Author(s)

YongKun Li¹, Yin Bo2*

Affiliation(s)

¹China Railway Tunnel Bureau Group Co., Ltd., Guangzhou 5111458, Guangdong, China.

²Changjiang Survey, Planning, Design and Research Co., Ltd., Wuhan 430010, Hubei, China.

Corresponding Author

Yin Bo

ABSTRACT

Tunnel Boring Machines (TBMs) are widely used in subway and tunnel projects due to their efficiency and safety. However, the complexity and high uncertainty of the construction environment lead to lag and limitations in traditional risk identification methods. Addressing the urgent need for intelligent management, this study combines the advantages of Knowledge Graphs (KG) in structured knowledge representation and Large Language Models (LLMs) in natural language understanding to construct a synergistically driven intelligent risk identification and decision support system for TBM construction.First, the study employs ontology modeling to build a risk knowledge graph covering risk factors, risk-causing mechanisms, and consequences, achieving entity-relation extraction through multi-source heterogeneous data mining. Second, the Large Language Model is adapted and fine-tuned by constructing an engineering corpus and injecting domain knowledge. Finally, a synergistic mechanism is designed to realize intelligent risk identification and decision-making.The innovations of this study include: (1) Proposing a "Knowledge Retrieval - Semantic Understanding - Logical Reasoning" three-layer architecture, effectively fusing explicit knowledge with reasoning capabilities; (2) Developing a multi-modal risk feature fusion identification model, improving identification accuracy and real-time performance; and (3) Building an intelligent recommendation engine for risk response plans, supporting multi-scenario simulation. Experimental and engineering application results indicate that the system significantly improves the precision and recall of risk identification, effectively enhancing the risk management level of TBM construction.

KEYWORDS

TBM construction; Risk management; Knowledge graph; Large language model; Intelligent decision support

CITE THIS PAPER

YongKun Li, Yin Bo. Intelligent identification and decision support system for TBM construction risk synergistically driven by knowledge graphs and large language models. World Journal of Engineering Research. 2026, 4(1): 1-18. DOI: https://doi.org/10.61784/wjer3073.

REFERENCES

[1] Hong Kairong, Du Yanliang, Chen Kui, et al. Development history, achievements, and prospects of full-face tunnel boring machine in China. Tunnel Construction, 2022, 42(5): 739-756.

[2] Wang Mengshu. Present status, existing problems and development countermeasures of shield and TBM tunneling technology in China. Tunnel Construction, 2014, 34(3): 179-187.

[3] Wang Yafeng. Key technologies for TBM jamming release of Gaoligongshan tunnel under unfavorable geologies. Tunnel Construction, 2021, 41(3): 441-448.

[4] Hasanpour R, Rostami J, Schmitt J, et al. Prediction of TBM jamming risk in squeezing grounds using bayesian and artificial neural networks. Journal of Rock Mechanics and Geotechnical Engineering, 2020, 12(1): 21-31.

[5] Zhu Qing, Wang Suozhi, Ding Yulin, et al. A method of safety-quality-schedule knowledge graph for intelligent management of drilling and blasting construction of railway tunnels. Geomatics and Information Science of Wuhan University, 2022, 47(8): 1155-1164.

[6] Park B, Lee C, Choi S-W, et al. Discrete-Element Analysis of the Excavation Performance of an EPB Shield TBM under Different Operating Conditions. Applied Sciences, 2021, 11(11): 5119.

[7] Chen J, Yang T, Zhang D, et al. Deep learning based classification of rock structure of tunnel face. Geoscience Frontiers, 2021, 12(1): 395-404.

[8] Kuang G, Li B, Hu X, et al. Review on machine learning-based defect detection of shield tunnel lining. Periodica Polytechnica Civil Engineering, 2022, 66(3): 943-957.

[9] Chen Xiangsheng, Zeng Shiqi, Han Wenlong, et al. Review and prospect of machine learning method in shield tunnel construction. Journal of Civil and Environmental Engineering, 2024, 46(1): 1-14.

[10] Sonmez H, Gokceoglu C, Nefeslioglu H A, et al. Estimation of rock modulus: For intact rocks with an artificial neural network and for rock masses with a new empirical equation. International Journal of Rock Mechanics and Mining Sciences, 2006, 43(2): 224-235.

[11] Gokceoglu C, Bal C, Aladag C H. Modeling of tunnel boring machine performance employing random forest algorithm. Geotechnical and Geological Engineering, 2023, 41(7): 4205-4226.

[12] Kilic K, Narihiro O, Ikeda H, et al. Soft ground micro TBM jack speed and torque prediction using machine learning models through operator data and micro TBM-log data synchronization. Scientific Reports, 2024, 14: 9728.

[13] Moghtader T, Sharafati A, Naderpour H, et al. Estimating maximum surface settlement caused by EPB shield tunneling utilizing an intelligent approach. Buildings, 2023, 13(4): 1051.

[14] Njock PG, Yin ZY, Xu HR, et al. Structural failure risk assessment of shield tunnel using large language model. Tunnelling and Underground Space Technology, 2025, 165: 106882.

[15] Lee J, Ahn S, Kim D, et al. Performance comparison of retrieval-augmented generation and fine-tuned large language models for construction safety management knowledge retrieval. Automation in Construction, 2024, 168: 105846.

[16] Clarke J, Laefer D. Evaluation of risk assessment procedures for buildings adjacent to tunnelling works. Tunnelling and Underground Space Technology, 2014, 40: 333-342.

[17] Sharafat A, Latif K, Seo J. Risk analysis of TBM tunneling projects based on generic bow-tie risk analysis approach in difficult ground conditions. Tunnelling and Underground Space Technology, 2021, 111: 103860.

[18] Yazdani-Chamzini A. Proposing a new methodology based on fuzzy logic for tunnelling risk assessment. Journal of Civil Engineering and Management, 2014, 20(1): 82-94.