Research Article
Prediction for Surface Subsidence of Shield Construction in Water-Rich Sand Egg Stratum Based on Edge Intelligence
@INPROCEEDINGS{10.1007/978-3-031-28990-3_14, author={Yanxia Gao and Yiwen Liu and Chunqiao Mi and Pengju Tang and Yuanquan Shi}, title={Prediction for Surface Subsidence of Shield Construction in Water-Rich Sand Egg Stratum Based on Edge Intelligence}, proceedings={Edge Computing and IoT: Systems, Management and Security. Third EAI International Conference, ICECI 2022, Virtual Event, December 13-14, 2022, Proceedings}, proceedings_a={ICECI}, year={2023}, month={3}, keywords={Surface deformation Upper soft and lower hard soil layer Curved shield construction Mindlin solution Random medium theory}, doi={10.1007/978-3-031-28990-3_14} }- Yanxia Gao
Yiwen Liu
Chunqiao Mi
Pengju Tang
Yuanquan Shi
Year: 2023
Prediction for Surface Subsidence of Shield Construction in Water-Rich Sand Egg Stratum Based on Edge Intelligence
ICECI
Springer
DOI: 10.1007/978-3-031-28990-3_14
Abstract
The refinement and intelligent control of shield tunneling is the development trend of modern tunnel construction technology. In order to better predict and control the surface subsidence caused by shield excavation, this paper takes the shield construction of Luoyang Metro Line 2 from Longmen Station to Longmen Avenue Station as the background, and proposes a method based on edge intelligence for shield construction in water-rich sand egg strata. Methods for predicting land subsidence. First, low latency and faster data processing are achieved by collecting a large amount of data containing dynamic information about geological conditions and surrounding environments during the shield tunneling process; then using the iFogSim simulator to create different configurations; second, establishing support A surface subsidence model based on vector regression was established, and the model was deployed on the edge equipment; finally, the model was evaluated using the monitoring data of surface subsidence of the water-rich sand egg formation in Luoyang area. The research results show that the edge computing-based system has lower latency and higher processing speed than only deploying cloud data centers. After Pearson-related parameter tuning and model comparison training with Linear as the kernel function, the mean square error of the predicted value and the collected value of the surface subsidence is better than the other two kernel functions. The method proposed in this paper can provide real-time prediction service for large-scale surface subsidence prediction caused by shield construction, and is more practical.
