摘要
行星齿轮箱振动信号在各频段的能量分布与其故障类型有关。利用Daubecics小波包将不同故障的振动信号分解到各个频带。BP神经网络的输入是各频带的能量——行星齿轮故障的特征向量,用神经网络识别故障类型。通过实验验证了该方法可以快速、准确地进行故障模式识别,达到良好的预期效果。利用此方法可以有效解决武装直升机武器系统复杂故障现象问题。
The energy distribution of each frequency sub band of vibration signal has relationship with the type of planetary gearbox defects.Vibration signals for different faults are decomposed into frequency sub bands by using Daubechies wavelet packet and the frequency energies as energy features of gearbox are selected.The energy features are then used as inputs to a back propagation neural network classifiers for identifying the fault types of planetay gearbox.The experimental results indicated that the proposed fault diagnosis method is effective and can quickly detect and accurately locate the fault diagnosis.This method can effectively solve the problem of complex armed helicopter weapon system fault phenomenon.
作者
罗佳
黄晋英
LUO Jia;HUANG Jin-ying(College of Mechanical Engineering,North University of China,Taiyuan 030051,China)
出处
《火力与指挥控制》
CSCD
北大核心
2020年第4期178-182,共5页
Fire Control & Command Control