Rotating Machinery Fault Classification Based on One-Dimensional Residual Network with Attention Mechanism and Bidirectional Gated Recurrent Unit
Zhilin Dong,Dezun Zhao,lingli Cui
DOI: https://doi.org/10.1088/1361-6501/ad41fb
IF: 2.398
2024-04-25
Measurement Science and Technology
Abstract:Conventional convolutional neural networks (CNNs) predominantly emphasize spatial features of signals and often fall short in prioritizing sequential features. As the number of layers increases, they are prone to issues such as vanishing or exploding gradients, leading to training instability and subsequent erratic fluctuations in loss values and recognition rates. To address this issue, a novel hybrid model, termed one-dimensional residual network with attention mechanism and bidirectional gated recurrent unit (1D-RAM-BGRU) is developed for rotating machinery fault classification. First, a novel one-dimensional residual network (1D-ResNet) with optimized structure is constructed to obtain spatial features and mitigate the gradient vanishing or exploding. Second, the attention mechanism (AM)is designed to catch important impact characteristics for fault samples. Next, temporal features are mined through the bidirectional gated recurrent unit (BGRU .) Finally, feature information is summarized through global average pooling, and the fully connected layer is utilized to output the final classification result for rotating machinery fault diagnosis. The developed technique which is tested on one set of planetary gear data and three different sets of bearing data, has achieved classification accuracy of 98.5%, 100%, 100%, and 100%, respectively. Compared with other methods, including CNN, CNN-BGRU, CNN-AM, and CAM-BGRU, the proposed technique has the highest recognition rate and stable diagnostic performance.
engineering, multidisciplinary,instruments & instrumentation