Deep Semi-supervised Domain Generalization Network for Rotary Machinery Fault Diagnosis under Variable Speed

Liao, Yixiao; Huang, Ruyi; Li, Jipu; Chen, Zhuyun; Li, Weihua

doi:10.1109/tim.2020.2992829

Cited by 96 publications

(39 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then, the kernel number of the convolution layer is investigated. The kernel number of convolution layers can be regarded as the number of eigenvectors obtained by convolution, and the kernel number of the convolution layer is set in the range of [2,20] with an interval of 2. From the reconstruction error in Figure 9a, the training error becomes smaller with the increasing of the kernel number, but the decreasing of training error becomes very slow when the kernel number is bigger than 10.…”

Section: The Number Of Convolution Kernelmentioning

confidence: 99%

“…The size of convolution kernel The kernel size of convolution layer is the sensing domain of convolution network in feature extraction, which can also be regarded as the window function of feature extraction. The kernel size affects the calculation amount in each convolution, and it is also investigated in the range of [2,20]. As shown in Figure 10a, the training error presents a decreasing trend as the increasing of the kernel number, but this trend slowed down when kernel number is bigger than 5.…”

Section: The Number Of Convolution Kernelmentioning

confidence: 99%

“…The parameters need to be determined in the shallow two-layer LSTM network are the node numbers of recurrent layer. The node numbers in these two hidden layers are explored in the range of [20,60] and [10,40], respectively. .…”

Section: Parameter Determinationmentioning

confidence: 99%

“…With the development of machine learning, more attention has been paid to data-driven methods [18][19][20][21][22] for their remarkable feature fusion ability, such as Gaussian mixture model (GMM), support vector machine (SVM), and deeplearning methods. Ren 23 studied GMM-based damage information representation method and utilized path-synthesis accumulation imaging to localize the defect.…”

mentioning

confidence: 99%

See 3 more Smart Citations

Deep‐learning‐based guided wave detection for liquid‐level state in porcelain bushing type terminal

Hong

Zhang

Liu

et al. 2020

Struct Control Health Monit

Self Cite

View full text Add to dashboard Cite

The structural security of civil energy equipment is significant for the steady operation of power supply system, and porcelain bushing type terminal is a typical energy equipment that needs long-term monitoring. As a nondestructive structural health monitoring method, ultrasonic guided wave (UGW) technology is extremely suitable for state detection of energy equipment. However, most current UGW methods still need to manually select the guided wave features, which rely heavily on the guidance of expert experience. This article presents a deep-learning method to directly utilize original-guided wave signals to quantitatively detect the liquid-level state. Firstly, the original signals were fed into convolutional autoencoder (CAE) to catch the low-dimension representation and realize the automatic feature extraction. Then, the low-dimension representations were orderly input into the long short-term memory (LSTM) recurrent neural network for liquidlevel regression. In feature extraction step, CAE method can effectively extract the useful features and remove the interference and signal distortion. In regression step, both the accuracy and the robustness of proposed method are better than backpropagation network and convolutional neural network. Experimental results show that proposed CAE-LSTM method can accurately inspect the liquid level by original signals and implement maintenance monitoring.

show abstract

Section: The Number Of Convolution Kernelmentioning

confidence: 99%

Section: The Number Of Convolution Kernelmentioning

confidence: 99%

Section: Parameter Determinationmentioning

confidence: 99%

mentioning

confidence: 99%

See 2 more Smart Citations

Deep‐learning‐based guided wave detection for liquid‐level state in porcelain bushing type terminal

Hong

Zhang

Liu

et al. 2020

Struct Control Health Monit

Self Cite

View full text Add to dashboard Cite

show abstract

“…Specifically, feature extraction was completed on the domain-level, and the classification was performed on the class-level simultaneously. Based on an adversarial idea, a deep semi-supervised learning model was employed for fault diagnosis of the transmission and bearing [ 31 ]. Jia et al employed a CNN with normalization to overcome the data imbalance and interpreted the learning process by visualization [ 32 ].…”

Section: Introductionmentioning

confidence: 99%

Intelligent Diagnosis towards Hydraulic Axial Piston Pump Using a Novel Integrated CNN Model

Tang

Zhu

Yuan

et al. 2020

Sensors

View full text Add to dashboard Cite

As a critical part of a hydraulic transmission system, a hydraulic axial piston pump plays an indispensable role in many significant industrial fields. Owing to the practical undesirable working environment and hidden faults, it is challenging to precisely and effectively detect and diagnose the varying fault in the engineering. Deep learning-based technology presents special strengths in processing mechanical big data. It can simultaneously complete the feature extraction and classification, and achieve the automatic information learning. The popular convolutional neural network (CNN) is exploited for its potent ability of image processing. In this paper, a novel combined intelligent method is developed for fault diagnosis towards a hydraulic axial piston pump. First, the conversion of signals to images is conducted via continuous wavelet transform; the effective feature is preliminarily extracted from the transformed time-frequency images. Second, a novel deep CNN model is constructed to achieve the fault classification. To disclose the potential learning in the disparate layers of the CNN model, the visualization of reduced features is performed by employing t-distributed stochastic neighbor embedding. The effectiveness and stability of the proposed model are validated through the experiments. With the proposed method, different fault types can be precisely identified and high classification accuracy is achieved in a hydraulic axial piston pump.

show abstract

A domain adaptation method for bearing fault diagnosis using multiple incomplete source data

Wang

Yang

et al. 2023

J Intell Manuf

View full text Add to dashboard Cite

Deep Semi-supervised Domain Generalization Network for Rotary Machinery Fault Diagnosis under Variable Speed

Cited by 96 publications

References 33 publications

Deep‐learning‐based guided wave detection for liquid‐level state in porcelain bushing type terminal

Deep‐learning‐based guided wave detection for liquid‐level state in porcelain bushing type terminal

Intelligent Diagnosis towards Hydraulic Axial Piston Pump Using a Novel Integrated CNN Model

A domain adaptation method for bearing fault diagnosis using multiple incomplete source data

Contact Info

Product

Resources

About