Abstract: In order to improve the accuracy of aero-engine component fault diagnosis ability, a method of two layers fault diagnosis based on gray relational analysis was proposed. Particle swarm optimization (PSO) was introduced to optimize gray relational weighted index under various degradation to build standard fault pattern module which was used to diagnose as the first layer. Then two most possible fault patterns were selected for further diagnosis as the second layer based on gray slope relational analysis. Simulation showes that the method proposed is more simple and with less calculation compared to the single layer diagnosis, while the method has better accuracy than two layers fault diagnosis based on experiential gray relational analysis.

Key words: Aero-engine; Fault diagnosis; Gray relational analysis; Particle swarm optimization (PSO)

摘要： 为了改善对航空发动机气路部件故障诊断能力,提出了一种基于灰色关联分析的两层诊断方法。该方法首先利用粒子群算法优化各蜕化程度下灰色关联加权指数,构建标准故障序列,利用灰色关联分析进行第一层定性诊断,再优选故障模式利用灰色斜率关联分析方法进行二次诊断,得到了气路部件故障诊断结果。仿真表明,改进二次灰色关联分析诊断方法比单层诊断方法结构更简单,计算量小,更适合于较多传感器的发动机诊断系统,比经验灰色关联分析方法诊断精度更高。

关键词: 航空发动机;故障诊断;灰色关联分析;粒子群算法