Publication Date
2-1-2024
Document Type
Article
Publication Title
Reliability Engineering and System Safety
Volume
242
DOI
10.1016/j.ress.2023.109776
Abstract
Degradation of engineered systems can result in poor performance and failure. Graph Convolutional Networks (GCNs) have been used to predict the remaining useful life (RUL) of engineered systems by analyzing condition monitoring data. Conventional GCNs typically stack multiple spectral graph convolutional layers, where each layer aggregates condition monitoring data and then projects the aggregated data into another feature space. However, conventional GCNs suffer from two issues. Firstly, repeated aggregation operations affect the temporal correlation of condition monitoring data. Secondly, repeated aggregation and projection operations may generate less significant features, resulting in poor prediction performance. To address these issues, we introduce a temporal convolutional operation to extract and preserve temporal features prior to repeated aggregation and projection operations. Additionally, we create an internal residual connection to skip some aggregation and projection operations to reduce the negative impact of the less significant features. Finally, we use an attention mechanism to extract the most significant features obtained from previous GCN layers and feed them to next GCN layers. We demonstrate the effectiveness of our method through three case studies. Our numerical results show that the proposed approach outperforms existing data-driven methods.
Keywords
Degradation modeling, Graph convolutional network, Residual network, Temporal convolutional network
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License
Department
Industrial and Systems Engineering
Recommended Citation
Yupeng Wei, Dazhong Wu, and Janis Terpenny. "Remaining useful life prediction using graph convolutional attention networks with temporal convolution-aware nested residual connections" Reliability Engineering and System Safety (2024). https://doi.org/10.1016/j.ress.2023.109776
