Risk assessment of existing buildings in tunnel construction based on an improved cumulative prospect theory method DOI Creative Commons

Laura Harrell ve Eugene B. Chang,

Nurhuda Nordin,

Xinhua Gu

et al.

Science Progress, Journal Year: 2025, Volume and Issue: 108(1)

Published: Jan. 1, 2025

As urbanization in China continues to rise, an increasing number of cities are constructing subway tunnels. However, due the complexity and ambiguity tunnel construction, there is a lack precise methods assess impact these constructions on surrounding buildings. Consequently, this study analyzes summarizes past experiences proposes IVTSFS-CPT-EDAS model based CPT-EDAS evaluation method. This establishes risk assessment approach specifically for construction existing The model's process was validated through real-world case study, including sensitivity analysis verify its effectiveness feasibility. findings indicate: (1) can more comprehensively delicately replicate actual decision-making environment, enhancing accuracy model. (2) expert evaluations indicates that improper material equipment configuration, inadequate excavation pressure control, non-compliance stratum solubility coefficient with requirements primary factors affecting building. (3) advantages proposed over other approaches enhancement results improvements method were demonstrated comparative evaluation. research expected provide valuable insights scientific management impacts nearby structures.

Language: Английский

Deep learning technologies for shield tunneling: Challenges and opportunities DOI
Cheng Zhou,

Yuyue Gao,

Elton J. Chen

et al.

Automation in Construction, Journal Year: 2023, Volume and Issue: 154, P. 104982 - 104982

Published: June 27, 2023

Language: Английский

Citations

60
A data-driven method to model stress-strain behaviour of frozen soil considering uncertainty DOI

Kai-Qi Li,

Zhen‐Yu Yin, Ning Zhang

et al.

Cold Regions Science and Technology, Journal Year: 2023, Volume and Issue: 213, P. 103906 - 103906

Published: May 31, 2023

Language: Английский

Citations

48
Success and challenges in predicting TBM penetration rate using recurrent neural networks DOI
Feng Shan, Xuzhen He, Danial Jahed Armaghani

et al.

Tunnelling and Underground Space Technology, Journal Year: 2022, Volume and Issue: 130, P. 104728 - 104728

Published: Sept. 1, 2022

Language: Английский

Citations

69
Safety prediction of shield tunnel construction using deep belief network and whale optimization algorithm DOI

Shuangshuang Ge,

Wei Gao, Shuang Cui

et al.

Automation in Construction, Journal Year: 2022, Volume and Issue: 142, P. 104488 - 104488

Published: July 19, 2022

Language: Английский

Citations

49
Effects of data smoothing and recurrent neural network (RNN) algorithms for real-time forecasting of tunnel boring machine (TBM) performance DOI Creative Commons
Feng Shan, Xuzhen He, Danial Jahed Armaghani

et al.

Journal of Rock Mechanics and Geotechnical Engineering, Journal Year: 2023, Volume and Issue: 16(5), P. 1538 - 1551

Published: Sept. 4, 2023

Tunnel boring machines (TBMs) have been widely utilised in tunnel construction due to their high efficiency and reliability. Accurately predicting TBM performance can improve project time management, cost control, risk management. This study aims use deep learning develop real-time models for the penetration rate (PR). The are built using data from Changsha metro project, performances evaluated unseen Zhengzhou Metro project. In one-step forecast, predicted follows trend of measured both training testing. autoregressive integrated moving average (ARIMA) model is compared with recurrent neural network (RNN) model. results show that univariate models, which only consider historical itself, perform better than multivariate take into account multiple geological operational parameters (GEO OP). Next, an RNN variant combining series last-step developed, it performs other models. A sensitivity analysis shows most important parameter, while a smaller impact on forecasting. It also found smoothed easier predict accuracy. Nevertheless, over-simplified lose real characteristics series. conclusion, accurately next-step rate, smoothing crucial provides practical guidance forecasting engineering.

Language: Английский

Citations

33
TBM performance prediction using LSTM-based hybrid neural network model: Case study of Baimang River tunnel project in Shenzhen, China DOI Creative Commons
Qihang Xu, Xin Huang, Baogang Zhang

et al.

Underground Space, Journal Year: 2023, Volume and Issue: 11, P. 130 - 152

Published: April 11, 2023

Accurately predicting tunnel boring machine (TBM) performance is beneficial for excavation efficiency enhancement and risk mitigation of TBM tunneling. In this paper, we develop a long short-term memory (LSTM) based hybrid intelligent model to predict two key parameters (advance rate cutterhead torque). The combines the LSTM, BN, Dropout Dense layers process raw data improve fitting quality. features, including ground formation properties, route curvature, location operational parameters, are divided into historical/real-time time-varying time-invariant output prediction data. effectiveness proposed verified on large monitoring database Baimang River Tunnel Project in Shenzhen, south China. We then discuss influence mode, neural network structure time division interval length historical accuracy. significance evaluation input features shows that has largest accuracy, properties secondary. It also found correlations between coincident with their interrelationships ease excavation. Finally, it results most affected by total propulsion force followed rotation speed cutterhead. established can provide useful guidance construction personnel roughly grasp possible status from when adjusting parameters.

Language: Английский

Citations

24
A spatiotemporal deep learning method for excavation-induced wall deflections DOI Creative Commons
Yuanqin Tao, Shaoxiang Zeng, Honglei Sun

et al.

Journal of Rock Mechanics and Geotechnical Engineering, Journal Year: 2024, Volume and Issue: 16(8), P. 3327 - 3338

Published: Jan. 29, 2024

Data-driven approaches such as neural networks are increasingly used for deep excavations due to the growing amount of available monitoring data in practical projects. However, most network models only use from a single point and neglect spatial relationships between multiple points. Besides, lack flexibility providing predictions days after activity. This study proposes sequence-to-sequence (seq2seq) two-dimensional (2D) convolutional long short-term memory (S2SCL2D) predicting spatiotemporal wall deflections induced by excavations. The model utilizes all points on entire extracts features combining 2D layers (LSTM) layers. S2SCL2D achieves long-term prediction through recursive seq2seq structure. excavation depth, which has significant impact deflections, is also considered using feature fusion method. An project Hangzhou, China, illustrate proposed model. results demonstrate that superior accuracy robustness than LSTM S2SCL1D (one-dimensional) models. demonstrates strong generalizability when applied an adjacent excavation. Based results, practitioners can plan allocate resources advance address potential engineering issues.

Language: Английский

Citations

14
Adaptive mutation sparrow search algorithm-Elman-AdaBoost model for predicting the deformation of subway tunnels DOI Creative Commons
Zhou Xiang-zhen, Wei Hu,

Zhongyong Zhang

et al.

Underground Space, Journal Year: 2024, Volume and Issue: 17, P. 320 - 360

Published: March 1, 2024

A novel coupled model integrating Elman-AdaBoost with adaptive mutation sparrow search algorithm (AM-SSA), called AMSSA-Elman-AdaBoost, is proposed for predicting the existing metro tunnel deformation induced by adjacent deep excavations in soft ground. The novelty that modified SSA proposes adjustment strategy to create a balance between capacity of exploitation and exploration. In AM-SSA, firstly, population initialized cat mapping chaotic sequences improve ergodicity randomness individual sparrow, enhancing global ability. Then individuals are adjusted Tent disturbance Cauchy avoid being too concentrated or scattered, expanding local Finally, producer-scrounger number formula introduced ability seek optimal. addition, it leads improved achieving better accuracy level convergence speed compared original SSA. To demonstrate effectiveness reliability 23 classical benchmark functions 25 IEEE Congress on Evolutionary Computation test (CEC2005), employed as numerical examples investigated comparison some well-known optimization algorithms. statistical results indicate promising performance AM-SSA variety constrained unknown spaces. By utilizing AdaBoost algorithm, multiple sets weak AMSSA-Elman predictor restructured into one strong successive iterations prediction output. Additionally, on-site monitoring data acquired from excavation project Ningbo, China, were selected training testing sample. Meanwhile, predictive outcomes those other different machine learning techniques. end, obtained this real-world geotechnical engineering field reveal feasibility hybrid model, illustrating its power superiority terms computational efficiency, accuracy, stability, robustness. More critically, observing real time daily basis, structural safety associated tunnels could be supervised, which enables decision-makers take concrete control protection measures.

Language: Английский

Citations

9
Prediction of shield machine attitude parameters based on decomposition and multi-head attention mechanism DOI
Qiushi Wang, Wenqi Ding, Kourosh Khoshelham

et al.

Automation in Construction, Journal Year: 2025, Volume and Issue: 171, P. 105973 - 105973

Published: Jan. 18, 2025

Language: Английский

Citations

1
Daily Streamflow Forecasting Based on the Hybrid Particle Swarm Optimization and Long Short-Term Memory Model in the Orontes Basin DOI Open Access
Hüseyin Çağan Kılınç

Water, Journal Year: 2022, Volume and Issue: 14(3), P. 490 - 490

Published: Feb. 7, 2022

Water, a renewable but limited resource, is vital for all living creatures. Increasing demand makes the sustainability of water resources crucial. River flow management, one key drivers sustainability, will be to protect communities from worst impacts on environment. Modelling and estimating river in hydrological process crucial terms effective planning, sustainable use resources. Therefore, this study, hybrid approach integrating long short-term memory networks (LSTM) particle swarm algorithm (PSO) was proposed. For purpose, three stations were utilized study along Orontes basin, Karasu, Demirköprü, Samandağ, respectively. The timespan Demirköprü Karasu between 2010 2019. Samandağ station data 2009–2018. datasets consisted daily values. In order validate performance model, first 80% used training, remaining 20% testing FMSs. Statistical methods such as linear regression more classical model autoregressive integrated moving average (ARIMA) during comparison assess proposed method’s demonstrate its superior predictive ability. estimation results models evaluated with RMSE, MAE, MAPE, SD, R2 statistical metrics. streamflow predictions revealed that PSO-LSTM provided promising accuracy presented higher compared benchmark models.

Language: Английский

Citations

38