摘要详情

ID / 提交时间

17 / 2025-04-16 17:17:19

标题

Bearing Fault Diagnosis Based on Dual-channel DenseNet-GRU Model

关键字

DenseNet, gated recurrent unit, fault diagnosis, rolling bearing

主题及专题

4、System Monitoring, Fault Diagnosis and Prediction

状态

全文待审

作者

皓张 / 长安大学

锐姚 / 长安大学

奥博贾 / 长安大学

康李 / 长安大学

群马 / 长安大学

萌惠 / 长安大学

摘要

In practical engineering applications, noise often contaminates the fault signals of rolling bearings, making the accurate diagnosis of compound faults challenging. To address this issue, this paper introduces an enhanced dual-channel DenseNet-GRU model for the diagnosis of compound faults in rolling bearings. The model constructs a DenseNet channel for initial feature extraction, while integrating a gated recurrent unit (GRU) with convolutional and pooling layers to form a GRU channel, aiming to extract linear features. By employing a dual-channel connection approach, the model minimizes potential information loss or error accumulation that may occur in single-model structures. In the identification module, a multi-label classification framework is established to recognize compound faults. The proposed model underwent evaluation using the Case Western Reserve University (CWRU) dataset, with findings indicating that the DC-DenseNet-GRU architecture consistently delivers robust performance across varying load and noise scenarios.