Tunnel construction management method and system

Abstract

The invention discloses a tunnel construction management method and system, and relates to the technical field of tunnel construction management, and the method comprises the steps: collecting tunnel construction parameters through a multi-source sensor network, obtaining original monitoring data, transmitting the original monitoring data to a data processing end through a data transmission architecture, carrying out the standardization processing and data cleaning of the original monitoring data, and carrying out the management of tunnel construction. The method comprises the steps of obtaining standardized monitoring data, constructing a tunnel three-dimensional model based on the standardized monitoring data, performing construction state evaluation through a simulation calculation system to obtain an evaluation result, generating a construction management instruction through a decision support mechanism according to the evaluation result, and performing tunnel construction management. According to the tunnel construction management system, construction management can be performed on the tunnel according to the monitoring data acquired automatically, and when construction management is performed, the terminal equipment is used for acquiring the corresponding construction management instruction in real time according to the acquired monitoring data, so that environments such as construction progress, safety monitoring and emergency response can be coordinated, the management portability is improved, and the management difficulty is reduced.

Description

Technical Field

[0001] This invention relates to the field of tunnel construction management technology, and in particular to a tunnel construction management method and system. Background Technology

[0002] As a crucial component of modern transportation infrastructure, tunnel engineering has undergone a transformation in its construction management technology, evolving from traditional manual management to information-based management, and finally to intelligent management. Early tunnel construction management relied primarily on manual experience and paper records, resulting in low efficiency and data loss. With the development of computer technology, information technology has been gradually introduced into tunnel construction management, such as using sensors for localized monitoring and utilizing databases to store construction data, thus improving management efficiency to some extent.

[0003] Currently, tunnel construction management primarily employs a combination of manual inspections and fixed monitoring points. Specifically, construction management personnel regularly inspect the tunnel site, observing conditions such as surrounding rock deformation, lining cracks, and water leakage, and recording relevant data using paper or spreadsheets. Simultaneously, a limited number of monitoring devices, including displacement gauges, stress gauges, and temperature and humidity sensors, are deployed at key locations within the tunnel to collect physical parameters at specific locations. This monitoring data is typically read periodically by designated personnel and subjected to simple statistical analysis; any abnormal data is then reported to relevant personnel for further action. However, when dealing with the construction management of large and complex tunnel networks, existing technologies and traditional management methods face significant challenges in coordinating the construction progress of different parts, ensuring safety monitoring, and facilitating emergency response. Summary of the Invention

[0004] The main objective of this invention is to propose a tunnel construction management method and system, which aims to solve the technical problems that existing technologies face when managing the construction of large and complex tunnel networks. Traditional management methods are facing significant challenges in coordinating the construction progress of various parts, safety monitoring, and emergency response.

[0005] To achieve the above objectives, in a first aspect, the present invention proposes a tunnel construction management method, comprising the following steps: Tunnel construction parameters are collected through a multi-source sensor network to obtain raw monitoring data; wherein, the multi-source sensor network includes temperature and humidity sensors, gas sensors and vibration sensors that are arranged in layers on the tunnel arch, sidewalls and road surface according to a preset spatial arrangement rule, as well as a radio frequency identification device positioning system composed of a tag and reader network architecture, and a three-dimensional laser scanner configured according to a site arrangement rule. The raw monitoring data is transmitted to the data processing terminal through a data transmission architecture; wherein, the data transmission architecture uses short-range wireless communication to build node topology relationships and converts analog signals into digital signals through a heterogeneous data fusion transmission mechanism; The original monitoring data is standardized and cleaned to obtain standardized monitoring data; A three-dimensional model of the tunnel is constructed based on the standardized monitoring data, and the construction status is evaluated through a simulation calculation system to obtain the evaluation results. Based on the assessment results, construction management instructions are generated through a decision support mechanism, and tunnel construction management is carried out.

[0006] In one embodiment, the step of acquiring tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data includes: Multiple scanning stations are set up inside the tunnel according to the station layout rules; Point cloud data was collected at each scanning station; Based on the coordinate calculation method, the point cloud data of each station is transformed and unified to obtain complete tunnel point cloud data.

[0007] In one embodiment, the step of standardizing and cleaning the original monitoring data to obtain standardized monitoring data includes: The original monitoring data is time-aligned using a time synchronization mechanism; wherein, the time synchronization mechanism is based on a time-alignment algorithm for a high-precision clock. The aligned monitoring data is converted into a standard format according to the format conversion rules; wherein, the format conversion rules adopt the field definition method of CSV standardized template; The standardized monitoring data is obtained by cleaning the standard format monitoring data using a data cleaning algorithm.

[0008] In one embodiment, the step of cleaning the standard-format monitoring data using a data cleaning algorithm to obtain the standardized monitoring data includes: The vibration signal acquired by the vibration sensor is filtered by a Butterworth filter, wherein the cutoff frequency of the Butterworth filter is determined according to the characteristics of the vibration signal. Detect missing data in the monitoring data; The missing data processing method is determined based on the data processing method selection logic; wherein, when the amount of missing data is less than a preset threshold, the mean filling method is used, and when the amount of missing data is greater than or equal to the preset threshold, the linear interpolation method is used. The missing data is filled in using the missing data processing method described above to obtain the standardized monitoring data.

[0009] In one embodiment, the step of constructing a three-dimensional model of the tunnel based on the standardized monitoring data and evaluating the construction status through a simulation calculation system to obtain the evaluation result includes: The point cloud data is sequentially processed by noise filtering, stitching and surface fitting to obtain a three-dimensional geometric model of the tunnel. The length, width, and lining thickness parameters of the tunnel are automatically calculated and extracted from the three-dimensional geometric model of the tunnel according to the geometric parameter extraction rules. The temperature and humidity data, gas concentration data, and vibration data from the standardized monitoring data are mapped onto the three-dimensional geometric model of the tunnel to obtain the three-dimensional model of the tunnel.

[0010] In one embodiment, the step of constructing a three-dimensional model of the tunnel based on the standardized monitoring data and evaluating the construction status through a simulation calculation system to obtain the evaluation result further includes: Based on the tunnel structural characteristics, the three-dimensional model of the tunnel is meshed according to the finite element meshing principle, wherein the finite element meshing principle adopts the element density control method based on the tunnel structural characteristics; The standardized monitoring data is input into a pre-trained machine learning prediction model to obtain the tunnel construction status prediction result. The machine learning prediction model is constructed using a support vector machine or a long short-term memory network and is trained through feature selection and training strategies. The evaluation result is generated based on the grid division result and the tunnel construction status prediction result.

[0011] In one embodiment, the step of generating construction management instructions and conducting tunnel construction management based on the evaluation results through a decision support mechanism includes: The construction plan is simulated using a virtual simulation engine; among them, a multi-parameter coupled simulation method is used to establish the correlation mapping relationship between environmental parameters and equipment status. The simulated construction plan is quantitatively evaluated based on the plan evaluation index quantification system, which includes energy consumption index and safety factor index, and evaluation weights are set for each index. The construction management instructions are generated based on the quantitative assessment results, and tunnel construction management is carried out.

[0012] In one embodiment, after the step of acquiring tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data, the method further includes: The standardized monitoring data is monitored through a real-time response system. When any parameter in the standardized monitoring data is detected to exceed a preset threshold, an early warning signal is triggered according to the early warning triggering logic; wherein, the preset threshold is dynamically adjusted through a threshold dynamic adjustment algorithm; A resource scheduling scheme is generated based on a resource scheduling optimization model; wherein the resource scheduling optimization model is constructed based on a matching algorithm between maintenance requirements and resource allocation.

[0013] In one embodiment, in the step of acquiring tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data, the short-range wireless communication is ZigBee communication or LoRa communication, and the heterogeneous data fusion transmission mechanism realizes the conversion of analog signals to digital signals through a unified interface.

[0014] Based on the same technical concept, in a second aspect, the present invention also proposes a tunnel construction management system for executing the tunnel construction management method described in the first aspect.

[0015] The technical solution of this invention collects tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data. This raw monitoring data is then transmitted to a data processing terminal via a data transmission architecture. The raw monitoring data undergoes standardization and data cleaning to obtain standardized monitoring data. Based on this standardized monitoring data, a three-dimensional model of the tunnel is constructed, and the construction status is evaluated using a simulation calculation system to obtain evaluation results. Based on these results, a decision support mechanism is used to generate construction management instructions and manage tunnel construction. This invention enables tunnel construction management based on automatically collected monitoring data. Since construction management is based on real-time access to corresponding construction management instructions using terminal devices derived from the acquired monitoring data, this invention can coordinate construction progress, safety monitoring, and emergency response, improving management portability and reducing management difficulty. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0017] Figure 1 A flowchart of the tunnel construction management method provided by the present invention.

[0018] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] It should be noted that if the embodiments of the present invention involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of the components in a specific posture. If the specific posture changes, the directional indicators will also change accordingly.

[0021] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the use of "and / or" or "and / or" throughout the text includes three parallel solutions. For example, "A and / or B" includes solution A, solution B, or a solution where both A and B are satisfied simultaneously. Furthermore, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0022] In the field of tunnel construction management, traditional methods rely on a combination of manual inspections and fixed monitoring points. This results in limited spatial coverage and insufficient temporal continuity of monitoring data, making it difficult to meet the dynamic perception requirements of large and complex tunnel networks for construction status. Specifically, manual inspections are limited by time intervals and personnel experience, making it impossible to capture initial changes in surrounding rock deformation or lining cracks in real time; fixed monitoring points are only deployed at preset key locations, failing to reflect the distribution of physical parameters across the entire tunnel cross-section; and monitoring data requires manual reading and simple statistical analysis, resulting in processing delays. Consequently, construction progress coordination, safety status assessment, and emergency response decisions lack timely and reliable data support, thus affecting the systematic nature and accuracy of overall construction management.

[0023] For example, in the construction of long-distance mountain tunnels traversing multiple geological structures, the stability of the surrounding rock is continuously affected by groundwater activity and changes in in-situ stress. Construction managers need to regularly inspect the tunnel arch for water leakage and sidewall cracks. However, due to the long tunnel length and complex internal environment, the frequency of manual inspections cannot cover the rapid evolution of potential risk points, and even small cracks may expand within the interval between inspections. At the same time, the displacement gauges and stress gauges deployed are only located at limited cross-sectional positions, making it impossible to monitor the vibration response and gas concentration distribution in the road surface area. Monitoring data is read and entered into spreadsheets by designated personnel on a weekly basis, and the delay in feedback of abnormal parameters leads to the failure to identify safety hazards in a timely manner, resulting in hindered construction progress and the accumulation of safety risks.

[0024] This invention proposes a tunnel construction management method and system.

[0025] Please see Figure 1 To facilitate understanding, this tunnel construction management method includes the following steps: S100. Collect tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data; wherein, the multi-source sensor network includes temperature and humidity sensors, gas sensors and vibration sensors arranged in layers on the tunnel arch, sidewalls and road surface according to a preset spatial arrangement rule, as well as a radio frequency identification device positioning system composed of a tag and reader network architecture, and a three-dimensional laser scanner configured according to a site arrangement rule. S200. The original monitoring data is transmitted to the data processing terminal through a data transmission architecture; wherein, the data transmission architecture uses short-range wireless communication to build node topology relationships and converts analog signals into digital signals through a heterogeneous data fusion transmission mechanism; S300. Standardize and clean the raw monitoring data to obtain standardized monitoring data. S400. Based on the standardized monitoring data, a three-dimensional model of the tunnel is constructed, and the construction status is evaluated through a simulation calculation system to obtain the evaluation results. S500. Based on the assessment results, generate construction management instructions through a decision support mechanism and carry out tunnel construction management.

[0026] It should be specifically and clearly stated that the software or programs required in the computer processing steps involved in the examples in this embodiment are all existing technologies. This embodiment only applies them and does not improve or redesign the software or programs themselves. Therefore, they will not be described in detail here.

[0027] Specifically, tunnel construction parameters are collected through a multi-source sensor network to obtain raw monitoring data. This multi-source sensor network can consist of various sensors; for example, temperature and humidity sensors, gas sensors, and vibration sensors can be manually deployed at different locations within the tunnel, and their data can be read periodically. Simultaneously, the positions of equipment and personnel within the tunnel can be tracked manually, or local geometric information of the tunnel can be obtained using simple two-dimensional scanning equipment. The data from these sensors and devices are considered the raw monitoring data.

[0028] It should be specifically and clearly stated that the processing and filtering examples in this embodiment are all existing technologies. This embodiment only applies them and does not improve or design them. Therefore, they will not be described in detail here.

[0029] The raw monitoring data is transmitted to the data processing terminal via a data transmission architecture. This data transmission architecture can employ wired connections, such as connecting various sensors and devices to the data processing terminal via network cables. For analog signals, they can be converted to digital signals manually or using a simple analog-to-digital converter for processing at the data processing terminal.

[0030] The raw monitoring data undergoes standardization and data cleaning to obtain standardized monitoring data. This process may include manual data inspection, manual adjustment of data with inconsistent formats, and manual correction or deletion of obviously erroneous or missing data. For example, sensor readings may be manually checked, and data anomalies may be judged based on experience.

[0031] A three-dimensional model of the tunnel is constructed based on this standardized monitoring data, and the construction status is assessed through a simulation calculation system to obtain the assessment results. The construction of the three-dimensional tunnel model can employ traditional measurement methods, such as total station measurements, followed by modeling using specialized software. The simulation calculation system can perform preliminary stability analysis of the tunnel structure based on preset empirical formulas or simple mechanical models, and provide a preliminary assessment of the construction status based on the analysis results.

[0032] Based on the assessment results, a decision support mechanism is used to generate construction management instructions and manage tunnel construction. This decision support mechanism can be based on a pre-set rule base; for example, when the assessment results show that a certain parameter exceeds the safe range, the system will suggest corresponding pre-set handling solutions. Managers can manually generate construction management instructions based on these suggestions and guide on-site construction personnel to perform the operations. For example, when an abnormal gas concentration is detected, a ventilation instruction can be manually issued.

[0033] In this embodiment, tunnel construction parameters are collected through a multi-source sensor network to obtain raw monitoring data. The raw monitoring data is then transmitted to a data processing terminal via a data transmission architecture. The raw monitoring data is standardized and cleaned to obtain standardized monitoring data. A three-dimensional model of the tunnel is constructed based on the standardized monitoring data, and the construction status is evaluated through a simulation calculation system to obtain evaluation results. Based on the evaluation results, a decision support mechanism is used to generate construction management instructions and manage tunnel construction. This invention enables tunnel construction management based on automatically collected monitoring data. Since construction management is based on the acquired monitoring data and the corresponding construction management instructions are obtained in real time using terminal devices, this invention can coordinate construction progress, safety monitoring, and emergency response, improving management portability and reducing management difficulty.

[0034] In one embodiment, the step of acquiring tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data includes: Multiple scanning stations are set up inside the tunnel according to the station layout rules; Point cloud data was collected at each scanning station; Based on the coordinate calculation method, the point cloud data of each station is transformed and unified to obtain complete tunnel point cloud data.

[0035] In this embodiment, by setting up multiple scanning stations within the tunnel, the 3D laser scanner can comprehensively scan the tunnel interior from different angles and positions, thus avoiding potential data loss due to a single scanning point. Although the point cloud data collected at each scanning station are independent, precise coordinate calculation methods allow for accurate coordinate transformation and unification of these data based on different local coordinate systems. This processing method seamlessly stitches together the originally scattered point cloud data, forming a continuous, complete, and high-precision tunnel point cloud dataset. This complete point cloud dataset provides a precise geometric foundation for subsequent construction of a 3D tunnel model based on standardized monitoring data, greatly improving the accuracy and reliability of the 3D tunnel model. This results in a more comprehensive and detailed assessment of the construction status based on the model, providing a more accurate decision-making basis for tunnel construction management.

[0036] In one embodiment, the step of standardizing and cleaning the original monitoring data to obtain standardized monitoring data includes: The original monitoring data is time-aligned using a time synchronization mechanism; wherein, the time synchronization mechanism is based on a time-alignment algorithm for a high-precision clock. The aligned monitoring data is converted into a standard format according to the format conversion rules; wherein, the format conversion rules adopt the field definition method of CSV standardized template; The standardized monitoring data is obtained by cleaning the standard format monitoring data using a data cleaning algorithm.

[0037] Specifically, time synchronization mechanisms refer to technical means to ensure time consistency between different devices or data sources in a distributed system. In multi-source sensor networks, because each sensor operates independently, its internal clock may drift, leading to inconsistent timestamps in the collected data. Time synchronization mechanisms calibrate the time of each sensor through a unified time base, ensuring that all data has an accurate and consistent time reference. This mechanism can employ standard protocols such as Network Time Protocol (NTP) or Precision Time Protocol (PTP), providing a unified time signal to each sensor node through a network server to achieve high-precision time synchronization; or it can use hardware synchronization methods, such as using high-precision time signals provided by GPS receivers, or distributing a unified clock signal to all sensors through a dedicated synchronization bus to ensure the synchronicity of data acquisition. The time stamp alignment algorithm of a high-precision clock is the core component of the time synchronization mechanism. It uses the precise time information provided by a high-precision clock source to calibrate and align the timestamps of data from different sensors. This algorithm aims to eliminate time deviations between data from different sensors, ensuring that data collected at the same time can be accurately correlated. The algorithm can be based on timestamp interpolation, which uses a high-precision clock to linearly or non-linearly interpolate the timestamps of sensor data and adjust them to a unified time axis; or it can use an event-triggered timestamp correction method, which uses the event occurrence time recorded by a high-precision clock to uniformly calibrate the timestamps of all relevant sensor data when a specific synchronization event is detected.

[0038] Format conversion rules refer to the specifications for uniformly converting raw monitoring data from different sources and in different formats into a standard data format. Data collected by multi-source sensor networks may have diverse formats, such as binary, XML, JSON, or custom text formats. Establishing unified format conversion rules ensures that all data conforms to a preset data structure and encoding standard before entering subsequent processing, facilitating unified parsing and processing. These rules can define a data model and corresponding conversion mapping table, mapping raw data fields to standard format fields and specifying data types, units, and encoding methods; or they can employ metadata-based conversion rules, defining data metadata and combining it with predefined conversion scripts or programs to achieve automated format conversion. The CSV standardized template field definition method is a specific implementation of format conversion rules. It leverages the simplicity and universality of the comma-separated values ​​(CSV) file format to define a standardized set of field names, data types, and arrangement order. By using CSV templates, it ensures that all converted monitoring data is presented in a unified tabular format, with each field having a clear meaning, facilitating data storage, transmission, and subsequent analysis and processing. This method can explicitly specify the header row of a CSV file, including all necessary field names, and specify the data type and unit for each field; or it can provide a CSV template file containing predefined column names and sample data to guide data conversion programs or manual operators to populate the original data into the corresponding fields, ensuring data format consistency.

[0039] Data cleaning algorithms refer to technical methods used to identify, correct, or delete erroneous, inconsistent, or incomplete data in raw monitoring data. In sensor data acquisition, problems such as outliers, missing values, duplicate records, or format errors may occur due to environmental interference, sensor malfunctions, or transmission errors. Data cleaning algorithms aim to improve data quality and ensure the accuracy and reliability of subsequent analysis. These algorithms can employ statistical methods, such as those based on the mean, standard deviation, or interquartile range (IQR), to detect and handle outliers; or they can use machine learning methods, such as Isolation Forest or Local Outlier Factor (LOF), to identify anomalous patterns in the data. For missing data, interpolation methods (such as linear interpolation and spline interpolation) or imputation methods (such as mean imputation and mode imputation) can be used; for duplicate data, unique identifiers can be used for detection and deletion.

[0040] In this embodiment, a series of data processing steps are introduced to address the inconsistencies in timing, format, and quality of raw monitoring data collected by a multi-source sensor network, providing high-quality input for subsequent tunnel 3D model construction and construction status assessment. First, a time synchronization mechanism is used, employing a high-precision clock time-stamp alignment algorithm, to precisely calibrate the raw monitoring data from different sensors with varying acquisition frequencies and timestamps. This step ensures consistency across all data in the time dimension, enabling accurate correlation of data collected by different sensors at the same time, laying the foundation for subsequent fusion analysis. Subsequently, to unify the data structure, this scheme employs format conversion rules to convert the time-stamped monitoring data into a standard format. Specifically, using the field definition method of a CSV standardized template, heterogeneous data is uniformly mapped to a predefined field structure, ensuring all data is presented in a unified and standardized tabular format. This unified format greatly simplifies the data parsing and processing, avoiding the complexity caused by data format diversity. Finally, to improve data quality, this scheme further processes the standard-format monitoring data using a data cleaning algorithm. This algorithm can identify and correct outliers, missing values, duplicate records, or format errors in the data, thereby eliminating data noise and inaccuracies. Through the organic combination of the above steps, the original monitoring data is comprehensively optimized and improved in terms of time, format, and quality, ensuring that the data input into the tunnel 3D model construction and simulation calculation system is accurate, reliable, and consistent, thus significantly improving the accuracy of construction status assessment and the effectiveness of decision support.

[0041] In one embodiment, the step of cleaning the standard-format monitoring data using a data cleaning algorithm to obtain the standardized monitoring data includes: The vibration signal acquired by the vibration sensor is filtered by a Butterworth filter, wherein the cutoff frequency of the Butterworth filter is determined according to the characteristics of the vibration signal. Detect missing data in the monitoring data; The missing data processing method is determined based on the data processing method selection logic; wherein, when the amount of missing data is less than a preset threshold, the mean filling method is used, and when the amount of missing data is greater than or equal to the preset threshold, the linear interpolation method is used. The missing data is filled in using the missing data processing method described above to obtain the standardized monitoring data.

[0042] Specifically, a Butterworth filter is a filter with a maximally flat passband response. It exhibits a flat frequency response within the passband and a monotonically decreasing attenuation characteristic in the stopband, effectively filtering out high-frequency or low-frequency noise while preserving the original waveform to the greatest extent possible. It can be implemented through software algorithms executed by a digital signal processor or microcontroller, such as digital filtering using IIR or FIR filter structures; or it can be an analog filter circuit constructed using passive components such as resistors, capacitors, and inductors, or active components such as operational amplifiers. The cutoff frequency is a key parameter of the filter, defining the frequency boundary through which the signal passes or is attenuated. Determining the cutoff frequency based on the characteristics of the vibration signal means that the filter can adaptively adjust according to the spectral characteristics, noise distribution, and the range of vibration frequencies of interest of the actually monitored vibration signal. For example, by performing spectral analysis such as Fourier transform on historical or real-time vibration signals, the main noise frequency components or target vibration frequency range can be identified, and a suitable cutoff frequency can be set; alternatively, the cutoff frequency can be dynamically adjusted according to different construction stages or environmental conditions based on empirical knowledge or a pre-set rule base. Detecting missing data in the monitored data refers to identifying incomplete or non-existent data points in the dataset. This can be achieved in various ways, such as checking for null values, non-numeric values, or specific placeholders in the data records; or identifying data segments with abnormal time intervals or consecutive missing data by analyzing the timestamp sequence of the data. The data processing method selection logic refers to dynamically selecting the most suitable processing method for the current data situation based on specific conditions or rules. This logic can improve the flexibility and efficiency of data processing. For example, a set of decision rules can be designed based on factors such as the quantity, type, distribution, or importance of missing data to automatically select different imputation or repair strategies. When the amount of missing data is less than a preset threshold, the mean imputation method is a simple and effective missing data processing technique that replaces missing values ​​with the average of the known data in the dataset. This method is suitable for situations where the amount of missing data is small and the data distribution is relatively uniform, maintaining the overall statistical characteristics of the dataset. In addition to mean imputation, when the amount of missing data is small, median imputation, mode imputation, or imputation using the previous valid observation can also be used. When the amount of missing data is greater than or equal to the preset threshold, the linear interpolation method is used to estimate missing values ​​by establishing a linear relationship between known data points. It assumes an approximate linear trend between data points, making it suitable for situations with a large amount of missing data but a relatively stable data trend, thus better preserving the temporal characteristics of the data. Besides linear interpolation, more complex interpolation methods such as spline interpolation, polynomial interpolation, or regression model-based predictive imputation can be used when the amount of missing data is large. The preset threshold is a critical value used to distinguish between different missing data processing strategies. It defines the boundary between "small missing data" and "large missing data".This threshold can be set based on experience, statistical analysis results, or different requirements for data integrity. For example, the threshold can be set by analyzing historical data to determine at what proportion of missing data mean imputation and linear interpolation each perform best.

[0043] In this embodiment, the monitoring data in standard format undergoes refined cleaning to ensure the accuracy of subsequent data analysis. Specifically, firstly, for the signals acquired by vibration sensors, which are susceptible to environmental noise interference, this solution introduces a Butterworth filter for filtering. The cutoff frequency of this filter is not fixed but dynamically determined based on the actual characteristics of the vibration signal. This allows the filtering process to adaptively remove noise while preserving effective vibration information to the maximum extent, avoiding over-filtering or under-filtering. After noise removal, the system further detects whether there is missing data in the monitoring data. To address different degrees of data loss, this solution designs an intelligent data processing method selection logic: when the detected amount of missing data is less than a preset threshold, a mean imputation method is used for fast and statistically friendly repair; when the amount of missing data reaches or exceeds the preset threshold, a linear interpolation method is used to more accurately estimate the missing value by utilizing the trend information between data points. This hierarchical processing strategy ensures that the most appropriate imputation method can be selected regardless of the degree of data loss, thereby effectively restoring the integrity and continuity of the data. Through the combined effect of filtering and intelligent filling, noise and missing values ​​in the original monitoring data were effectively processed, significantly improving the quality and reliability of the standardized monitoring data, and providing a solid data foundation for subsequent tunnel 3D model construction and construction status assessment.

[0044] In one embodiment, the step of constructing a three-dimensional model of the tunnel based on the standardized monitoring data and evaluating the construction status through a simulation calculation system to obtain the evaluation result includes: The point cloud data is sequentially processed by noise filtering, stitching and surface fitting to obtain a three-dimensional geometric model of the tunnel. The length, width, and lining thickness parameters of the tunnel are automatically calculated and extracted from the three-dimensional geometric model of the tunnel according to the geometric parameter extraction rules. The temperature and humidity data, gas concentration data, and vibration data from the standardized monitoring data are mapped onto the three-dimensional geometric model of the tunnel to obtain the three-dimensional model of the tunnel.

[0045] Specifically, by refining the raw point cloud data, firstly, noise filtering removes interference, then multi-source point clouds are integrated using stitching technology, and finally, a high-precision 3D geometric model of the tunnel is constructed through surface fitting. This process ensures the accuracy and completeness of the tunnel's geometric information. Subsequently, through preset geometric parameter extraction rules, key construction quality parameters, such as the tunnel's length, width, and lining thickness, can be automatically and efficiently quantified from the complex geometric model. The accurate acquisition of these geometric parameters provides a solid foundation for subsequent construction quality assessment. More importantly, this solution organically combines standardized temperature and humidity data, gas concentration data, and vibration data with the aforementioned precisely constructed 3D geometric model of the tunnel using spatial mapping technology. This combination allows the originally discrete and heterogeneous monitoring data to be presented in an intuitive and spatial way, forming a unified and information-rich 3D tunnel model. This model not only shows the physical form of the tunnel but also incorporates real-time environmental and structural status information, thus comprehensively and accurately reflecting various key indicators during the tunnel construction process. In this way, the solution effectively solves the technical problem of how to integrate multi-source heterogeneous data into a comprehensive three-dimensional model with spatial semantics, providing a more comprehensive and accurate data foundation for subsequent simulation calculations and construction status assessments, and significantly improving the reliability of the assessment and the scientific nature of the decision-making.

[0046] In one embodiment, the step of constructing a three-dimensional model of the tunnel based on the standardized monitoring data and evaluating the construction status through a simulation calculation system to obtain the evaluation result further includes: Based on the tunnel structural characteristics, the three-dimensional model of the tunnel is meshed according to the finite element meshing principle, wherein the finite element meshing principle adopts the element density control method based on the tunnel structural characteristics; The standardized monitoring data is input into a pre-trained machine learning prediction model to obtain the tunnel construction status prediction result. The machine learning prediction model is constructed using a support vector machine or a long short-term memory network and is trained through feature selection and training strategies. The evaluation result is generated based on the grid division result and the tunnel construction status prediction result.

[0047] Specifically, the finite element method (FEM) mesh generation principle refers to discretizing a continuous 3D tunnel model into a finite number of interconnected elements and nodes for numerical analysis. This principle aims to ensure the quality and applicability of the mesh to accurately simulate the mechanical behavior of the tunnel during construction. For example, structured mesh generation, such as hexahedral or tetrahedral meshes, can be used to accommodate different geometries and analysis requirements. Another approach is to use unstructured mesh generation for greater flexibility in adapting to complex boundaries and geometric features. The element density control method based on tunnel structural characteristics refers to dynamically adjusting the size and density of mesh elements during mesh generation according to key parts of the tunnel structure (such as lining, surrounding rock interfaces, and support structures) and expected stress concentration areas.

[0048] For example, in areas of stress concentration or drastic geometric changes, a finer mesh can be used to improve computational accuracy; while in areas of gradual stress change, a coarser mesh can be used to improve computational efficiency. A pre-trained machine learning prediction model refers to an algorithmic model that learns and optimizes through historical or simulation data, enabling it to identify patterns in the data and predict future tunnel construction conditions. This model has already completed its training process and possesses predictive capabilities before practical application. For example, this model can be a regression model used to predict tunnel settlement or deformation rate; or it can be a classification model used to determine whether the tunnel is currently in a stable state or has potential risks. Support Vector Machines (SVMs) are a type of supervised learning model that can be used for classification and regression analysis.

[0049] By finding an optimal hyperplane to maximize the margin between data points of different categories, effective data classification or fitting can be achieved. Long Short-Term Memory (LSTM) networks are a special type of recurrent neural network, particularly suitable for processing and predicting time series data. They effectively solve the gradient vanishing or exploding problems of traditional recurrent neural networks by introducing gating mechanisms (input gate, forget gate, output gate), enabling them to learn and remember long-term dependencies. Feature selection and training strategies refer to the process of selecting the most relevant input features and using appropriate training methods to optimize model performance when building a machine learning model. Feature selection can reduce model complexity and improve generalization ability, for example, through correlation analysis, principal component analysis, or recursive feature elimination. Training strategies include data partitioning (training set, validation set, test set), optimization algorithms (such as gradient descent and its variants), regularization techniques (such as L1 and L2 regularization), and hyperparameter tuning (such as grid search and random search) to ensure the model can learn effectively and avoid overfitting.

[0050] In the specific implementation process, the first step is to mesh the 3D tunnel model obtained through the above steps. For example, finite element analysis software can be used to mesh the tunnel based on the different materials and geometric characteristics of the tunnel lining, surrounding rock, and support structure, employing a unit density control method based on the tunnel structural characteristics. For instance, smaller element sizes can be set in stress-concentrated areas such as the tunnel arch and sidewalls, as well as at the interface between the lining and surrounding rock, to obtain a finer mesh and improve the accuracy of subsequent mechanical analysis; while in the surrounding rock parts far from these critical areas, relatively larger element sizes can be used to balance computational efficiency. Simultaneously, standardized monitoring data, such as temperature and humidity data, gas concentration data, and vibration data, are input into a pre-trained machine learning prediction model. For example, a prediction model based on a long short-term memory network can be used, which has been trained using historical monitoring data and corresponding tunnel construction status data. During training, feature selection strategies, such as selecting vibration frequency, vibration amplitude, temperature change rate, and gas concentration change rate as key input features, and cross-validation and other training strategies are used to optimize the model parameters, enabling it to accurately predict the future deformation trend or potential instability risk of the tunnel. Finally, the structural information provided by the 3D tunnel model after finite element mesh generation is combined with the tunnel construction status prediction results output by the machine learning prediction model to generate the final evaluation result. For example, the prediction model may indicate that the deformation of a certain area will exceed a preset threshold in the future, while the finite element model can provide details of the stress distribution and structural response in that area. Combining these two pieces of information yields a comprehensive and forward-looking evaluation report that clearly identifies potential risk areas, risk levels, and possible causes.

[0051] In this embodiment, the 3D tunnel model can be finely discretized, enabling subsequent structural mechanics analysis to more accurately simulate the complex mechanical behaviors of the tunnel during construction, such as stress and deformation, especially achieving high-precision analysis results in critical structural areas. Simultaneously, standardized monitoring data is input into a pre-trained machine learning prediction model, employing advanced models such as support vector machines or long short-term memory networks, combined with feature selection and training strategies. This allows the system to learn from massive amounts of monitoring data and identify potential construction risk patterns, thereby accurately predicting the tunnel construction status. This method, combining physical modeling and data-driven prediction, significantly improves the depth and foresight of construction status assessment, enabling earlier detection of potential safety hazards or structural problems, and providing strong technical support for timely adjustments to construction plans and prevention of accidents.

[0052] In one embodiment, the step of generating construction management instructions and conducting tunnel construction management based on the evaluation results through a decision support mechanism includes: The construction plan is simulated using a virtual simulation engine; among which, a multi-parameter coupled simulation method is used to establish the correlation mapping relationship between environmental parameters and equipment status. The simulated construction plan is quantitatively evaluated based on the plan evaluation index quantification system, which includes energy consumption index and safety factor index, and evaluation weights are set for each index. The construction management instructions are generated based on the quantitative assessment results, and tunnel construction management is carried out.

[0053] In one embodiment, after the step of acquiring tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data, the method further includes: The standardized monitoring data is monitored through a real-time response system. When any parameter in the standardized monitoring data is detected to exceed a preset threshold, an early warning signal is triggered according to the early warning triggering logic; wherein, the preset threshold is dynamically adjusted through a threshold dynamic adjustment algorithm; A resource scheduling scheme is generated based on the resource scheduling optimization model; wherein, the resource scheduling optimization model is constructed based on a matching algorithm between maintenance requirements and resource allocation.

[0054] Specifically, by introducing a decision support mechanism, the tunnel construction status assessment results obtained in the preceding steps are transformed into specific construction management instructions. First, a virtual simulation engine is used to simulate various possible construction schemes. This engine, through a multi-parameter coupled simulation method, can establish a complex correlation mapping between environmental parameters and equipment status, thereby accurately predicting the actual effects of different construction schemes under various conditions in a virtual environment. For example, it can simulate the impact of different ventilation schemes on air quality and equipment energy consumption under specific temperature, humidity, and gas concentration conditions, or the impact of different support schemes on tunnel structural stability. Subsequently, these simulated construction schemes are quantitatively evaluated using a scheme evaluation index quantification system. This quantification system specifically includes energy consumption and safety factor indicators, and sets evaluation weights for these indicators, enabling the evaluation process to comprehensively consider multiple dimensions such as economy, efficiency, and safety. For example, during the evaluation process, schemes with high safety factors can be prioritized based on preset weights, while also considering schemes with lower energy consumption. Finally, based on these quantitative evaluation results, the system can generate optimal construction management instructions. This method ensures that the generated instructions are based on scientific simulation and quantitative evaluation, rather than experience or intuition, thereby significantly improving the scientific nature and effectiveness of construction management. It effectively solves the problem of transforming complex evaluation results into optimal management instructions, making tunnel construction management more precise and efficient.

[0055] In this embodiment, a real-time response mechanism is introduced into the tunnel construction management method to address potential emergencies during construction. Based on standardized monitoring data collected by a multi-source sensor network, the real-time response system continuously monitors this standardized data. Once the system detects that any key parameter (such as temperature, humidity, gas concentration, vibration, etc.) exceeds a preset dynamic adjustment threshold, the early warning trigger logic immediately activates an early warning signal. This early warning signal not only indicates potential risks but also drives the operation of a resource scheduling optimization model. This model, based on a matching algorithm between maintenance needs and resource allocation, comprehensively considers the urgency of the current warning, the type of maintenance required, and the characteristics and location of existing available resources (such as personnel, equipment, and materials) to quickly generate the optimal resource scheduling plan. In this way, this solution transforms from passive post-event processing to proactive real-time intervention, ensuring timely and efficient resource allocation in response to abnormal situations, thereby effectively reducing construction risks, ensuring construction safety, and optimizing resource utilization efficiency. Compared to solutions that rely solely on evaluation results to generate management instructions, this significantly improves the real-time performance and responsiveness of construction management.

[0056] In one embodiment, in the step of acquiring tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data, the short-range wireless communication is ZigBee communication or LoRa communication, and the heterogeneous data fusion transmission mechanism realizes the conversion of analog signals to digital signals through a unified interface.

[0057] In this embodiment, by specifically defining short-range wireless communication as ZigBee or LoRa communication, and clarifying that the heterogeneous data fusion transmission mechanism achieves the conversion of analog signals to digital signals through a unified interface, the acquisition and transmission process of tunnel construction parameters is optimized. Specifically, when acquiring tunnel construction parameters through a multi-source sensor network, the analog signals collected by devices such as temperature and humidity sensors, gas sensors, and vibration sensors are first converted into digital signals by their respective analog-to-digital converters (ADCs) or conversion circuits integrated in the sensor modules. Subsequently, these digital signals undergo standardization processing through a unified interface, such as standardizing the data frame format and adding timestamps, to ensure compatibility of data from different types of sensors. Then, these digital signals processed by the unified interface are wirelessly transmitted to the data processing end through ZigBee or LoRa communication modules. If ZigBee communication is used, sensor nodes can form a self-organizing mesh network, reliably transmitting data to the aggregation node through multi-hop routing, and then the aggregation node uploads it to the data processing end; if LoRa communication is used, sensor nodes can directly transmit data to the LoRa gateway within the coverage area, and the gateway is responsible for forwarding the data to the data processing end. This specific choice of communication technology can provide low-power, high-reliability, or long-distance data transmission capabilities based on the characteristics of the tunnel environment, effectively overcoming the impact of the complex electromagnetic environment and physical obstacles within the tunnel on wireless communication. Simultaneously, the introduction of a unified interface enables the efficient and accurate conversion of heterogeneous analog signals from different sensors into digital signals of a unified format, avoiding compatibility issues and data processing complexities caused by inconsistent data formats. This provides a high-quality raw data foundation for subsequent standardization processing, data cleaning, and the construction of 3D tunnel models.

[0058] Based on the same technical concept, in a second aspect, the present invention also proposes a tunnel construction management system for executing the tunnel construction management method described in the first aspect.

[0059] Specifically, the system first communicates with a multi-source sensor network through its data acquisition module. This network includes temperature and humidity sensors, gas sensors, vibration sensors, a radio frequency identification (RFID) positioning system, and a 3D laser scanner to collect tunnel construction parameters and obtain raw monitoring data. Subsequently, the system's data transmission module is responsible for establishing node topology relationships via short-range wireless communication and converting analog signals to digital signals using a heterogeneous data fusion transmission mechanism, transmitting the raw monitoring data to the data processing end. At the data processing end, the system's data processing module performs standardization and data cleaning on the received raw monitoring data. This includes time-scale alignment using a time synchronization mechanism, conversion to a standard format according to format conversion rules, and data cleaning algorithms to obtain standardized monitoring data. Next, the system's model building and evaluation module constructs a 3D tunnel model based on the standardized monitoring data. This involves performing noise filtering, stitching, and surface fitting on the point cloud data to obtain a 3D geometric model of the tunnel, and automatically calculating and extracting the tunnel's length, width, and lining thickness parameters according to geometric parameter extraction rules. Simultaneously, the temperature and humidity data, gas concentration data, and vibration data from the standardized monitoring data are mapped to the 3D geometric model of the tunnel. Building upon this foundation, the module assesses the construction status through a simulation computing system. This may involve finite element mesh generation based on tunnel structural characteristics, inputting standardized monitoring data into a pre-trained machine learning prediction model to obtain tunnel construction status prediction results, and finally generating an assessment result based on the mesh generation and prediction results. Finally, the system's decision support and instruction generation module generates construction management instructions and manages tunnel construction based on the assessment results through a decision support mechanism. This may involve simulating construction plans using a virtual simulation engine and quantitatively evaluating the simulated construction plans according to a quantitative system of plan evaluation indicators. Furthermore, the system may include a real-time response and early warning module to monitor standardized monitoring data and trigger an early warning signal when any parameter exceeds a preset threshold, thereby generating a resource scheduling plan based on a resource scheduling optimization model. Through this integrated system design, the aforementioned complex tunnel construction management methods can be executed efficiently and accurately on a unified platform, achieving fully automated management from data acquisition to decision generation.

[0060] The above description is merely an exemplary embodiment of the present invention and is not intended to limit the scope of the present invention. Any equivalent structural transformations made based on the technical concept of the present invention and the contents of the specification and drawings of the present invention, or direct / indirect applications in other related technical fields, are included within the protection scope of the present invention.

Claims

1. A tunnel construction management method, characterized in that, Includes the following steps: Tunnel construction parameters are collected through a multi-source sensor network to obtain raw monitoring data; wherein, the multi-source sensor network includes temperature and humidity sensors, gas sensors and vibration sensors that are arranged in layers on the tunnel arch, sidewalls and road surface according to a preset spatial arrangement rule, as well as a radio frequency identification device positioning system composed of a tag and reader network architecture, and a three-dimensional laser scanner configured according to a site arrangement rule. The raw monitoring data is transmitted to the data processing terminal through a data transmission architecture; wherein, the data transmission architecture uses short-range wireless communication to build node topology relationships and converts analog signals into digital signals through a heterogeneous data fusion transmission mechanism; The original monitoring data is standardized and cleaned to obtain standardized monitoring data; A three-dimensional model of the tunnel is constructed based on the standardized monitoring data, and the construction status is evaluated through a simulation calculation system to obtain the evaluation results. Based on the assessment results, construction management instructions are generated through a decision support mechanism, and tunnel construction management is carried out.

2. The tunnel construction management method as described in claim 1, characterized in that, The step of acquiring tunnel construction parameters and obtaining raw monitoring data through a multi-source sensor network includes: Multiple scanning stations are set up inside the tunnel according to the station layout rules; Point cloud data was collected at each scanning station; Based on the coordinate calculation method, the point cloud data of each station is transformed and unified to obtain complete tunnel point cloud data.

3. The tunnel construction management method as described in claim 2, characterized in that, The steps of standardizing and cleaning the original monitoring data to obtain standardized monitoring data include: The original monitoring data is time-aligned using a time synchronization mechanism; wherein, the time synchronization mechanism is based on a time-alignment algorithm for a high-precision clock. The aligned monitoring data is converted into a standard format according to the format conversion rules; wherein, the format conversion rules adopt the field definition method of CSV standardized template; The standardized monitoring data is obtained by cleaning the standard format monitoring data using a data cleaning algorithm.

4. The tunnel construction management method as described in claim 3, characterized in that, The step of cleaning the standard-format monitoring data using a data cleaning algorithm to obtain the standardized monitoring data includes: The vibration signal acquired by the vibration sensor is filtered by a Butterworth filter, wherein the cutoff frequency of the Butterworth filter is determined according to the characteristics of the vibration signal. Detect missing data in the monitoring data; The missing data processing method is determined based on the data processing method selection logic; wherein, when the amount of missing data is less than a preset threshold, the mean filling method is used, and when the amount of missing data is greater than or equal to the preset threshold, the linear interpolation method is used. The missing data is filled in using the missing data processing method described above to obtain the standardized monitoring data.

5. The tunnel construction management method as described in claim 4, characterized in that, The steps of constructing a three-dimensional model of the tunnel based on the standardized monitoring data and evaluating the construction status through a simulation calculation system to obtain the evaluation results include: The point cloud data is sequentially subjected to noise filtering, stitching and surface fitting to obtain a three-dimensional geometric model of the tunnel; The length, width, and lining thickness parameters of the tunnel are automatically calculated and extracted from the three-dimensional geometric model of the tunnel according to the geometric parameter extraction rules. The temperature and humidity data, gas concentration data, and vibration data from the standardized monitoring data are mapped onto the three-dimensional geometric model of the tunnel to obtain the three-dimensional model of the tunnel.

6. The tunnel construction management method as described in claim 5, characterized in that, The step of constructing a three-dimensional model of the tunnel based on the standardized monitoring data and evaluating the construction status through a simulation calculation system to obtain the evaluation results further includes: Based on the tunnel structural characteristics, the three-dimensional model of the tunnel is meshed according to the finite element meshing principle, wherein the finite element meshing principle adopts the element density control method based on the tunnel structural characteristics; The standardized monitoring data is input into a pre-trained machine learning prediction model to obtain the tunnel construction status prediction result. The machine learning prediction model is constructed using a support vector machine or a long short-term memory network and is trained through feature selection and training strategies. The evaluation result is generated based on the grid division result and the tunnel construction status prediction result.

7. The tunnel construction management method according to any one of claims 1 to 6, characterized in that, The step of generating construction management instructions and conducting tunnel construction management through a decision support mechanism based on the evaluation results includes: The construction plan is simulated using a virtual simulation engine; among which, a multi-parameter coupled simulation method is used to establish the correlation mapping relationship between environmental parameters and equipment status. The simulated construction plan is quantitatively evaluated based on the plan evaluation index quantification system, which includes energy consumption index and safety factor index, and evaluation weights are set for each index. The construction management instructions are generated based on the quantitative assessment results, and tunnel construction management is carried out.

8. The tunnel construction management method according to any one of claims 1 to 6, characterized in that, After the step of acquiring tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data, the method further includes: The standardized monitoring data is monitored through a real-time response system. When any parameter in the standardized monitoring data is detected to exceed a preset threshold, an early warning signal is triggered according to the early warning triggering logic; wherein, the preset threshold is dynamically adjusted through a threshold dynamic adjustment algorithm; A resource scheduling scheme is generated based on the resource scheduling optimization model; wherein, the resource scheduling optimization model is constructed based on a matching algorithm between maintenance requirements and resource allocation.

9. The tunnel construction management method according to any one of claims 1 to 6, characterized in that, In the step of collecting tunnel construction parameters through a multi-source sensor network to obtain raw monitoring data, the short-range wireless communication is ZigBee communication or LoRa communication, and the heterogeneous data fusion transmission mechanism realizes the conversion of analog signals to digital signals through a unified interface.

10. A tunnel construction management system, characterized in that, Regarding the tunnel construction management method as described in any one of claims 1 to 9.

Images