualitative attributive(quantity attributive) is proposed to realize the uncertainty in attributive measuring.Secondly,on the basis of castigation function theory,according to the different needs of castigation degree and sensitivity,three kinds of consecutive castigation function are developed to realize effective control of attributive configuration.Finally,attributive aggregation and sorting algorithm are presented with cloud model.The feasibility and effectiveness of the method are validated through application of examples.