摘要
针对盾构机振动信号的故障特征频率不易提取的问题,提出一种基于LCD(局部特征尺度分解)和FastICA(快速独立分量分析)相结合的故障诊断方式。通过对待测信号进行LCD分解,将其得到的各内禀尺度分量(ISC)再利用快速谱峭度-相关系数的筛选准则处理,获取真实的ISC分量来重构信号。最后将重构信号与待测信号组成FastICA的输入矩阵,实行降噪得到所需故障特征频率得以识别相应的故障类型。实验结果表明相比于固有时间尺度分解,本方法能够减小外界噪声的干扰,更好的提取出机械振动信号的故障特征频率。
Aiming at solving the problem of the difficulty in extracting the the fault characteristic frequency of the vibration signal of a shield machine,put forward in the present paper is a fault diagnosis method of combining LCD(Local Characteristic-scale Decomposition)with Fast ICA(fast Independent Component Analysis).By applying the LCD decomposition to the signal to be measured,the Intrinsic Scale Component(ISC)components obtained are screened by the criteria of the fast spectral kurtosis-correlation coefficient to reconstruct the signal through obtaining the true ISC component.Then,the input matrix of Fast ICA is composed of the reconstructed signal and the signal to be measured,and finally,the required fault characteristic component can be identified to identify the corresponding fault type by reducing the noise.The experimental results show that compared with the Intrinsic Time-scale Decomposition,this method can reduce the interference of external noise and help better extract the fault characteristic frequency of the mechanical vibration signal.
作者
李凤远
冀勇
张贻凡
陆春华
LI Fengyuan;JI Yong;ZHANG Yifan;LU Chunhua(State Key Laboratory of Shield Machines and Boring Technology,Zhengzhou450001,China;Xi'an Triumph Electronic Co.Ltd.of Science and Technology,Xi'an710065,China;The6th Engineering Co.Ltd.of the17th Bureau Group of China Railway,Xiamen361000,China)
出处
《国防交通工程与技术》
2019年第4期22-26,共5页
Traffic Engineering and Technology for National Defence
