Stacked pruning sparse denoising autoencoder based intelligent fault diagnosis of rolling bearings

This paper proposes a new stacked pruning sparse denoising autoencoder (sPSDAE) model for intelligent fault diagnosis of rolling bearings. Different from the traditional autoencoder, the proposed sPSDAE model, including a fully connected autoencoder network, uses the superior features extracted in a...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Applied soft computing Jg. 88; S. 106060
Hauptverfasser: Zhu, Haiping, Cheng, Jiaxin, Zhang, Cong, Wu, Jun, Shao, Xinyu
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 01.03.2020
Schlagworte:
Deep learning
Fault diagnosis
Pruning operation
Rolling bearing
Stacked pruning sparse denoising autoencoder
Deep learning
Fault diagnosis
Stacked pruning sparse denoising autoencoder
Pruning operation
Rolling bearing
ISSN:1568-4946, 1872-9681
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:This paper proposes a new stacked pruning sparse denoising autoencoder (sPSDAE) model for intelligent fault diagnosis of rolling bearings. Different from the traditional autoencoder, the proposed sPSDAE model, including a fully connected autoencoder network, uses the superior features extracted in all the previous layers to participate in the subsequent layers. This means that some new channels are created to connect the front layers and the back layers, which reduces information loss. To improve the training efficiency and precision of the sPSDAE model, a pruning operation is added into the sPSDAE model so as to prohibit non-superior units from participating in all the subsequent layers. Meanwhile, a feature fusion mechanism is introduced to ensure the uniqueness of the feature dimensions. After that, the sparse expression of the sPSDAE model is strengthened, thereby improving the generalization ability. The proposed method is evaluated by using a public bearing dataset and is compared with other popular fault diagnosis models. The results show that the ability of the sPSDAE model to extract features is significantly enhanced and the phenomenon of gradient disappearance is further reduced. The proposed model achieves higher diagnostic accuracy than other popular fault diagnosis models. •A new sPSDAE model is proposed to diagnose bearing faults.•Proposed fully connected network enhances information transfer efficiency significantly.•Pruning operation is designed to reduce the network redundancy.•Proposed feature fusion method shares feature information and guarantees the uniqueness of the feature dimensions.•A variety of methods are used to optimize the sparsity of the model.
ISSN:1568-4946
1872-9681
DOI:10.1016/j.asoc.2019.106060