基于领域加权稀疏的交通机电轴承故障诊断

Fault diagnosis of traffic electromechanical bearings based on domain-weighted sparsity

轴承是交通机电设备中最常发生故障的核心部件之一,一旦发生故障,整个系统将被关闭,甚至引发灾难性后果,因此,开展交通机电轴承故障检测具有重要意义。针对低信噪比条件下交通机电轴承冲击特征难提取问题,提出一种基于领域加权稀疏的交通机电轴承故障诊断方法。首先基于经典稀疏模型,引入基尼指数作为权重以区分信号各分量稀疏系数的贡献;其次,考虑系数之间相关性,研究领域系数去噪替代阈值降噪方法,提高对稀疏重构分量估计的准确性;最后,对重构稀疏信号进行包络检测以识别故障。利用机电轴承故障仿真信号和地铁轴箱轴承实测信号验证了所提方法的有效性。结果表明本文方法能有效实现交通机电轴承弱故障特征提取。

Bearing is one of the core components of traffic electromechanical equipment that most often fails,and once the failure occurs,the whole system will be shut down,and even lead to catastrophic consequences,therefore,it is of great significance to carry out the fault detection of traffic electromechanical bearings.Aiming at the problem of difficult extraction of impulse features of traffic electromechanical bearings under low signal-to-noise ratio conditions,a fault diagnosis method of traffic electromechanical bearings based on domain-weighted sparse was proposed.Firstly,based on the classical sparse model,the Gini index was introduced as the weight to distinguish the contribution of sparse coefficients of each signal component;secondly,considering the correlation between the coefficients,the domain coefficients denoising was taken as an alternative to the threshold noise reduction method,to improve the accuracy of the estimation of the sparse reconstructed components;lastly,envelope detection was carried out on the reconstructed sparse signals to identify the faults.The validity of the proposed method was verifiy by using the simulated signals of electromechanical bearing faults and the measured signals of underground axle-box bearings,and the results show that the propose method could effectively achieve the weak fault feature extraction of traffic electromechanical bearings.