Addressing the issue that single sensor data is insufficient to fully characterize the fault state information of rolling bearings,leading to suboptimal fault diagnosis results,a fault diagnosis method for rolling bearings based on multi-source data fusion is proposed.To enable the model to fully characterize the equipment's operational state information,vibration and current signals are combined through channel concatenation to construct multi-channel data.Meanwhile,to suppress irrelevant information interference in multi-channel data,a Convolutional Block Attention Module(CBAM)is introduced into the Self-Calibrated Convolutional(SCConv)neural network for adaptive weighting of different channel data.In a series of comparative experiments,the proposed method achieved a classification accuracy of 100%,demonstrating excellent robustness and adaptability.
关键词
滚动轴承/故障诊断/多源数据融合/卷积注意力/自校正卷积
Key words
rolling bearings/fault diagnosis/multi-source data fusion/convolutional block attention/Self-Calibrated Convolutional
维普
万方数据