小波和神经网络在柴油机故障诊断中的应用  

Application on Diesel Engine Fault Diagnosis Based on Wavelet and Neural Network

作　　者：吴虎胜[1] 吕建新[1] 王茂生[1] 许阳懿[1] 

机构地区：[1]中国人民武装警察部队工程学院,西安710086

出　　处：《农机化研究》2010年第10期207-210,共4页Journal of Agricultural Mechanization Research

摘　　要：柴油机以其良好的动力性、可靠性、经济性在运输车辆和农用机械中广泛应用,但对其施行及时的不解体故障诊断却并非易事。为此,以配气机构故障为例,提出将小波包分解与神经网络结合的故障诊断方法。先对振动信号应用小波阀值法降噪,再进行小波包分解,构造小波包特征向量作为故障样本,并用训练好的BP神经网络进行故障识别,试验结果证明了该方法的有效性。Diesel engine is widly used in transport vehicle and farm machine because of its good dynamic performance,economy and reliability,but it is not easy to be diagnosed without dismantlement and in time. Making valve train for example,the diagnostic method about combination of wavelet and neural network is presented in this paper. First,it denoises the vibration signals with wavelet thresholding and adopts wavelet packet to decompose the signal,then constructs the wavelet packet energy eigenvector as fault samples. Finally,use the BP neural network to diagnose the fault. The practical example shows the method is available.

关 键 词：柴油机 故障诊断 小波包 神经网络 

分 类 号：TK428[动力工程及工程热物理—动力机械及工程] TP389.1[自动化与计算机技术—计算机系统结构]