论文部分内容阅读
液压泵源是民机液压系统的动力部分,对于民航飞机飞行的安全起着重要的作用。由于民机液压泵结构复杂,故障机理繁多,因此其故障诊断难以用常规方法实现。提出了基于小波包分解和RBF神经网络的民机液压泵源故障诊断方法。采集民机液压泵轴向和径向的振动信号,利用小波包分解将振动信号进行分解,得到的各频带信号的能量作为神经网络的输入。经过RBF神经网络的计算以实现民机液压泵源的故障诊断。
Hydraulic pump source is the power part of civil aircraft hydraulic system, plays an important role in the safety of civil aircraft flight. Due to the complicated structure of hydraulic pump of civil aircraft and the numerous failure mechanisms, it is difficult to diagnose the fault with the conventional method. A fault diagnosis method of hydraulic pump source based on wavelet packet decomposition and RBF neural network is proposed. The axial and radial vibration signals of civil hydraulic pump are collected, and the vibration signals are decomposed by wavelet packet decomposition. The energy of each frequency band signal is taken as the input of neural network. After RBF neural network calculation to achieve the diagnosis of civil aircraft hydraulic pump fault.