基于时频融合深度网络的矿用钻机轴承故障诊断

Bearing Fault Diagnosis of Mining Drilling Rig with Time-frequency-fused Deep Network

针对矿用钻机低速重载工作特性导致其轴承故障特征微弱、噪声大的问题,考虑单一模态下故障诊断的局限性,提出了一种基于时频融合深度网络的矿用钻机轴承故障诊断方法,对时域和时频域两种模态特征进行联合提取与分析。所设计的诊断网络在不同模块中区分性地嵌入不同注意力机制,实现多维度关键故障特征提取。最后通过钻机实验台数据集和凯斯西储大学轴承数据集进行验证。结果表明:所提方法能自动提取丰富的钻机轴承故障特征,比仅在时域或时频域分析具有更高的准确率和抗噪能力。

To solve the problems of weak and noisy bearing fault features caused by the low-speed and heavy-load operating characteristics of mining drilling rigs,a fault diagnosis method was proposed for mining rig bearings,named time-frequency-fused deep network.It considered the limitations of fault diagnosis with single modality,and then jointly characterizes two modal features of the time domain and time-frequency domain.The designed diagnostic network differentially embeded specific attention mechanism in different modules to extract multi-dimensional key fault features.Finally,the proposed method was validated on the experimental equipment and the Case Western Reserve University bearing dataset.The results show that the proposed method may automatically extract sufficient fault features combining two domains.It has higher accuracy and noise immunity than those with a single domain.