TY - JOUR
T1 - Transformer-based hierarchical latent space VAE for interpretable remaining useful life prediction
AU - Jing, Tao
AU - Zheng, Pai
AU - Xia, Liqiao
AU - Liu, Tianyuan
N1 - Funding Information:
This research is partially funded by the Mainland-Hong Kong Joint Funding Scheme ( MHX/001/20 ), Innovation and Technology Commission (ITC) , Hong Kong Special Administration Region , and National Key R&D Programs of Cooperation on Science and Technology Innovation with Hong Kong, Macao and Taiwan ( SQ2020YFE020182 ), Ministry of Science and Technology (MOST) of the People’s Republic of China , and the state key laboratory of ultra-precision machining technology and the Research and Innovation Office of the Hong Kong Polytechnic University (Project codes: BBR2 and BBX7).
Publisher Copyright:
© 2022 Elsevier Ltd
PY - 2022/10
Y1 - 2022/10
N2 - Data-driven prediction of remaining useful life (RUL) has emerged as one of the most sought-after research in prognostics and health management (PHM). Nevertheless, most RUL prediction methods based on deep learning are black-box models that lack a visual interpretation to understand the RUL degradation process. To remedy the deficiency, we propose an intrinsically interpretable RUL prediction method based on three main modules: a temporal fusion separable convolutional network (TF-SCN), a hierarchical latent space variational auto-encoder (HLS-VAE), and a regressor. TF-SCN is used to extract the local feature information of the temporal signal. HLS-VAE is based on a transformer backbone that mines long-term temporal dependencies and compresses features into a hierarchical latent space. To enhance the streaming representation of the latent space, the temporal degradation information, i.e., health indicators (HI), is incorporated into the latent space in the form of inductive bias by using intermediate latent variables. The latent space can be used as a visual representation with self-interpretation to evaluate RUL degradation patterns visually. Experiments based on turbine engines show that the proposed approach achieves the same high-quality RUL prediction as black-box models while providing a latent space in which degradation rate can be captured to provide the interpretable evaluation.
AB - Data-driven prediction of remaining useful life (RUL) has emerged as one of the most sought-after research in prognostics and health management (PHM). Nevertheless, most RUL prediction methods based on deep learning are black-box models that lack a visual interpretation to understand the RUL degradation process. To remedy the deficiency, we propose an intrinsically interpretable RUL prediction method based on three main modules: a temporal fusion separable convolutional network (TF-SCN), a hierarchical latent space variational auto-encoder (HLS-VAE), and a regressor. TF-SCN is used to extract the local feature information of the temporal signal. HLS-VAE is based on a transformer backbone that mines long-term temporal dependencies and compresses features into a hierarchical latent space. To enhance the streaming representation of the latent space, the temporal degradation information, i.e., health indicators (HI), is incorporated into the latent space in the form of inductive bias by using intermediate latent variables. The latent space can be used as a visual representation with self-interpretation to evaluate RUL degradation patterns visually. Experiments based on turbine engines show that the proposed approach achieves the same high-quality RUL prediction as black-box models while providing a latent space in which degradation rate can be captured to provide the interpretable evaluation.
KW - Interpretable estimation
KW - Prognostics
KW - Remaining useful life
KW - Transformer encoder
KW - VAE
UR - http://www.scopus.com/inward/record.url?scp=85141265417&partnerID=8YFLogxK
U2 - 10.1016/j.aei.2022.101781
DO - 10.1016/j.aei.2022.101781
M3 - Journal article
AN - SCOPUS:85141265417
SN - 1474-0346
VL - 54
JO - Advanced Engineering Informatics
JF - Advanced Engineering Informatics
M1 - 101781
ER -