Real-time unsupervised monitoring of earth pressure balance shield-induced sinkholes in mixed-face ground conditions via convolutional variational autoencoders

Tunnelling and Underground Space Technology, Journal Year: 2024, Volume and Issue: 152, P. 105908 - 105908

Published: June 27, 2024

This study introduces a real-time unsupervised monitoring framework for sinkhole formation events during earth pressure balance (EPB) shield tunneling operations. A feature extractor (FE) is constructed by coupling variational Autoencoders structure with convolutional neural network layers (VAE-CNN) to manage the complexity of EPB operational data, including non-linearity and temporal dependencies. The consists two main phases: offline modeling online monitoring. In phase, an FE model trained using data-intensive techniques define subspace characterizing behavior multivariate data without formations. squared prediction error (SPE) statistics control limits are computed detection. During unseen propagated generate SPE values determine based on whether these surpass limit. Sensor validity index violation counts were used isolate most influential variables, while results demonstrated superiority proposed VAE-CNN method, achieving 100% detection rate 0.9% false alarm rate. variables identified include cutter resolutions per minute, jack speed, screw pressure, torque, seal components. system shows great potential early warnings operations mitigate risks.

Language: Английский

Design of Disc Cutter Wear Test System and Research on Wear Law Based on Eddy Current Testing Technology

Rock Mechanics and Rock Engineering, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 26, 2024

Language: Английский