Dynamic Interactional And Cooperative Network For Shield Machine
Dazhi Gao, Rongyang Li, Hongbo Wang, Lingfeng Mao, Huansheng Ning
TL;DR
This work addresses the limitations of manual shield machine monitoring by integrating geological information into a dynamic interactional network that links the shield machine, geology, and control terminal. It introduces two task-focused models: a CNN-based rate-prediction system with attention and residuals (augmented by $Smooth L_{1}$ loss) and a VAE–LSTM anomaly-detection framework, both leveraging geology-excavation data. Empirical results show a rate-prediction $R^2$ of 92.2% with $MSE$ of 0.0065, and anomaly-detection accuracy of 98.2%, outperforming several baselines. The findings highlight the practical value of geology-aware modeling for robust, intelligent SM operation and pave the way for more integrated, data-driven tunnel construction systems.
Abstract
The shield machine (SM) is a complex mechanical device used for tunneling. However, the monitoring and deciding were mainly done by artificial experience during traditional construction, which brought some limitations, such as hidden mechanical failures, human operator error, and sensor anomalies. To deal with these challenges, many scholars have studied SM intelligent methods. Most of these methods only take SM into account but do not consider the SM operating environment. So, this paper discussed the relationship among SM, geological information, and control terminals. Then, according to the relationship, models were established for the control terminal, including SM rate prediction and SM anomaly detection. The experimental results show that compared with baseline models, the proposed models in this paper perform better. In the proposed model, the R2 and MSE of rate prediction can reach 92.2\%, and 0.0064 respectively. The abnormal detection rate of anomaly detection is up to 98.2\%.