基于模糊网络与TOPSIS的备件品种确定方法

Fuzzy networks and TOPSIS based method of determining spare parts variety

舰船携行备件品种确定过程是一种典型的多属性决策问题,针对其属性评估时面临的复杂性与模糊性、评估过程透明性不足的问题,提出了一种集成模糊网络与熵权法来扩展TOPSIS的方法。该方法以语言值Z-numbers作为评估值表示形式,分别构造了效益和成本属性下的模糊网络子系统,将合并子系统规则的模糊隶属度函数作为聚合模糊隶属度函数,通过案例计算得到Spearman rho相关系数值。结果表明:模糊网络的应用提高了决策过程的透明性,熵权法的应用提高了属性权重确定过程中的客观性,证明了该方法较其他经典方法具有更高的准确性。

The process of determining spare parts variety is a typical multi-criteria decision-making problem.To address the complexity and fuzziness faced in criteria evaluation,as well as the lack of transparency in the evaluation process,a method integrating fuzzy networks and entropy weight method was proposed to extend TOPSIS.This method took the language value Z-numbers as the representation form of evaluation value,whereby the subsystems of fuzzy networks under the benefit and cost criteria were constructed respectively,while the fuzzy membership function of subsystem rules were merged into the aggregated fuzzy membership function.The ranking produced by the proposed method was comparatively validated by Spearman rho ranking correlation.The results show that the transpa-rency of the extended TOPSIS method is improved significantly.The ranking produced by the proposed method is comparatively validated by Spearman rho ranking correlation.And thus the proposed method outperforms the existing approaches in term of ranking performance.