基于小波和神经网络的柴油机失火故障检测

Diesel Misfire Fault Diagnosis based on Wavelet and Neural Network

采用LabVIEW虚拟仪器搭建柴油机故障测试平台,采集了柴油机在正常、进气管阻塞、排气管阻塞、供油不足4种状态下的振动信号。利用小波分析对信号使用改进阈值消噪方法进行消噪,得出缸盖振动信号的频谱图,将柴油机振动信号的频率与能量的特征作为振动信号的特征值,设计了基于BP神经网络的柴油机故障诊断识别系统,识别准确率均达到80%以上。

Using the LabVIEW virtual instrument to build engine fault diagnosis system,vibration signals under four kinds of conditions of engines in normal,intake pipe jams,tailpipes jams,propulsion shortage were collected.Using wavelet analysis and improved threshold de-noising method to de-noising,the vibration signal spectrum graph was obtained.Used diesel vibration signal and energy characteristics as the eigenvalue of vibration signals,the diesel engine misfire fault diagnosis system was designed based on the BP neural network,more than 80% of accuracy rate of diagnosis was achieved.