航空发动机转子故障诊断系统设计与实现  

DESIGN AND IMPLEMENTATION OF FAULT DIAGNOSIS SYSTEM FOR AEROENGINE ROTOR SYSTEM

作　　者：张俊红[1] 贾晓杰[1] 马文朋[1] 

机构地区：[1]天津大学内燃机燃烧学国家重点实验室,天津300072

出　　处：《计算机应用与软件》2015年第11期217-221,276,共6页Computer Applications and Software

基　　金：国家自然科学基金委员会与中国民用航空局联合项目(U1233201);天津市科技支撑计划项目(13ZCZDGX00200)

摘　　要：为准确、高效地诊断航空发动机转子系统故障,采用面向对象的程序设计方法,在VC++开发环境下,设计开发了航空发动机转子故障诊断系统。系统包含四个模块:数据采集模块采用多线程处理技术,配合数据采集设备实现多通道数据同步采集;信号分析和特征提取模块利用封装的改进经验模态分解算法类CEmd进行故障信号处理并从中提取故障特征;模式识别模块基于Lib SVM和GAlib库模板建立了基于支持向量机的故障识别模型。在航空发动机转子模拟实验台上对该系统进行测试,结果表明参数设置适当时,系统对转子典型故障的诊断精度达到了98%。In order to diagnose the faults in rotor system of aeroengine accurately and efficiently, by using object-oriented programming method we designed and implemented a fault diagnosis system for aeroengine rotor system in VC ＋ ＋ development environment. The system consists of four modules ： The data acquisition module adopts multithreading approach and implements synchronous collection of multichannel data in cooperation with data acquisition equipment. In modules of signal analysis and feature extraction, they utilise the improved EMD algorithm which is encapsulated as class CEmd to process the fault signals and extract from them the features. Based on LibSVM and GAlib library templates, the pattern recognition module establishes an SVM-based fault recognition model. The system has been tested on aeroengine rotor simulator rig, and the result showed that the system reached a high diagnosis precision of 98% for typical rotor faults when the parameters were properly set.

关 键 词：航空发动机转子 故障诊断系统 经验模态分解 支持向量机 VC++开发环境 

分 类 号：TP302.1[自动化与计算机技术—计算机系统结构] TP311.52[自动化与计算机技术—计算机科学与技术]