/Deep-Convolutional-Neural-Networks-with-Wide-First-layer-Kernels

这是一个首层卷积为宽卷积的深度神经网络Deep Convolutional Neural Networks with Wide First-layer Kernels (WDCNN)的实现，该模型具有优越的抗噪能力，可用于轴承的智能故障诊断。

Primary LanguagePython

Deep-Convolutional-Neural-Networks-with-Wide-First-layer-Kernels

这是一个首层卷积为宽卷积的深度神经网络Deep Convolutional Neural Networks with Wide First-layer Kernels (WDCNN)的实现，该模型具有优越的抗噪能力，可用于轴承的智能故障诊断。

通过借鉴该模型，发表科研论文两篇

Intelligent Motor Bearing Fault Diagnosis Using Channel Attention-Based CNN

Fault diagnosis method for imbalanced bearing data based on W-DCGAN

模型结构

Architecture of the proposed WDCNN model

不同数据量对模型性能的影响

Diagnosis results using different numbers of training samples

t-SNE可视化

Feature visualization via t-SNE

Feature visualization via t-SNE

抗噪性分析

Results of the proposed WDCNN and WDCNN-AdaBN of six domain shifts on the Datasets A, B and C, compared with FFT-SVM, FFT-MLP and FFT-DNN

卷积可视化

Visualization of all convolutional neuron activations in WDCNN