基于梯形模糊数联系数的评价方法及其应用

Assessment method using connection number based on trapezoidal fuzzy numbers and its application

针对当联系分量系数取值具有峰值较宽的分布特征时,利用三角模糊数表示差异度系数可能存在误差较大的问题,通过分析差异度系数取值与等级标准阈值的对应关系,确定了差异度系数的最优取值区间,构造了以梯形模糊数表示的差异度系数,提出了基于梯形模糊数联系数的评价方法。将该评价方法应用于巢湖流域水资源系统脆弱性评价,结果表明:置信水平75%条件下巢湖流域水资源系统脆弱性等级置信区间为[2.89,3.44];所提出的方法不仅能得出评价等级的置信区间,有效避免差异度系数峰值较宽时产生的误差,还可以采用"期望-方差排序法"对不同子区域的脆弱性等级进行优劣排序。

Using a triangular fuzzy number to express the difference degree coefficient might lead to a large error when the connection components show a broad peak distribution characteristic. The optimal interval of the difference degree coefficient was determined, the difference degree coefficient was formulated with the trapezoidal fuzzy numbers, and an assessment method using the connection number based on trapezoidal fuzzy numbers was developed through analysis of the relationship between the difference degree coefficient and grade standard threshold. The method was applied to vulnerability assessment of the water resources system in the Chaohu Basin. The results show that the confidence interval of the vulnerability grade in the water resources system in the Chaohu Basin is [2. 89, 3. 44] with a 75% confidence level. This method can not only obtain the confidence interval of the evaluation grade and effectively avoid the error induced by the broad peak of the difference degree coefficient, but also determine the priorities of vulnerability grades in different sub-regions using the expectation-variance method.