Answers are grounded in this paper and cite their sources.
Precise monitoring of ground deformation in hydraulic structures is essential for ensuring the safety and stability of infrastructure. This study proposes a comprehensive method for Global Navigation Satellite System (GNSS) deformation monitoring in hydraulic structures. The proposed method combines the strengths of convolutional neural networks (CNN), bidirectional long short-term memory (BiLSTM) networks, and Self-Attention mechanisms to improve the accuracy and efficiency of deformation analysis. The CNN model extracts multi-scale features from time series, providing comprehensive insights into deformation. The BiLSTM networks capture temporal dependencies, enhancing predictive capabilities. The self-attention mechanism focuses on relevant spatial relationships, improving interpretability. Experimental results demonstrate that the CNN-BiLSTM model reduces RMSE by approximately 26.1% compared with the CNN-LSTM model, decreasing from 0.138 mm to 0.102 mm. By enabling earlier and more reliable detection of anomalous deformation, the proposed method supports timely maintenance and risk-informed decision-making for dams, dikes and other critical water infrastructure, thereby contributing to their long-term safe operation and to the sustainability of the water and energy services they provide. Improved accuracy also reduces false alarms and unnecessary field surveys, lowering the resource and energy costs of continuous monitoring. Consequently, these findings contribute to advancing GNSS deformation monitoring for hydraulic structures and to sustainable, resilient infrastructure management, providing valuable insights for engineering decision-making and risk management.
No takes yet. Share an insight, caveat, or question.
He et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: