关键词: 模糊加权影响情景下的非线性测度体系（WINGS）, 网络分析法, 超矩阵, 加权矩阵

Abstract: It is vital to construct a Cluster-Weighted Matrix（CWM） among clusters when converting an unweighted supermatrix of ANP（Analytic Network Process） system to the weighted supermatrix. However, existing three construction methods have the following disadvantages. First, pairwise comparison method has disordered comparison mechanism. Second, the method of supposing equal weight matrix is usually inefficient unless the CWM is a nominal matrix. Third, the construction method based on DEMATEL（Decision Making Trial and Evaluation Lab） not only ignores the Importance of Each Cluster（IEC）, also is the inaccuracy of expert judgments overlooked. To overcome the above drawbacks, this paper suggests a new fuzzy WINGS（Weighted Influence Non-linear Gauge System） to construct CWM of ANP. On one hand, in the new approach, the inaccuracy of expert judgments is considered by improved fuzzy DELPHI procedures. On the other hand, the idea of fuzzy WINGS is suggested in detail, and IEC is added when analyzing the influence relation among clusters, which makes the direct influence matrix be more rational. By a numerical case, the approach is validated to be scientific and it can be well applied to solving the real world issues.

Key words: fuzzy Weighted Influence Non-linear Gauge System（WINGS）, analytic network process, supermatrix, Cluster-Weighted Matrix（CWM）