Abstract: This paper studies the sorting algorithm of triangular fuzzy number for elements of the reciprocal judgment matrix in fuzzy analytical hierarchy process and proposes improved fuzzy probability algorithm. The paper estimates and corrects the satisfied consistency of reciprocal judgment matrix under the improved satisfied consistency index, and uses the fuzzy probability algorithm to transfer feature vector of the triangular fuzzy number judgment matrix into exact value, with a consequence of getting sort vector. The results from analyzing examples show that sort vector approaches to ideal judgment and the algorithm are more accurate.

Key words: triangular fuzzy number, analytical hierarchy process, reciprocal judgment matrix, consistency judgment, triangular fuzzy number sequence

摘要： 研究了模糊层次分析法中以三角模糊数为元素的互反判断矩阵排序方法问题，提出一种改进模糊概率算法。采用改进的满意一致性判断指标对互反判断矩阵判断与修正；用模糊概率算法将判断矩阵特征向量中的三角模糊数转换成准确值，最终得出排序向量。实例分析结果表明，排序向量更加趋近理想判断，算法更具准确性。

关键词: 三角模糊数, 层次分析法, 互反判断矩阵, 一致性判断, 三角模糊数排序