APU故障智能诊断系统设计

The designer of APU fault intelligent diagnosis system

摘要:

针对飞机辅助动力装置(APU)故障诊断智能化、自动化水平较低问题,设计了基于MATLAB语言的APU智能故障诊断系统。通过小波包分析数字滤波算法实现对APU监控传感器输出数据的滤波处理,创造性地利用“Min-Max”算法对APU监控传感器是否发生故障进行诊断,确保传感器输出的数据可以真实地反应APU的工 作状况。智能故障诊断系统通过分析经过处理后的数据,实现征兆空间到故障空间的映射而实现故障的智能诊断。最后利用南航沈阳维修基地APS3200型APU的相关数据对系统进行检测分析,实验结果表明该智能故障诊断系统具有学习速度快、噪声干扰抑制能力强及故障诊断结果准确等特点,对提高APU维修效率、节省维修费用具有重要意义,具有一定的实际推广应用价值。

关键词:辅助动力装置;智能故障诊断;小波包分析;Min-Max算法;改进BP网络

Abstract:

In accordance with the low level of aircraft power unit’s intelligent fault diagnosis and maintenance automation, intelligent fault diagnosis system based on matlab was presented. Wavelet packet analysis was used to deal with the data from the monitoring sensor of APU; In order to ensure the monitoring sensor can truly reflect the capability of APU, “Min-Max” algorithm was creative used to judge it’s data. Through analysis the processed data to achieving symptom space mapping fault space by the intelligent fault diagnosis system. Finally the APS3200 types APU’s data was used to detected and analyzed the system. The experimental results shows that the intelligent fault diagnosis system has the character of learn fast, strong immunity of noise and fault diagnosis accurate. The system has an important meaning in improve APU maintenance efficiency, saving cost and practical application.

Key words:APU; intelligent fault diagnosis; wavelet packet analysis; “Min-Max” algorithm; improved BP network