Autopilot model for shield tunneling machines using support vector regression and its application to previously constructed tunnels

Kubota, Yoshimi; Yabuki, Nobuyoshi; Fukuda, Tomohiro

doi:10.1111/mice.13019

Cited by 5 publications

(2 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Shield tunnelling parameters are predicted by inputting only the parameters of next ring to be excavated in traditional intelligence methods, such as random forest [11,28] and support vector regression [14,23]. However, all shield parameters are time-series data recorded chronologically, and the construction state is influenced by the adjacent construction process.…”

Section: Introductionmentioning

confidence: 99%

An intelligent construction model in whole process for shield tunnelling

Lu,

Liu,

Kong

et al. 2024

Preprint

View full text Add to dashboard Cite

Predicting shield tunnelling parameters in the whole construction process is of great importance, which can effectively control ground stability and improve tunnelling efficiency. A novel deep learning method is developed considering transfer learning, incremental learning and Bi-LSTM fusing with available data of the next ring to be excavated (ADNRE) to predict shield tunnelling parameters in the whole process. Before construction, transfer learning uses data from similar projects to determine initial network parameters, then solve the insufficient data in the prophase of the project. As the shield machine begins to excavate, incremental learning is used to continuously accept new data and adjust model parameters in real time during the whole process. A feature fusion module in Bi-LSTM is proposed to integrate ADNRE and data of the adjacent excavated rings. The proposed Bi-LSTM method can consider the mutation of stratum conditions during tunnelling. The applicability of the proposed method is explored by predicting the shield cutter head torque of a tunnel project in Qingdao, China. The influence of fine-tuning epochs and project similarity on model performance is further discussed. Overall, the proposed method can provide reasonable whole process prediction for shield tunnelling parameters, which improves construction safety and efficiency.

show abstract

Section: Introductionmentioning

confidence: 99%

An intelligent construction model in whole process for shield tunnelling

Lu,

Liu,

Kong

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Fu et al [20] developed a deep learning model that integrates the graph convolutional network and LSTM model to predict deviations in the tail section of shield machines. Although existing research has already successfully predicted shield attitudes for the next 1 to 10 s, the operation and adjustment of shield machines typically require a longer timeframe [21,22]. Predictions made within 10 s or less may not provide operators with sufficient reaction time for effective adjustments.…”

Section: Introductionmentioning

confidence: 99%

An Intelligent Multi-Ring Shield Movement Performance Prediction and Control Method

Zeng,

Yu,

Chen

et al. 2024

Applied Sciences

View full text Add to dashboard Cite

Accurate control of the shield attitude can ensure precise tunnel excavation and minimize impact on the surrounding areas. However, neglecting the total thrust force may cause excessive disturbance to the strata, leading to collapse. This study proposes a Bayesian optimization-based temporal attention long short-term memory model (BOTA-LSTM) for multi-objective prediction and control of shield tunneling, including shield attitude and total thrust. The model can achieve multi-ring predictions of shield attitude and total thrust by allocating larger weights to significant moments through a temporal attention mechanism. The hyperparameters of the proposed model are automatically selected through Bayesian hyperparameter optimization, which can effectively address the issue of complex parameter selection and optimization difficulties in multi-ring, multi-objective tasks. Based on the predictive results of the optimal model, an intelligent control method that considers both shield attitude and total thrust is proposed. Compared to a method that solely predicts and corrects for the next ring, the proposed multi-ring correction method provides the opportunity for further adjustments, if the initial correction falls short of expectations. A shield tunneling project in Hangzhou is used to demonstrate the effectiveness of the proposed model. The results show that the BOTA-LSTM model outperforms the models without the integration of a temporal attention mechanism and Bayesian hyperparameter optimization. The proposed multi-ring intelligent correction method can adjust the shield attitude and total thrust to a reasonable range, providing references for practical engineering applications.

show abstract