Jing Lu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jing Lu has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
- Min Hu
SILC Business School, Shanghai University, Shanghai 201800, China
3 shared publications Min Hu
SHU-SUCG Research Centre for Building Industrialization, Shanghai University, Shanghai, China
2 shared publicationsWei Xu
SHU-SUCG Research Centre for Building Industrialization, Shanghai University, Shanghai, China
2 shared publicationsWenBo Zhou
Shanghai Tunnel Engineering Co., Ltd., Shanghai, China
2 shared publicationsZhaoYu Wu
Shanghai Tunnel Engineering Co., Ltd., Shanghai, China
2 shared publications- Bingjian Wu
SHU-SUCG Research Center for Building Industrialization, Shanghai University, Shanghai 200072, China
1 shared publication
Latest Publications
- Open Access
- 17 December 2025
- Published by MDPI AG in Smart Cities
- Vol. 8 (6) , 212
Bridging the Information Gap in Smart Construction: An LLM-Based Assistant for Autonomous TBM Tunneling
The development of autonomous tunneling is crucial for building the intelligent underground infrastructure that smart cities require. However, in complex urban environments, the need for frequent manual intervention during Tunnel Boring Machine (TBM) operation remains a challenge, hindering overall efficiency and safety. To address the human–machine collaboration gap, this study analyzes practical experiences from six tunnel projects that use autonomous driving systems. Building on this foundation, we develop an intelligent assistant powered by a large language model (LLM). The assistant constructs a complete service architecture and intervention mechanism, proposes a phased intention recognition framework, and uses conversational interaction to achieve efficient human–machine communication. Experimental results demonstrate the strong classification performance of our intention recognition model. Furthermore, engineering case studies validate the assistant’s effectiveness in enhancing operational transparency, increasing user trust, bridging the human–machine information gap, and ultimately ensuring safer and more reliable tunneling. This research provides a feasible and innovative technological path for human–machine collaboration in the construction of critical urban infrastructure.
- 17 December 2025
- Published by MDPI AG in Smart Cities
- Vol. 8 (6) , 212
Open AccessBridging the Information Gap in Smart Construction: An LLM-Based Assistant for Autonomous TBM Tunneling
The development of autonomous tunneling is crucial for building the intelligent underground infrastructure that smart cities require. However, in complex urban environments, the need for frequent manual intervention during Tunnel Boring Machine (TBM) operation remains a challenge, hindering overall efficiency and safety. To address the human–machine collaboration gap, this study analyzes practical experiences from six tunnel projects that use autonomous driving systems. Building on this foundation, we develop an intelligent assistant powered by a large language model (LLM). The assistant constructs a complete service architecture and intervention mechanism, proposes a phased intention recognition framework, and uses conversational interaction to achieve efficient human–machine communication. Experimental results demonstrate the strong classification performance of our intention recognition model. Furthermore, engineering case studies validate the assistant’s effectiveness in enhancing operational transparency, increasing user trust, bridging the human–machine information gap, and ultimately ensuring safer and more reliable tunneling. This research provides a feasible and innovative technological path for human–machine collaboration in the construction of critical urban infrastructure.
- Open Access
- 27 November 2025
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 12 (1)
Toward safe and efficient shield tunneling: counterfactual reinforcement learning based multi-subsystem collaborative optimization
Shield tunneling must satisfy safety requirements while maintaining efficiency, but the dynamic interactions among subsystems and continuously evolving ground conditions make manual coordination unreliable. To address this challenge, this study proposes the Counterfactual Reinforcement Learning-Based Shield Multi-Subsystem Collaborative Optimization (CRL-SMSCO), a multi-agent reinforcement learning framework with a centralized critic and decentralized actors. CRL-SMSCO performs counterfactual credit assignment to quantify each subsystem’s marginal contribution to global tunneling outcomes and jointly optimizes subsystem parameters. Safety is enforced by a safety-oriented action masking unit that restricts the feasible action space in real time, together with a hierarchical reward that prioritizes safety over efficiency. These features enable CRL-SMSCO to achieve interpretable subsystem coordination and rigorous safety enforcement. In the Nanjing Metro case study, experiments show that CRL-SMSCO improves average training reward by 3.9% over multi-agent deep deterministic policy gradient (MADDPG) and by 3.5% over multi-agent proximal policy optimization (MAPPO), while also yielding lower variance and higher minimum rewards on the held-out test segment. In the engineering application, relative to experience-based manual operation under similar geology, CRL-SMSCO increases average tunneling speed by 13.2%, reduces equipment loads over 1.7%, and decreases ground deformation by 82%, with all indicators maintained within permissible limits. These results demonstrate that CRL-SMSCO provides significant practical value in shield tunneling, offering an effective framework for managing other safety-critical coupled systems.
- 27 November 2025
- Published by Springer Nature in Complex & Intelligent Systems
- Vol. 12 (1)
Open AccessToward safe and efficient shield tunneling: counterfactual reinforcement learning based multi-subsystem collaborative optimization
Shield tunneling must satisfy safety requirements while maintaining efficiency, but the dynamic interactions among subsystems and continuously evolving ground conditions make manual coordination unreliable. To address this challenge, this study proposes the Counterfactual Reinforcement Learning-Based Shield Multi-Subsystem Collaborative Optimization (CRL-SMSCO), a multi-agent reinforcement learning framework with a centralized critic and decentralized actors. CRL-SMSCO performs counterfactual credit assignment to quantify each subsystem’s marginal contribution to global tunneling outcomes and jointly optimizes subsystem parameters. Safety is enforced by a safety-oriented action masking unit that restricts the feasible action space in real time, together with a hierarchical reward that prioritizes safety over efficiency. These features enable CRL-SMSCO to achieve interpretable subsystem coordination and rigorous safety enforcement. In the Nanjing Metro case study, experiments show that CRL-SMSCO improves average training reward by 3.9% over multi-agent deep deterministic policy gradient (MADDPG) and by 3.5% over multi-agent proximal policy optimization (MAPPO), while also yielding lower variance and higher minimum rewards on the held-out test segment. In the engineering application, relative to experience-based manual operation under similar geology, CRL-SMSCO increases average tunneling speed by 13.2%, reduces equipment loads over 1.7%, and decreases ground deformation by 82%, with all indicators maintained within permissible limits. These results demonstrate that CRL-SMSCO provides significant practical value in shield tunneling, offering an effective framework for managing other safety-critical coupled systems.
- Open Access
- 27 December 2024
- Published by MDPI AG in Smart Cities
- Vol. 8 (1) , 6
Mechanism-Driven Intelligent Settlement Prediction for Shield Tunneling Through Areas Without Ground Monitoring
Ground settlement is a crucial indicator for assessing the safety of shield tunneling and its impact on the surrounding environment. However, most existing settlement prediction methods are based on historical data, which can only be applied with effective monitoring conditions. To overcome this limitation, this paper proposes the mechanism-driven intelligent settlement prediction method (MISPM), which considers the mechanisms of settlement and attitude movements during construction to design new features that can indirectly reflect settlement. Simulation experiments were used to compare the impact of different candidate features and algorithms on prediction performance, verifying the validity and accuracy of the model. The efficacy of MISPM in predicting settlement changes in advance was substantiated by practical engineering applications. Results showed that MISPM could accurately predict settlement changes even without ground monitoring, thereby corroborating its reliability and applicability in supporting safe tunneling in complex geological environments. In the construction of urban infrastructure, this method has the potential to enhance the efficiency of tunnel construction and ensure environmental safety, which is of great significance for the development of smart cities.
- 27 December 2024
- Published by MDPI AG in Smart Cities
- Vol. 8 (1) , 6
Open AccessMechanism-Driven Intelligent Settlement Prediction for Shield Tunneling Through Areas Without Ground Monitoring
Ground settlement is a crucial indicator for assessing the safety of shield tunneling and its impact on the surrounding environment. However, most existing settlement prediction methods are based on historical data, which can only be applied with effective monitoring conditions. To overcome this limitation, this paper proposes the mechanism-driven intelligent settlement prediction method (MISPM), which considers the mechanisms of settlement and attitude movements during construction to design new features that can indirectly reflect settlement. Simulation experiments were used to compare the impact of different candidate features and algorithms on prediction performance, verifying the validity and accuracy of the model. The efficacy of MISPM in predicting settlement changes in advance was substantiated by practical engineering applications. Results showed that MISPM could accurately predict settlement changes even without ground monitoring, thereby corroborating its reliability and applicability in supporting safe tunneling in complex geological environments. In the construction of urban infrastructure, this method has the potential to enhance the efficiency of tunnel construction and ensure environmental safety, which is of great significance for the development of smart cities.
- Open Access
- 31 January 2023
- Published by Springer Nature in Scientific Reports
- Vol. 13 (1) , 1-17
A multistage model for rapid identification of geological features in shield tunnelling
Decision-making on shield construction parameters depends on timely and accurate geological condition feedback. Real-time mastering of geological condition around the shield during tunnelling is necessary to achieve safe and efficient construction. This paper proposes a Rapidly Geological Features Identification (RGFI) method that balances the model's generalizability and the accuracy of geological identification. First, a k-means algorithm is used to redefine the stratum based on the key mechanical indexes of strata. An XGBoost model is then used to determine the stratum composition of the excavation face based on the tunnelling parameters. If the result is compound strata, a deep neural network with an attention mechanism is used to predict the percentage of each stratum. The attention mechanism assigns weights to the features of the tunnelling parameters according to the stratum composition. The simulation results in the interval between Qian-Zhuang and Ke-Ning Road of Nanjing Metro show that the method can effectively determine the geological conditions on the excavation face. Furthermore, the method was used in the Hangzhou-Shaoxing intercity railroad tunnel project, where the 'ZhiYu' self-driving shield was used for tunnelling control. It helped the 'ZhiYu' shield to adjust the construction parameters quickly and improve the safety and quality of the project.
- 31 January 2023
- Published by Springer Nature in Scientific Reports
- Vol. 13 (1) , 1-17
Open AccessA multistage model for rapid identification of geological features in shield tunnelling
Decision-making on shield construction parameters depends on timely and accurate geological condition feedback. Real-time mastering of geological condition around the shield during tunnelling is necessary to achieve safe and efficient construction. This paper proposes a Rapidly Geological Features Identification (RGFI) method that balances the model's generalizability and the accuracy of geological identification. First, a k-means algorithm is used to redefine the stratum based on the key mechanical indexes of strata. An XGBoost model is then used to determine the stratum composition of the excavation face based on the tunnelling parameters. If the result is compound strata, a deep neural network with an attention mechanism is used to predict the percentage of each stratum. The attention mechanism assigns weights to the features of the tunnelling parameters according to the stratum composition. The simulation results in the interval between Qian-Zhuang and Ke-Ning Road of Nanjing Metro show that the method can effectively determine the geological conditions on the excavation face. Furthermore, the method was used in the Hangzhou-Shaoxing intercity railroad tunnel project, where the 'ZhiYu' self-driving shield was used for tunnelling control. It helped the 'ZhiYu' shield to adjust the construction parameters quickly and improve the safety and quality of the project.
- Open Access
- 24 June 2022
- Published by Springer Nature in Research Square
A Multi-stage Model for Rapid Identification of Geological Features in Shield Tunnelling
Decision-making on shield construction parameters depends on timely and accurate geological properties feedback. Real-time mastering of geological status around the shield during tunneling is necessary to achieve safe and efficient construction. However, balancing the model generalizability and the accuracy of geological identification remains a significant challenge. This paper proposes a Rapidly Geological Features Identification (RGFI) method that predicts the percentage of each stratum in the excavation face under different geological environments. The method consists of three stages: k-means-based geological type redefinition, eXtreme Gradient Boosting based geological type classification and convolutional neural network stratum distribution prediction with an attention mechanism. The method produces good generalization performance and achieves higher accuracy when simulating tunnel construction data for self-driving shield in the interval between Qian-Zhuang and Ke-Ning Road of Line 5 of Nanjing Metro in China. Furthermore, the method successfully provided more accurate geological information for the Hangzhou-Shaoxing intercity railroad tunnel project. It helped the ‘ZhiYu’ shield to adjust the construction parameters quickly and improve the safety and quality of the project.
- 24 June 2022
- Published by Springer Nature in Research Square
Open AccessA Multi-stage Model for Rapid Identification of Geological Features in Shield Tunnelling
Decision-making on shield construction parameters depends on timely and accurate geological properties feedback. Real-time mastering of geological status around the shield during tunneling is necessary to achieve safe and efficient construction. However, balancing the model generalizability and the accuracy of geological identification remains a significant challenge. This paper proposes a Rapidly Geological Features Identification (RGFI) method that predicts the percentage of each stratum in the excavation face under different geological environments. The method consists of three stages: k-means-based geological type redefinition, eXtreme Gradient Boosting based geological type classification and convolutional neural network stratum distribution prediction with an attention mechanism. The method produces good generalization performance and achieves higher accuracy when simulating tunnel construction data for self-driving shield in the interval between Qian-Zhuang and Ke-Ning Road of Line 5 of Nanjing Metro in China. Furthermore, the method successfully provided more accurate geological information for the Hangzhou-Shaoxing intercity railroad tunnel project. It helped the ‘ZhiYu’ shield to adjust the construction parameters quickly and improve the safety and quality of the project.