Shield construction intelligent management system and method based on convergence of multiple production elements

Abstract

The invention relates to the technical field of shield engineering, in particular to the technical field of civil engineering and shield engineering of a shield construction intelligent management system based on multi-production-factor convergence, and particularly relates to a shield construction intelligent management and control system and method based on multi-production-factor convergence. The objective of the invention is to solve the problems of decision delay and resource waste caused by dependence on an isolated subsystem in the prior art. The system sequentially comprises a multi-production-factor sensing unit, a heterogeneous data fusion central unit, an intelligent decision engine unit, a control instruction execution unit and a visual mutual feedback unit in the data processing direction. The system has the advantages that information barriers of all links of shield construction are broken through, full-process closed-loop management from data acquisition, intelligent analysis to precise management and control is achieved, the safety, efficiency and intelligent level of shield construction are remarkably improved, and the system is suitable for shield tunnel engineering under various geological conditions.

Description

Technical Field

[0001] This invention relates to the field of intelligent management systems for tunnel boring machine (TBM) construction based on the convergence of multiple production factors, and in particular to an intelligent management system for TBM construction based on the convergence of multiple production factors. Background Technology

[0002] Currently, the tunnel boring machine (TBM) construction process suffers from problems such as scattered data on multiple production factors, including personnel, equipment, and environment, and low monitoring efficiency. Data sharing still faces serious barriers. Traditional TBM management and control systems often rely on isolated subsystems, leading to decision-making delays, resource waste, and safety risks, which in turn affect the overall management level and make it difficult to achieve efficient project management and control.

[0003] Current tunnel boring machine (TBM) construction management systems have evolved beyond simple data monitoring into systems integrating real-time monitoring, intelligent decision-making, process collaboration, and risk prevention and control. Their management functions primarily include: real-time parameter monitoring and intelligent early warning, precise attitude and axis control, and auxiliary decision-making during tunneling. The current system's advanced nature relies heavily on the integration of multiple technologies: BIM+GIS (Geographic Information System): Constructing a 3D visualized digital model of the tunnel and its surrounding environment, enabling unified management of macro-geography and micro-structure. Digital Twin: Creating a virtual model fully synchronized with the physical tunnel boring machine, which can be used for simulation, scenario planning, and fault prediction, achieving true preventative maintenance. 5G and IoT: Ensuring ultra-low latency and high reliability transmission of massive amounts of data, providing the possibility for remote real-time control.

[0004] Despite significant progress, the system still faces challenges such as incomplete data silos, low model accuracy, slow data update speed, and decision-making delays. Summary of the Invention

[0005] The purpose of this invention is to solve the problems of decision-making delays and resource waste caused by relying on isolated subsystems in existing technologies. Based on a shield tunneling construction intelligent management system that integrates multiple production factors, this invention addresses the problems of dispersed production factors, poor information exchange, and low level of intelligence in current shield tunneling construction management. It proposes a shield tunneling construction intelligent management and control system and method based on the integration of multiple production factors. Through data aggregation and fusion and artificial intelligence technology, it achieves intelligent management and control of the entire shield tunneling construction process, thereby improving the overall efficiency of the project.

[0006] The specific solution of this invention is: Design an intelligent management system for tunnel boring machine (TBM) construction based on the convergence of multiple production factors. Along the data processing direction, it sequentially includes a multi-production factor sensing unit, a heterogeneous data fusion hub unit, an intelligent decision engine unit, a control command execution unit, and a visualization feedback unit, as detailed below: Multi-production factor sensing unit: includes multiple IoT nodes. Each IoT node is used to collect raw data of production factors such as construction personnel, equipment, materials, environment, and processes in real time. The edge computing terminal performs preliminary preprocessing on the collected raw data and then transmits the preprocessed data to the heterogeneous data fusion hub unit. The heterogeneous data fusion hub unit adopts a federated learning framework and consists of multiple data processing nodes and a model aggregation node. The data processing nodes are distributed near each edge computing terminal and perform screening on the pre-processed data transmitted from the edge computing terminal. The processed data is then transmitted to the intelligent decision engine unit. The intelligent decision engine unit includes a data storage module, an algorithm model library, and a decision analysis module; The data storage module uses a distributed database to store the fused data transmitted from the heterogeneous data fusion hub, and the storage capacity can be expanded according to actual needs. The algorithm model library contains a variety of deep learning models, including recurrent neural networks (RNN) for construction progress prediction, convolutional neural networks (CNN) for quality defect identification, long short-term memory networks (LSTM) for safety risk prediction, and deep learning models based on attention mechanisms for dynamic cost accounting. The decision analysis module calls the corresponding models in the algorithm model library to perform data mining and analysis on the fused data. The analysis mainly includes: construction progress analysis, quality defect analysis, safety risk analysis, and cost accounting analysis. The control command execution unit includes a command generation module and a command issuance module. The command generation module generates a distributed adaptive control scheme based on the analysis results and decision suggestions output by the intelligent decision engine. The command issuance module issues the generated control scheme to the field execution system through 5G communication technology. The command issuance module tracks and provides feedback on the execution status of the commands in real time. The visualization feedback unit includes a data visualization module and a human-computer interaction module. The data visualization module uses 3D modeling technology and big data visualization tools to present the control information output by the intelligent decision engine in the form of 3D figures, charts, and curves, constructing a full-element control dashboard. The human-computer interaction module supports managers in configuring parameters and intervening with commands. Managers can set various parameters of the system or directly issue control commands through input devices such as touch screens, mice, and keyboards. The human-computer interaction module also has data query and report generation functions, allowing managers to query historical data and analysis results, and generate various statistical reports. Finally, the feedback effect is transmitted to the multi-production factor sensing unit, which then re-collects and processes the feedback data, entering the next control cycle to form a closed-loop feedback.

[0007] The screening process includes: firstly, data denoising, which uses wavelet transform algorithm to remove noise from the data and improve the signal-to-noise ratio; secondly, data normalization, which converts data of different types and magnitudes into a unified numerical range to facilitate subsequent data analysis and model training; and finally, a dynamically updated construction digital twin model is constructed based on the processed data.

[0008] The control plan includes instructions for dispatching construction personnel, instructions for adjusting the parameters of tunnel boring machines, instructions for allocating material supplies, and instructions for optimizing construction procedures.

[0009] The decision analysis module calls the corresponding models from the algorithm model library, which are derived from historical project data, historical design data, and pre-design analysis data. After screening out the construction differences with historical project data, construction differences with historical design data, and differences with the mechanical data of each force point in the pre-design analysis, it forms mining analysis points and analyzes the reasons for the differences to ensure safety.

[0010] It also includes a production factor sensing unit installed on the site adjustment physical component, and the production factor sensing unit is connected in parallel to the transmission path in the data processing direction to form a parallel loop of the production factor sensing unit, so as to introduce data change information after the adjustment component in real time.

[0011] For construction progress, an RNN model is used to analyze historical progress data and current construction parameters to predict future construction progress and compare it with the planned progress, thus providing early warning of construction progress deviations. For quality defects, a CNN model is used to analyze image and sensor data collected during construction to identify potential quality defects and trace their causes. For safety risks, an LSTM model is used to analyze environmental monitoring data, equipment operation data, and personnel behavior data to predict potential safety risks, such as gas leaks, equipment failures, and personnel entering dangerous areas. For cost accounting, a deep learning model based on an attention mechanism analyzes material consumption data, equipment energy consumption data, and personnel work hours data to achieve dynamic cost accounting and prediction, and generate decision-making suggestions.

[0012] The dynamic cost accounting introduces a dynamic cost accounting formula.

[0013] This invention provides an intelligent management and control method for tunnel boring machine (TBM) construction based on the convergence of multiple production factors, comprising the following steps: S10: The sensing unit collects data from multiple production factors in real time, and the edge terminal realizes data aggregation and preprocessing; S20: The data fusion center further processes the received data and constructs a dynamically updated twin model; S30: Conduct intelligent decision analysis based on fused data and twin models; S40: Based on intelligent decision analysis suggestions, generate and execute a distributed adaptive control scheme; S50: The results after execution are presented after visualization processing, and manual intervention and control can be performed during the process.

[0014] In specific implementation, in step S10, data from the sensing module on the temporarily installed firmware is also collected, and information from the installed firmware is introduced to achieve closed-loop feedback of the installed component.

[0015] The added components include pre-support components, temporary reinforcement mesh fasteners, and segment spare parts; the data sensed by the sensing module includes the deformation degree and mechanical strength of the pre-support components, the mechanical strength analysis of the temporary reinforcement fasteners, and the weight information of the segment spare parts.

[0016] The beneficial effects of this invention are as follows: This invention breaks down information barriers in all aspects of shield tunneling by comprehensively converging and deeply integrating multiple production factors. It achieves closed-loop management of the entire process from data collection and intelligent analysis to precise control, significantly improving the safety, efficiency, and intelligence level of shield tunneling. It is applicable to shield tunnel projects under various geological conditions.

[0017] During operation, construction collaboration efficiency improved by 15%, and the accident rate decreased by 20%. This system effectively addresses the management needs of tunnel boring machine (TBM) construction under complex geological conditions, providing a reliable intelligent control solution for TBM construction. Based on big data and machine learning, the system can analyze historical optimal tunneling parameters and, combined with current geological conditions, recommend matching combinations of key parameters such as thrust, rotation speed, and velocity to operators, thereby assisting in achieving efficient and stable tunneling and reducing over-reliance on the personal experience of operators. Attached Figure Description

[0018] Figure 1 This is a flowchart of the method proposed in this invention; Detailed Implementation

[0019] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention. Example 1

[0020] A smart management system for tunnel boring machine (TBM) construction based on the convergence of multiple production factors, see [link / reference]. Figure 1 Along the data processing direction, it includes, in sequence, a multi-production factor sensing unit, a heterogeneous data fusion hub unit, an intelligent decision engine unit, a control command execution unit, and a visualization feedback unit, as detailed below: Multi-production factor sensing unit: includes multiple IoT nodes. Each IoT node is used to collect raw data of production factors such as construction personnel, equipment, materials, environment, and processes in real time. The edge computing terminal performs preliminary preprocessing on the collected raw data and then transmits the preprocessed data to the heterogeneous data fusion hub unit. The heterogeneous data fusion hub unit adopts a federated learning framework and consists of multiple data processing nodes and a model aggregation node. The data processing nodes are distributed near each edge computing terminal and perform screening on the pre-processed data transmitted from the edge computing terminal. The processed data is then transmitted to the intelligent decision engine unit. The intelligent decision engine unit includes a data storage module, an algorithm model library, and a decision analysis module; The data storage module uses a distributed database to store the fused data transmitted from the heterogeneous data fusion hub, and the storage capacity can be expanded according to actual needs. The algorithm model library contains a variety of deep learning models, including recurrent neural networks (RNN) for construction progress prediction, convolutional neural networks (CNN) for quality defect identification, long short-term memory networks (LSTM) for safety risk prediction, and deep learning models based on attention mechanisms for dynamic cost accounting. The decision analysis module calls the corresponding models in the algorithm model library to perform data mining and analysis on the fused data. The analysis mainly includes: construction progress analysis, quality defect analysis, safety risk analysis, and cost accounting analysis. The control command execution unit includes a command generation module and a command issuance module. The command generation module generates a distributed adaptive control scheme based on the analysis results and decision suggestions output by the intelligent decision engine. The command issuance module issues the generated control scheme to the field execution system through 5G communication technology. The command issuance module tracks and provides feedback on the execution status of the commands in real time. The visualization feedback unit includes a data visualization module and a human-computer interaction module. The data visualization module uses 3D modeling technology and big data visualization tools to present the control information output by the intelligent decision engine in the form of 3D figures, charts, and curves, constructing a full-element control dashboard. The human-computer interaction module supports managers in configuring parameters and intervening with commands. Managers can set various parameters of the system or directly issue control commands through input devices such as touch screens, mice, and keyboards. The human-computer interaction module also has data query and report generation functions, allowing managers to query historical data and analysis results, and generate various statistical reports. Finally, the feedback effect is transmitted to the multi-production factor sensing unit, which then re-collects and processes the feedback data, entering the next control cycle to form a closed-loop feedback.

[0021] The screening process includes: firstly, data denoising, which uses wavelet transform algorithm to remove noise from the data and improve the signal-to-noise ratio; secondly, data normalization, which converts data of different types and magnitudes into a unified numerical range to facilitate subsequent data analysis and model training; and finally, a dynamically updated construction digital twin model is constructed based on the processed data.

[0022] The control plan includes instructions for dispatching construction personnel, instructions for adjusting the parameters of tunnel boring machines, instructions for allocating material supplies, and instructions for optimizing construction procedures.

[0023] The decision analysis module calls the corresponding models from the algorithm model library, which are derived from historical project data, historical design data, and pre-design analysis data. After screening out the construction differences with historical project data, construction differences with historical design data, and differences with the mechanical data of each force point in the pre-design analysis, it forms mining analysis points and analyzes the reasons for the differences to ensure safety.

[0024] It also includes a production factor sensing unit installed on the site adjustment physical component, and the production factor sensing unit is connected in parallel to the transmission path in the data processing direction to form a parallel loop of the production factor sensing unit, so as to introduce data change information after the adjustment component in real time.

[0025] For construction progress, an RNN model is used to analyze historical progress data and current construction parameters to predict future construction progress and compare it with the planned progress, thus providing early warning of construction progress deviations. For quality defects, a CNN model is used to analyze image and sensor data collected during construction to identify potential quality defects and trace their causes. For safety risks, an LSTM model is used to analyze environmental monitoring data, equipment operation data, and personnel behavior data to predict potential safety risks, such as gas leaks, equipment failures, and personnel entering dangerous areas. For cost accounting, a deep learning model based on an attention mechanism analyzes material consumption data, equipment energy consumption data, and personnel work hours data to achieve dynamic cost accounting and prediction, and generate decision-making suggestions.

[0026] The dynamic cost accounting introduces a dynamic cost accounting formula. Dynamically predicted total cost = Actual costs incurred + Estimated costs to be incurred upon completion.

[0027] This invention provides an intelligent management and control method for tunnel boring machine (TBM) construction based on the convergence of multiple production factors, comprising the following steps: S10: The sensing unit collects data from multiple production factors in real time, and the edge terminal realizes data aggregation and preprocessing; S20: The data fusion center further processes the received data and constructs a dynamically updated twin model; S30: Conduct intelligent decision analysis based on fused data and twin models; S40: Based on intelligent decision analysis suggestions, generate and execute a distributed adaptive control scheme; S50: The results after execution are presented after visualization processing, and manual intervention and control can be performed during the process.

[0028] This embodiment includes a multi-production factor sensing unit, a heterogeneous data fusion hub unit, an intelligent decision engine unit, a control command execution unit, and a visualization feedback unit. The multi-production factor sensing unit collaborates with edge computing terminals and IoT nodes to achieve real-time aggregation of multi-source factors. Based on this unit, the heterogeneous data fusion hub unit performs noise reduction and normalization on the multi-source data, transmitting the dynamic data fusion model to the intelligent decision engine unit. The intelligent decision engine unit mines and analyzes the fused data, deriving a series of decision results. The control command execution unit generates control plans based on the decision results and issues them to the field for execution. The feedback results after field execution are presented by the visualization feedback unit using 3D modeling technology, and the feedback effect is transmitted back to the multi-production factor sensing unit, forming a closed-loop control system. This invention, by constructing a multi-production factor-based full-domain perception and intelligent collaborative control mechanism, achieves precise management of the entire lifecycle of shield tunneling construction, significantly improving construction collaboration efficiency, reducing the accident rate, and providing an integrated intelligent control solution for shield tunneling under complex geological conditions. Example 2

[0029] The working principle of this embodiment is the same as that of embodiment 1. The specific difference is that in step S10, data from the sensing module on the temporarily installed firmware is also collected, and information from the installed firmware is introduced to realize closed-loop feedback of the installed component.

[0030] The added components include pre-support components, temporary reinforcement mesh fasteners, and segment spare parts; the data sensed by the sensing module includes the deformation degree and mechanical strength of the pre-support components, the mechanical strength analysis of the temporary reinforcement fasteners, and the weight information of the segment spare parts.

[0031] In this embodiment, the status of auxiliary firmware added to the device for adjustment and correction can be introduced in a timely manner, and the status of the added firmware can be analyzed to effectively ensure the accuracy of various data after the device is corrected, and further ensure the scientific, accurate and rapid control of the overall management system.

[0032] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A shield tunneling construction intelligent management system based on the convergence of multiple production factors, characterized in that: Along the data processing direction, it includes, in sequence, a multi-production factor sensing unit, a heterogeneous data fusion hub unit, an intelligent decision engine unit, a control command execution unit, and a visualization feedback unit, as detailed below: Multi-production factor sensing unit: includes multiple IoT nodes. Each IoT node is used to collect raw data of production factors such as construction personnel, equipment, materials, environment, and processes in real time. The edge computing terminal performs preliminary preprocessing on the collected raw data and then transmits the preprocessed data to the heterogeneous data fusion hub unit. The heterogeneous data fusion hub unit adopts a federated learning framework and consists of multiple data processing nodes and a model aggregation node. The data processing nodes are distributed near each edge computing terminal and perform screening on the pre-processed data transmitted from the edge computing terminal. The processed data is then transmitted to the intelligent decision engine unit. The intelligent decision engine unit includes a data storage module, an algorithm model library, and a decision analysis module; The data storage module uses a distributed database to store the fused data transmitted from the heterogeneous data fusion hub, and the storage capacity can be expanded according to actual needs. The algorithm model library contains a variety of deep learning models, including recurrent neural networks (RNN) for construction progress prediction, convolutional neural networks (CNN) for quality defect identification, long short-term memory networks (LSTM) for safety risk prediction, and deep learning models based on attention mechanisms for dynamic cost accounting. The decision analysis module calls the corresponding models in the algorithm model library to perform data mining and analysis on the fused data. The analysis mainly includes: construction progress analysis, quality defect analysis, safety risk analysis, and cost accounting analysis. The control command execution unit includes a command generation module and a command issuance module. The command generation module generates a distributed adaptive control scheme based on the analysis results and decision suggestions output by the intelligent decision engine. The command issuance module issues the generated control scheme to the field execution system through 5G communication technology. The command issuance module tracks and provides feedback on the execution status of the commands in real time. The visualization feedback unit includes a data visualization module and a human-computer interaction module. The data visualization module uses 3D modeling technology and big data visualization tools to present the control information output by the intelligent decision engine in the form of 3D figures, charts, and curves, constructing a full-element control dashboard. The human-computer interaction module supports managers in configuring parameters and intervening with commands. Managers can set various parameters of the system or directly issue control commands through input devices such as touch screens, mice, and keyboards. The human-computer interaction module also has data query and report generation functions, allowing managers to query historical data and analysis results, and generate various statistical reports. Finally, the feedback effect is transmitted to the multi-production factor sensing unit, which then re-collects and processes the feedback data, entering the next control cycle to form a closed-loop feedback.

2. The intelligent management system for shield tunneling construction based on the convergence of multiple production factors as described in claim 1, characterized in that: The screening process includes: firstly, data denoising, which uses wavelet transform algorithm to remove noise from the data and improve the signal-to-noise ratio; secondly, data normalization, which converts data of different types and magnitudes into a unified numerical range to facilitate subsequent data analysis and model training; and finally, a dynamically updated construction digital twin model is constructed based on the processed data.

3. The intelligent management system for tunnel boring machine construction based on the convergence of multiple production factors as described in claim 1, characterized in that: The control plan includes instructions for dispatching construction personnel, instructions for adjusting the parameters of tunnel boring machines, instructions for allocating material supplies, and instructions for optimizing construction procedures.

4. The intelligent management system for shield tunneling construction based on the convergence of multiple production factors as described in claim 1, characterized in that: The decision analysis module calls the corresponding models from the algorithm model library, which are derived from historical project data, historical design data, and pre-design analysis data. After screening out the construction differences with historical project data, construction differences with historical design data, and differences with the mechanical data of each force point in the pre-design analysis, it forms mining analysis points and analyzes the reasons for the differences to ensure safety.

5. The intelligent management system for shield tunneling construction based on the convergence of multiple production factors as described in claim 4, characterized in that: It also includes a production factor sensing unit installed on the site adjustment physical component, and the production factor sensing unit is connected in parallel to the transmission path in the data processing direction to form a parallel loop of the production factor sensing unit, so as to introduce data change information after the adjustment component in real time.

6. The intelligent management system for shield tunneling construction based on the convergence of multiple production factors as described in claim 1, characterized in that: Regarding construction progress, an RNN model is used to analyze historical progress data and current construction parameters to predict future construction progress and compare it with the planned progress to achieve early warning of construction progress deviation; regarding quality defects, a CNN model is used to analyze image data and sensor data collected during construction to identify potential quality defects and trace the causes of the defects. For safety risks, LSTM models are used to analyze environmental monitoring data, equipment operation data, and personnel behavior data to predict potential safety risks, such as gas leaks, equipment failures, and personnel accidentally entering dangerous areas. For cost accounting, deep learning models based on attention mechanisms analyze material consumption data, equipment energy consumption data, and personnel work hours data to achieve dynamic cost accounting and prediction, and generate decision recommendations.

7. The intelligent management system for shield tunneling construction based on the convergence of multiple production factors as described in claim 1, characterized in that: The dynamic cost accounting introduces a dynamic cost accounting formula.

8. A method for intelligent control of tunnel boring machine (TBM) construction based on the convergence of multiple production factors, using the intelligent management system for TBM construction based on the convergence of multiple production factors as described in claim 1, characterized in that: Includes the following steps: S10: The sensing unit collects data from multiple production factors in real time, and the edge terminal realizes data aggregation and preprocessing; S20: The data fusion center further processes the received data and constructs a dynamically updated twin model; S30: Conduct intelligent decision analysis based on fused data and twin models; S40: Based on intelligent decision analysis suggestions, generate and execute a distributed adaptive control scheme; S50: The results after execution are presented after visualization processing, and manual intervention and control can be performed during the process.

9. The intelligent control method for shield tunneling construction based on the convergence of multiple production factors as described in claim 8, characterized in that: In step S10, data from the sensing module on the temporarily installed firmware is also collected, and information from the installed firmware is introduced to achieve closed-loop feedback of the installed component.

10. The intelligent control method for shield tunneling construction based on the convergence of multiple production factors as described in claim 9, characterized in that: The added components include pre-support components, temporary reinforcement mesh fasteners, and segment spare parts; the data sensed by the sensing module includes the deformation degree and mechanical strength of the pre-support components, the mechanical strength analysis of the temporary reinforcement fasteners, and the weight information of the segment spare parts.

Images