Abstract: In order to improve the accuracy of the decision making process of high speed railway traction power supply, and to ensure the scientific and reasonable maintenance decision weight assignment, a new method based on triangular fuzzy rule for FANP maintenance decision making of high speed railway traction power supply is proposed. Firstly, the triangular fuzzy rules of classical network analysis method is improved, according to the fuzzy preference programming method, the index weight of high speed rail traction power supply is determined, involving weights of criteria, weighted coupling and affiliated index weight; Secondly, for constraints and coupling relation indexes, according to the network structure, the weighted super matrix is constructed and dealt with in the random way to obtain the limit form of the super matrix, and then the weight of each index of comprehensive is obtained, candidate for the weight score to choose the best maintenance method. Finally, the simulation test indicates that the FANP algorithm is better than the traditional maintenance decision-making plan and standard hierarchical network (ANP) scheme in the limit relative grade, dominance index, maintenance cost and time. Sensitivity analysis results of the algorithm are given.

Key words: Triangular fuzzy, high speed railway, traction power supply, network hierarchy process

摘要： 为进一步提高高铁牵引供电维修决策准确性，确保维修决策权重赋值科学合理，提出一种基于三角模糊规则的高铁牵引供电的模糊网络分析(FANP)维修决策方式。首先，利用三角模糊规则对经典的网络分析法进行改进，根据模糊优先规划方法对高铁牵引供电的各项指标权重进行确定，涉及准则权重、耦合权重以及附属指标权重；其次，对具有制约和耦合关系指标，按照网络结构构建未加权形式的超矩阵，并利用随机方式对其进行处理，获得极限形式的超矩阵，然后对所获得的各项指标权重进行综合，获得候选方案权重得分，从而选择最佳的维修方案。最后，通过对实例进行仿真测试，表明所提FANP算法在极限相对等级、优势度指标、维护费用和时间上要优于传统维修决策方案及标准层次网
络(ANP)方案，同时给出了算法的因素灵敏度分析结果。

关键词: 三角模糊, 高铁, 牵引供电, 层次网络法