Optimized CNN-BiLSTM-Attention with hybrid signal denoising: a novel interpretable framework for prediction of shield tunneling advance speed

Key Points

• The proposed framework successfully predicts tunneling advance speed using advanced algorithms.
• Prediction accuracy improved by 25% with optimized signal denoising techniques in the model.
• The analysis leverages a combination of CNN and BiLSTM architectures with attention mechanisms for detailed insights.
• Significance lies in the model's interpretability, allowing for better understanding and adjustments in tunneling operations.