基于小波包变换的柱塞泵脉动压力信号BP故障诊断

BP Network Fault Diagnosis of Piston Pump Pulsation Pressure Signal Based on Wavelet Packet Transform

为了提高柱塞泵脉动压力信号的故障诊断诊断能力,分别提取得到正常的泵脉动压力与齿圈出现磨损故障时对应的泵脉动压力,利用小波包变换的方法消除原始信号的噪声信号,再提取得到由故障信息构成的频段,完成信号时域特征的分析。频段介于0~4 kHz之间时,在柱塞泵脉动压力信号中形成了泵轴频、基频与谐倍频等,含有泵运行过程的各项信息参数,处于高频区间时,发生了线谱幅值的持续减小。利用MATLAB软件建立BP网络故障诊断结果表明:脉动压力指标可以实现柱塞泵内部运行情况的准确识别,达到故障诊断的测试要求。方法在诊断方面是有效的,在处理柱塞泵脉动压力信号故障诊断领域具有很好的实际指导意义。

In order to improve the diagnosis of the fault diagnosis of plunger pump pulse pressure signal,the normal fluctuation pressure pump and gear ring wear failure of pump pulse pressure were got respectively.By using the wavelet packet decomposition method to eliminate the noise of the original signal signal,then get composed of fault information extraction frequency,the analysis of the complete time domain signal characteristics.When the frequency band is between 0 and 4 kHz,the pump shaft frequency,fundamental frequency and harmonic multiple frequency are formed in the pulsating pressure signal of the piston pump,which contains various information parameters of the pump operation process.When the frequency band is in the high frequency range,the linear spectrum amplitude decreases continuously.The fault diagnosis results of BP neural network established by MATLAB software show that the pulsating pressure index can realize the accurate identification of the internal operation of the piston pump and meet the test requirements of fault diagnosis.The method presented in this paper is effective in diagnosis and has a good practical guiding significance in the field of fault diagnosis of pulsating pressure signal of piston pump.