A New Deep Transfer Learning Based on Sparse Auto-Encoder for Fault Diagnosis

By Long Wen, Liang Gao, Xinyu Li. IEEE Transactions on Systems, Man & Cybernetics. Systems, 2019

927 citations recorded by the source corpus.

Fault diagnosis plays an important role in modern industry. With the development of smart manufacturing, the data-driven fault diagnosis becomes hot. However, traditional methods have two shortcomings: 1) their performances depend on the good design of handcrafted features of data, but it is difficult to predesign these features and 2) they work well under a general assumption: the training data and testing data should be drawn from the same distribution, but this assumption fails in many engineering applications. Since deep learning (DL) can extract the hierarchical representation features of raw data, and transfer learning provides a good way to perform a learning task on the different but related distribution datasets, deep transfer learning (DTL) has been developed for fault diagnosis. In this paper, a new DTL method is proposed. It uses a three-layer sparse auto-encoder to extract the features of raw data, and applies the maximum mean discrepancy term to minimizing the discrepancy penalty between the features from training data and testing data. The proposed DTL is tested on the famous motor bearing dataset from the Case Western Reserve University. The results show a good improvement, and DTL achieves higher prediction accuracies on most experiments than DL. The prediction accuracy of DTL, which is as high as 99.82%, is better than the results of other algorithms, including deep belief network, sparse filter, artificial neural network, support vector machine and some other traditional methods. What is more, two additional analytical experiments are conducted. The results show that a good unlabeled third dataset may be helpful to DTL, and a good linear relationship between the final prediction accuracies and their standard deviations have been observed.

Subfields: Control and Systems Engineering, Mechanics of Materials

View Paper: https://www.semanticscholar.org/paper/0d17094660ebe6eb3489351f28894712fffc012e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Control and Systems Engineering

In Mechanics of Materials

Citations in the corpus

A transfer convolutional neural network for fault diagnosis based on ResNet-50 (2019, background): “Wen et al. [19] investigated a sparse autoencoder (SAE)-based deep transfer learning for fault diagnosis and achieved the accuracy as high as 99.82%.”
A New Convolutional Neural Network-Based Data-Driven Fault Diagnosis Method (2018, methodology): “Several DL methods had already been applied on the fault diagnosis, such as deep belief network (DBN) [12], sparse auto-encoder [13], stacked denois-”
WaveletKernelNet: An Interpretable Deep Neural Network for Industrial Intelligent Diagnosis (2019, background): “There are many kinds of neural network structures in deep learning, such as deep belief network (DBN) [9], [10], auto-encoder (AE) [11], [12] and convolutional neural network (CNN) [13], [14].”
Online Fault Diagnosis Method Based on Transfer Convolutional Neural Networks (2020, methodology): “[24] presented an SAEbased transfer learning approach for bearing fault diagnosis, where the features of raw data are extracted by SAE and the maximum mean discrepancy term is applied to minimize the discrepancy penalty between the feature spaces from the source domain and the target domain.”
Deep Transfer Learning for Bearing Fault Diagnosis: A Systematic Review Since 2016 (2023, methodology): “i) Statistic-based methods: As mentioned in previous paragraphs, MMD and its variants are widely used in statistic-based transfer learning methods [50], [69], [111], [126], [131], [132].”
Deep Convolutional Transfer Learning Network: A New Method for Intelligent Fault Diagnosis of Machines With Unlabeled Data (2019, background): “Currently, several transfer-learning-based intelligent fault diagnosis methods have been proposed [13], [23], [24].”
Deep Transfer Network with Joint Distribution Adaptation: A New Intelligent Fault Diagnosis Framework for Industry Application (2018, methodology): “In [37], the authors developed a SAE based domain adaptation method for bearing diagnosis across diverse operating conditions, where a maximum mean discrepancy (MMD) term is utilized to measure the domain discrepancy.”
Fault Description Based Attribute Transfer for Zero-Sample Industrial Fault Diagnosis (2021, background): “Hence, Long et al. [15] integrated the deep neural network with the maximum mean discrepancy item to present a deep transfer learning for fault diagnosis, in which none of samples of the target faults were necessary for highly accurate performance.”
A Review of Deep Transfer Learning and Recent Advancements (2022, methodology): “Even with implementation limitations (freezing part of a layer), our evaluation demonstrates the high potential of the EXPANSE approach for a successful transfer in deep learning that can be applied to different DL models and tasks.”
A Digital-Twin-Assisted Fault Diagnosis Using Deep Transfer Learning (2019, methodology): “[33] proposed a supervised DTL based on auto-encoder, they demonstrated the superior diagnosis accuracy of proposed method by comparing with other algorithms.”