制造企业供应商排序决策:基于SVM和TFN-RS的改进TOPSIS

考虑制造企业供应链的绿色内涵,构建供应商排序指标体系。为解决指标数量众多、计算方式复杂、协调困难等问题,构造一种基于SVM和TFN-RS的改进TOPSIS方法来进行供应商排序决策。首先根据候选供应商的主要数据信息,应用基于SVM的分类模型对其进行初步筛选,以缩小候选供应商数量;通过专家对准则下属指标进行考察和掌握,利用专家的智慧和经验对准则进行TFN评分,设计一种TFN-RS方法来计算供应商在准则上的分值;运用CRITIC方法对各准则进行赋权;最后通过相对熵替代欧氏距离的改进TOPSIS对各供应商进行排序。以某轴承制造企业的滚球供应商排序决策为例,验证了该方法的实用效果。

Vendors sequencing for manufacturing enterprise:Improved TOPSIS based on SVM and trapezoidalfuzzy number-rough set method

Considering the green connotation of manufacturing enterprise supply chain, an indicator system of vendors sequencing is established. To solve the problems such as large index number, complex calculation, and coordination difficulties, an improved TOPSIS based on SVM and trapezoidal fuzzy number rough set method is proposed. According to the main data, a SVM based classification model is applied for the preliminary screening of candidate vendors. Through the investigation and mastery of experts to the indicators of criteria, a trapezoidal fuzzy number rough set method is designed to calculate the value of a vendor in the criteria by using the wisdom and experience of experts. The criteria weight is determined by CRITIC. Finally, the vendors are sequenced by the improved TOPSIS replacing Euclidean distance with relative entropy. An example of ball vendors sequencing decision of a bearing manufacturing enterprise is used to verify the practical effect of this method.