Tunnel structure prediction method, system and device based on digital twinning and storage medium

Abstract

The invention discloses a tunnel structure prediction method, system and equipment based on digital twinning and a storage medium, and belongs to the technical field of power system energy storage equipment, and the method comprises the steps: collecting time sequence monitoring data of a tunnel structure, and carrying out the spatial stability test; constructing a longitudinal deformation curve, and calculating a plurality of structural deformation indexes; comparing the structural deformation index with a preset threshold value to generate a classification result of the structural state; and generating early warning information based on the classification result and historical data. According to the method, the continuous and smooth longitudinal deformation curve in the monitoring interval is constructed, the curve fitting precision is improved by encrypting the nodes in the local area, and the real deformation state of the tunnel structure can be accurately described; structural response parameters are extracted from derivative results, a multi-dimensional deformation evaluation index is constructed and compared with a preset threshold value, and accurate classification of structural states is achieved; according to the method, early warning can be triggered when any index is abnormal, and the recognition capability of structural abnormity under complex working conditions is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel engineering safety monitoring, and in particular to a tunnel structure prediction method, system, device and storage medium based on digital twins. Background Art

[0002] Tunnel construction in urban areas or areas with a high concentration of cultural relics often faces complex geological environments. The strata often contain fractured zones, aquifers, soft rock, and expansive soil layers, resulting in highly variable surrounding rock mass. Frequent adjustments to excavation and support plans are required during construction, which can easily lead to safety issues such as excessive surrounding rock deformation. Environmental and cultural heritage protection pressures also arise. During construction, surrounding rock deformation and blasting vibrations can significantly impact ancient buildings, surface roads, and surrounding structures. Failure to monitor and control these effects can lead to irreversible damage to cultural relics or structural safety hazards. Data processing efficiency is low. Discrete monitoring data requires complex numerical models to reconstruct deformation curves. Traditional polynomial or Gaussian curve fitting yields significant errors, especially in areas of localized deformation. Early warning accuracy is insufficient. Existing methods often rely on single-indicator threshold alarms, failing to fully incorporate the mechanical properties of tunnel structures. This leads to ambiguous disease location and a high rate of false alarms. Summary of the Invention

[0003] In view of the above-mentioned problems, the present invention is proposed.

[0004] Therefore, the technical problem solved by the present invention is: how to achieve high-precision modeling and response prediction of the longitudinal deformation of the tunnel structure under complex geological environments and strict protection requirements, improve the processing efficiency of monitoring data and the accuracy of early warning judgment, so as to effectively support structural health management and risk prevention and control during the construction process.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions: a tunnel structure prediction method based on digital twin, which includes:

[0006] Collect time-series monitoring data of tunnel structures and conduct spatial stability tests;

[0007] Based on the inspected data, a longitudinal deformation curve is constructed and several structural deformation indices are calculated;

[0008] By comparing the structural deformation index with the preset threshold, the classification result of the structural state is generated;

[0009] Generate early warning information based on classification results and historical data;

[0010] The warning information includes the structural risk level;

[0011] The deformation curve is completed by a fitting method, and the node distribution is adjusted in the selected area to form a high-order continuous curve.

[0012] As a preferred solution of the tunnel structure prediction method based on digital twins described in the present invention, the steps of collecting time series monitoring data of the tunnel structure and performing spatial stability testing include:

[0013] Several monitoring points with spatial numbers are arranged on the tunnel structure.

[0014] The three-dimensional coordinate data of the monitoring points are collected by monitoring equipment at several consecutive time points.

[0015] The collected data are organized into a data set with a time index in the order of monitoring time;

[0016] During the set time period, the three-dimensional coordinate changes of each monitoring point in the same measurement section are compared.

[0017] Identify the monitoring points whose position changes exceed the stability threshold in the current cycle and remove the data of the corresponding monitoring points from the data set.

[0018] As a preferred solution of the tunnel structure prediction method based on digital twins described in the present invention, wherein: the longitudinal deformation curve is constructed based on the tested data, including

[0019] Construct a data input sequence based on the spatial location of the monitoring points and the settlement values.

[0020] Determine the initial node distribution and boundary interpolation conditions based on the preset expression function form,

[0021] The expression curve structure is constructed and all parameter values ​​controlling the curve shape are obtained by solving the constraint equations, thus forming a continuous and smooth longitudinal deformation curve within the monitoring interval.

[0022] The beneficial effects of this preferred technical solution are: by constructing a data input sequence based on the spatial position of the monitoring point and the settlement value, the position-response correspondence in the original monitoring data can be fully retained, which is conducive to maintaining the engineering geometry authenticity during subsequent curve fitting. Furthermore, by adopting a preset expression function form (such as B-spline) and determining the initial node distribution and boundary interpolation conditions, a curve expression model with high-order continuity can be constructed, which has good local control capabilities while maintaining global smoothness. By solving the constraint equation to obtain all the parameter values ​​that control the shape of the curve, the constructed curve can accurately reconstruct the longitudinal deformation trend under given physical boundary conditions. Overall, the fitting process improves the model's ability to express settlement curves under complex geology and ensures the continuity and smoothness of deformation data in space.

[0023] As a preferred scheme of the tunnel structure prediction method based on digital twinning, the constructing the longitudinal deformation curve further comprises,

[0024] Identifying a region with a curve change degree greater than a set value in the monitoring interval, adding several additional nodes to construct a new node distribution;

[0025] Re-fitting the longitudinal deformation curve using the updated node distribution, and extracting the structure response parameters according to the derivative calculation processing at the node or monitoring point position determined by the re-fitted longitudinal deformation curve.

[0026] As a preferred scheme of the tunnel structure prediction method based on digital twinning, the comparing the structure deformation index with the preset threshold value,

[0027] Based on the structure response parameters, an evaluation index reflecting different deformation characteristics is constructed;

[0028] According to the comparison of different evaluation indexes with corresponding set threshold values, the classification to which the current structure state belongs is determined;

[0029] An abnormal state is generated under the condition that at least one index exceeds the preset threshold value.

[0030] The beneficial effects of the preferred technical scheme are: by constructing multiple evaluation indexes for reflecting different deformation characteristics from the structure response parameters, multi-dimensional discrimination of the tunnel structure health state can be realized without relying on a single judgment basis, effectively enhancing the comprehensiveness and robustness of structure state identification. After comparing each evaluation index with its corresponding set threshold value, the current structure state can be divided into a specific health classification, and the abnormal state indication mechanism is triggered when at least one index exceeds the threshold value, ensuring that the system can identify risks in the early abnormal stage of the structure.

[0031] As a preferred scheme of the tunnel structure prediction method based on digital twinning, the constructing the expression curve structure and obtaining all parameter values of the control curve shape by solving the constraint equation to form a continuous and smooth longitudinal deformation curve in the monitoring interval comprises: reconstructing the measured tunnel settlement data into a tunnel longitudinal deformation curve using a quartic B-spline fitting method, and the expression is:

[0032]

[0033] Where S(x) is the tunnel longitudinal deformation curve, c i is the control vertex, B i,4 (x) is the 4th order B-spline basis function;

[0034] Supplement interpolation nodes are extended at both ends of the monitoring interval, and constraint equations are required to pass through all interpolation nodes, a longitudinal deformation curve is refitted to form a continuous and smooth longitudinal deformation curve in the monitoring interval, and the expression is:

[0035]

[0036] Wherein, a j Supplement interpolation nodes are extended at both ends of the monitoring interval, m j is the settlement constraint value of the corresponding node, x k is the coordinate of the monitoring point, y k is the measured settlement value;

[0037] Derivative calculation is performed to extract the structural response parameters, the longitudinal deformation curvature radius is the basic geometric feature of the tunnel, the curvature radius of the continuous and smooth longitudinal deformation curve in the monitoring interval of the tunnel is the inverse of the curvature of each point on the longitudinal deformation curve, and the expression is:

[0038]

[0039] Wherein, R is the curvature radius of the continuous and smooth longitudinal deformation curve in the monitoring interval of the tunnel, S'(x) and S''(x) are the first and second derivatives of the continuous and smooth longitudinal deformation curve function in the monitoring interval, respectively.

[0040] As a preferred scheme of the tunnel structure prediction method based on digital twinning, wherein: the evaluation index reflecting different deformation characteristics is constructed based on the structural response parameters, including,

[0041] According to the calculated curvature radius of the continuous and smooth longitudinal deformation curve in the monitoring interval, the bending deformation mode and the dislocation deformation mode are evaluated and calculated for the tunnel structure health condition;

[0042] The bending deformation mode is expressed as:

[0043]

[0044] Wherein, Δ is the joint opening amount of the tunnel, l s is the ring width of the tunnel, D is the outer diameter of the tunnel, and R is the curvature radius of the longitudinal deformation curve of the tunnel;

[0045] The dislocation deformation mode is expressed as:

[0046]

[0047] Wherein, δ is the dislocation amount of the tunnel;

[0048] The joint opening amount evaluation index and the benching amount evaluation index of the tunnel are calculated according to the curvature radius of the continuous and smooth longitudinal deformation curve in the monitoring interval.

[0049] The three data are compared with the preset threshold value respectively, and if at least one index exceeds the preset threshold value, an abnormal state is generated.

[0050] The application provides a tunnel structure prediction system based on digital twinning.

[0051] To solve the above technical problems, the application provides the following technical scheme: a tunnel structure prediction system based on digital twinning, comprising a data collection and inspection module, a structure deformation index calculation module, a data comparison module and a warning module.

[0052] The data collection and inspection module collects time series monitoring data of the tunnel structure and performs spatial stability inspection.

[0053] The structure deformation index calculation module constructs a longitudinal deformation curve based on the inspected data and calculates a plurality of structure deformation indexes.

[0054] The data comparison module compares the structure deformation indexes with preset threshold values and generates a classification result of the structure state.

[0055] The warning module generates warning information based on the classification result and historical data.

[0056] The application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements the steps of the tunnel structure prediction method based on digital twinning when executing the computer program.

[0057] The application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program is executed by a processor to implement the steps of the tunnel structure prediction method based on digital twinning.

[0058] The application has the following beneficial effects: the longitudinal deformation curve is continuously and smoothly constructed in the monitoring interval, and the node in the local area is encrypted to improve the curve fitting precision, so that the real deformation state of the tunnel structure can be accurately described; the structure response parameters are extracted from the derivative results, the multi-dimensional deformation evaluation indexes are constructed, and the structure state is accurately classified by comparing with the preset threshold values; compared with the traditional method which depends on a single index, the application can trigger the warning when any index is abnormal, and the identification ability of the structure abnormality under complex working conditions is improved. BRIEF DESCRIPTION OF DRAWINGS

[0059] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0060] Figure 1 A general flowchart of a tunnel structure prediction method based on digital twinning is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0061] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort should be within the scope of protection of the present application.

[0062] Embodiment 1, refer to Figure 1 For an embodiment of the present application, the embodiment provides a tunnel structure prediction method based on digital twinning, comprising:

[0063] S1, collecting time series monitoring data of the tunnel structure and performing spatial stability test.

[0064] S2, constructing a longitudinal deformation curve based on the tested data, and calculating a plurality of structure deformation indexes.

[0065] S3, comparing the structure deformation indexes with the preset threshold value to generate a classification result of the structure state.

[0066] S4, generating an early warning information based on the classification result and historical data.

[0067] The early warning information includes a structure risk level;

[0068] The deformation curve is completed by a fitting method, and the node distribution is adjusted in a selected area to form a high-order continuous curve.

[0069] The present application aims to solve the problems of strong monitoring data dispersion, insufficient curve fitting accuracy, single deformation index, inaccurate disease positioning, high false alarm rate and the like in the process of traditional tunnel structure monitoring and early warning, especially in the areas with complex geological conditions or local severe deformation, the traditional method is difficult to realize effective identification and accurate early warning of the structure health state.

[0070] The present invention collects time-series monitoring data of tunnel structures, performs spatial stability testing and preprocessing, constructs a continuous and smooth longitudinal deformation curve, and improves fitting accuracy by encrypting nodes in areas with severe curvature. Structural response parameters are extracted from the fitting curve, and multiple structural deformation assessment indicators are constructed. These indicators are compared with corresponding preset thresholds to generate a structural status classification result. Then, combined with historical data, early warning information including the structural risk level and location is output.

[0071] This technical solution can improve the accuracy of fitting curves in local areas, enhance the dimensionality and pertinence of structural response assessments, and enable timely detection of structural anomalies.

[0072] Example 2 is an embodiment of the present invention, which provides a tunnel structure prediction method based on digital twins based on the previous embodiment, including:

[0073] The digital twin application system architecture is built through four core layers: the physical acquisition layer (which uses manual inspections and automated monitoring technologies to achieve real-time collection and status perception of tunnel structure data, and receives feedback for optimized maintenance decision-making instructions), the twin data layer (which builds a multi-dimensional feature mapping system based on tunnel mechanism research, historical data accumulation, and industry standards to provide a scientific benchmark for data processing and ensure the credibility of data and models), the data fusion layer (which cleans, stores, and standardizes heterogeneous data, relies on historical data analysis to evaluate the overall structural health of the tunnel, and ensures data consistency and availability), and the data application layer (which optimizes construction plans and dynamically monitors construction, uses high-precision deformation reconstruction technology to provide disease warnings and make disposal decisions, and establishes full-life cycle monitoring archives, forming a twin mutual feedback mechanism that runs through "data collection-analysis-decision-making-feedback", providing closed-loop technical support for the precise management of tunnel diseases). It plays an important role in tunnel monitoring and the efficient early warning of tunnel structure deformation in operation.

[0074] S1. Collect time-series monitoring data of the tunnel structure and conduct spatial stability tests.

[0075] Several monitoring points with spatial numbers are arranged on the tunnel structure.

[0076] The three-dimensional coordinate data of the monitoring points are collected by monitoring equipment at several consecutive time points.

[0077] The collected data are organized into a data set with a time index in the order of monitoring time;

[0078] During the set time period, the three-dimensional coordinate changes of each monitoring point in the same measurement section are compared.

[0079] Identify the monitoring points whose position changes exceed the stability threshold in the current cycle and remove the data of the corresponding monitoring points from the data set.

[0080] Automatic data collection is achieved by establishing a benchmark network and designing monitoring points; the collected data is transmitted to the twin data layer.

[0081] Based on the measurement automation monitoring system established in the physical acquisition layer, representative monitoring data types are selected in combination with industry standards and specifications to conduct further analysis of the longitudinal deformation curve of the tunnel structure, and the accuracy and efficiency of tunnel structure deformation early warning are improved through the application of the twin data layer.

[0082] Benchmarks are set up outside the deformation zone of the tunnel and at stable locations covering the deformation zone, and working base points are set up at locations with good visibility. Before each monitoring session, the benchmarks need to be periodically inspected using a total station to measure their three-dimensional coordinates and assess their stability. This is primarily to eliminate the impact of factors such as benchmark displacement on the monitoring data, safeguard the accuracy of the benchmark network, and thus ensure the accuracy and reliability of subsequent monitoring data analysis. Monitoring points are points that directly reflect the deformation of the subway tunnel. Monitoring sections are arranged at a certain distance at deformation-sensitive and complex stress-bearing locations within the deformation zone of the subway, and monitoring points are distributed according to the tunnel cross section.

[0083] S2. Construct a longitudinal deformation curve based on the inspected data and calculate several structural deformation indices.

[0084] Construct a data input sequence based on the spatial location of the monitoring points and the settlement values.

[0085] Determine the initial node distribution and boundary interpolation conditions based on the preset expression function form,

[0086] The expression curve structure is constructed and all parameter values ​​controlling the curve shape are obtained by solving the constraint equations, thus forming a continuous and smooth longitudinal deformation curve within the monitoring interval.

[0087] The constructing of the longitudinal deformation curve further comprises:

[0088] Identify areas within the monitoring interval where the curve change is greater than a set value, and add several additional nodes to construct a new node distribution;

[0089] The longitudinal deformation curve is refitted using the updated node distribution, and the structural response parameters are extracted based on the derivative calculation process at the node or monitoring point positions determined by the refitted longitudinal deformation curve.

[0090] In a preferred embodiment of the present invention, the longitudinal deformation curve is constructed based on the verified data. All parameter values ​​controlling the curve shape are obtained by solving constraint equations to form a continuous and smooth longitudinal deformation curve within the monitoring interval. This includes using a high-order, differentiable, quartic B-spline fitting method to accurately reconstruct the measured tunnel settlement data into the tunnel longitudinal deformation curve S(x). The principle is that the B-spline function constitutes a set of basic functions in the spline function space, which can effectively reconstruct discrete data into a continuous function, avoiding the error accumulation problem in solving the complex tunnel longitudinal settlement curve equation. The expression is:

[0091]

[0092] Among them, S(x) is the longitudinal deformation curve of the tunnel, c i To control the vertex, B i,4 (x) is the 4th-order B-spline basis function.

[0093] By controlling the vertex c i With the quartic B-spline basis function B i,4 (x) to construct the tunnel longitudinal deformation curve S(x). S(x) has high-order differentiability and local controllability, ensuring that the curve is continuous at the nodes and the curvature calculation accuracy meets engineering precision requirements, solving the edge distortion problem of traditional polynomial fitting.

[0094] In order to ensure the continuity and boundary smoothness of the curve, in addition to the known monitoring data points, interpolation conditions are added to form a set of equations.

[0095] The interpolation nodes are added at both ends of the monitoring interval and are required to pass through all interpolation nodes. The constraint equation is constructed and the longitudinal deformation curve is refitted to form a continuous and smooth longitudinal deformation curve within the monitoring interval. The expression is:

[0096]

[0097] Among them, a j To set up supplementary interpolation nodes at appropriate lengths at both ends of the monitoring interval, m j is the settlement constraint value of the corresponding node, which is used to control the smoothness of the curve boundary. k is the coordinate of the monitoring point, y k To measure the settlement value, ensure that the fitting curve passes through all the measured points and realize the continuous reconstruction of discrete data.

[0098] This ensures that the curve is continuous at the nodes and the curvature calculation accuracy is high, solves the edge distortion problem of traditional polynomial fitting, and provides a basis for subsequent calculations.

[0099] Constructing the longitudinal deformation curve based on the tested data In an optional embodiment of the present invention, the longitudinal deformation curve is constructed using a multi-segment least squares fitting method based on a sliding window:

[0100] After completing the spatial stability test of the monitoring data, the monitoring points arranged along the tunnel axis are sorted in spatial order, and a sliding interval of fixed length is set (for example, 5 to 10 points form a section). The three-dimensional coordinate data of the monitoring points and the settlement values ​​within each sliding interval constitute the fitting input data set.

[0101] Within each sliding interval, a quadratic or cubic polynomial is fitted using the least squares method to obtain the curve expression for the corresponding segment. The function value and first-order derivative of the fitted function at the interval endpoints are recorded. Subsequently, endpoint interpolation and smooth connection strategies are used to construct a longitudinal continuous curve for the entire monitoring interval.

[0102] The output curve of this method has a first-order derivative result at each monitoring point, which can be used to subsequently extract structural response parameters (such as curvature and offset) and enter the structural index calculation process.

[0103] This method is particularly suitable for construction scenarios where the monitoring points are evenly distributed and require low-latency online updates. It can achieve rapid reconstruction and local updates, making it easy to update prediction results in real time after data access.

[0104] An optional embodiment of the present invention is to construct a longitudinal deformation curve based on the tested data:

[0105] The Gaussian process regression (GPR) method was used to construct the longitudinal deformation curve.

[0106] After preprocessing the monitoring data, the lateral positions and corresponding settlement values ​​of all valid monitoring points are input into the Gaussian process regression model. As a nonparametric fitting model, GPR can adaptively construct a kernel function based on the training data to estimate the deformation trend value and prediction confidence interval at any location.

[0107] The GPR model used in this embodiment has preset kernel function types and hyperparameters (such as RBF kernel). The continuous function of the fitting output can calculate the derivative results at any point for the subsequent structural response parameter extraction step.

[0108] It should be further explained that the beneficial effects of the current preferred embodiment are: by using a quartic B-spline function with high-order differentiability and local control characteristics to fit and reconstruct the measured settlement data, the accuracy of the curvature calculation can be improved while maintaining the continuity of the curve, and the fluctuation distortion problem that is prone to occur in the boundary area of ​​traditional polynomial fitting can be effectively avoided; by supplementing interpolation nodes outside the monitoring interval and constructing a set of constraint equations, the fitting curve meets the interpolation accuracy requirements at all monitoring points, further improving the smoothness and stability of the curve in the boundary area.

[0109] Although the quartic B-spline fitting can make up for this defect by encrypting the nodes of the overall curve, it can also improve the representation ability of local deformation by utilizing the local controllability of B-spline.

[0110] First, the p-order derivative of the original data is calculated by the central difference method to identify the high complexity area, where p-1 is the degree of the B-spline and is defined at the node parameter position by the formula The feature point set {f i}, the expression is:

[0111]

[0112] For node x, where Local node encryption follows the formula f(X), which is expressed as:

[0113]

[0114] The characteristic function f(x) represents the amount of detail at a node position x. Higher f(x) values ​​correspond to smaller node spans at that location. Based on the geometric characteristics of the curve, the local node refinement method adjusts node positions, identifies high-complexity sections of the curve, and allocates more nodes to these areas. The fitted curve S(x), which is refined by estimating the curve's derivative characteristics, achieves higher resolution in areas with severe local deformation. This effectively reduces fitting errors in local areas caused by local misalignment and excessive settlement, improving curve reconstruction accuracy.

[0115] The curvature radius of longitudinal deformation is the basic geometric characteristic of the tunnel. The curvature radius R of the tunnel longitudinal deformation curve is the reciprocal of the curvature kc at each point on the longitudinal deformation curve.

[0116]

[0117] Where R is the curvature radius of the continuous and smooth longitudinal deformation curve in the monitoring section of the tunnel, S′(x) and S″(x) are the first and second order derivatives of the continuous and smooth longitudinal deformation curve function in the monitoring section, respectively.

[0118] S3. Compare the structural deformation index with the preset threshold to generate a classification result of the structural state.

[0119] Based on the structural response parameters, evaluation indicators reflecting different deformation characteristics are constructed;

[0120] Compare different evaluation indicators with corresponding set thresholds to determine the classification to which the current structural state belongs;

[0121] When at least one indicator exceeds the preset threshold, an abnormal state is generated.

[0122] The preferred embodiment of the present invention is to compare the structural deformation index with the preset threshold value, and to evaluate and calculate the health status of the tunnel structure by performing bending deformation mode and staggered deformation mode according to the curvature radius of the continuous and smooth longitudinal deformation curve in the monitoring interval obtained by calculation;

[0123] Bending deformation mode, the expression is:

[0124]

[0125] Where Δ is the tunnel joint opening, l s is the tunnel ring width, D is the tunnel outer diameter, and R is the curvature radius of the tunnel longitudinal deformation curve;

[0126] The staggered deformation mode is expressed as:

[0127]

[0128] Where, δ is the tunnel stagger;

[0129] The tunnel joint opening and stagger evaluation indices are calculated based on the curvature radius of the continuous and smooth longitudinal deformation curve within the monitoring interval.

[0130] The three data are compared with the corresponding preset thresholds respectively. If at least one indicator exceeds the preset threshold, an abnormal state is generated.

[0131] An optional embodiment of the present invention is to compare the structural deformation index with the preset threshold value: within the monitoring interval, calculate the numerical variation of the first-order derivative and the second-order derivative of the longitudinal deformation curve at each monitoring point, and obtain the gradient change rate and curvature change coefficient reflecting the local structural mutation trend. These two parameters are compared with the upper and lower limit intervals extracted from the historical healthy sample data to construct the judgment criteria:

[0132] When the gradient change rate (i.e., the absolute value of the change in adjacent derivatives) is greater than the set threshold; or the curvature change coefficient (i.e., the difference between adjacent second-order derivatives) is greater than the set threshold; the monitoring point is marked as a local outlier.

[0133] If two or more continuous monitoring points are marked as abnormal, it is determined that there is a potential structural risk in the corresponding tunnel section, and warning information is generated.

[0134] In an optional embodiment of the present application, a feature vector group (such as longitudinal curvature, node settlement gradient, historical change trend, etc.) is constructed based on the structural response parameters, and the feature vector is input into a state recognition model that has been trained based on the K-Means clustering algorithm. The specific processing flow is as follows: a structural parameter feature vector is formed every monitoring period, including the curvature radius, the displacement gradient of the joint, and the historical trend residual; the structural parameter feature vector is compared with the cluster centers formed in the training phase in terms of Euclidean distance, and if the current vector distance from the "abnormal state cluster center" is less than the set threshold, it is classified as an abnormal state.

[0135] It is further explained that the beneficial effect of the current preferred embodiment is to construct a structural evaluation index based on the curvature radius of the longitudinal deformation curve, to clearly define the corresponding bending deformation mode and joint deformation mode, and to convert them into actual engineering quantities reflecting the degree of structural deformation such as joint opening and joint displacement, so as to realize the quantitative transformation from geometric shape change to structural risk index. Compared with the traditional method of relying only on absolute settlement for judgment, the present embodiment can comprehensively reflect the continuity distortion and segmental displacement of the structure, and improve the pertinence and practicality of classification and judgment.

[0136] S4, generating warning information based on the classification result and historical data.

[0137] The warning information includes a structural risk level;

[0138] The deformation curve is completed by a fitting method, and the node distribution is adjusted in the selected area to form a high-order continuous curve.

[0139] Embodiment 3 is an embodiment of the present application, which provides a tunnel structure prediction system based on digital twinning, including a collected data preprocessing module, a structural deformation index calculation module, a data comparison module, and a warning module.

[0140] The collected data preprocessing module is to collect time series monitoring data of the tunnel structure, and to perform spatial stability test and data preprocessing;

[0141] The structural deformation index calculation module is to construct a longitudinal deformation curve based on the preprocessed data, and to calculate a plurality of structural deformation indexes;

[0142] The data comparison module is to compare the structural deformation index with a preset threshold to generate a classification result of the structural state;

[0143] The warning module is to generate warning information based on the classification result and historical data.

[0144] This embodiment also provides an electronic device, which is suitable for a tunnel structure prediction method based on digital twins, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement a tunnel structure prediction method based on digital twins as proposed in the above embodiment.

[0145] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, a tunnel structure prediction method based on digital twins as proposed in the above embodiment is implemented.

[0146] The storage medium proposed in this embodiment and the method for implementing a tunnel structure prediction method based on digital twin proposed in the above embodiment belong to the same inventive concept. Technical details not fully described in this embodiment can be found in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0147] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general-purpose hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer's floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0148] It should be noted that the above embodiments are only used to illustrate the technical solutions 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 preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A tunnel structure prediction method based on digital twins, characterized by: include, Collect time-series monitoring data of tunnel structures and conduct spatial stability tests; Based on the inspected data, a longitudinal deformation curve is constructed and several structural deformation indices are calculated; By comparing the structural deformation index with the preset threshold, the classification result of the structural state is generated; Generate early warning information based on classification results and historical data; The warning information includes the structural risk level; The deformation curve is completed by a fitting method, and the node distribution is adjusted in the selected area to form a high-order continuous curve.

2. The tunnel structure prediction method based on digital twin according to claim 1, characterized in that: The collecting of time series monitoring data of the tunnel structure and the performing of spatial stability test include: Several monitoring points with spatial numbers are arranged on the tunnel structure. The three-dimensional coordinate data of the monitoring points are collected by monitoring equipment at several consecutive time points. The collected data are organized into a data set with a time index in the order of monitoring time; During the set time period, the three-dimensional coordinate changes of each monitoring point in the same measurement section are compared. Identify the monitoring points whose position changes exceed the stability threshold in the current cycle and remove the data of the corresponding monitoring points from the data set.

3. The tunnel structure prediction method based on digital twin according to claim 2, characterized in that: The longitudinal deformation curve is constructed based on the tested data, including Construct a data input sequence based on the spatial location of the monitoring points and the settlement values. Determine the initial node distribution and boundary interpolation conditions based on the preset expression function form, The expression curve structure is constructed and all parameter values ​​controlling the curve shape are obtained by solving the constraint equations, thus forming a continuous and smooth longitudinal deformation curve within the monitoring interval.

4. The tunnel structure prediction method based on digital twin according to claim 3, characterized in that: The constructing of the longitudinal deformation curve further comprises: Identify areas within the monitoring interval where the curve change is greater than a set value, and add several additional nodes to construct a new node distribution; The longitudinal deformation curve is refitted using the updated node distribution, and the structural response parameters are extracted based on the derivative calculation process at the node or monitoring point positions determined by the refitted longitudinal deformation curve.

5. The tunnel structure prediction method based on digital twin according to claim 4, characterized in that: The structural deformation index is compared with a preset threshold value. Based on the structural response parameters, evaluation indicators reflecting different deformation characteristics are constructed; Compare different evaluation indicators with corresponding set thresholds to determine the classification to which the current structural state belongs; When at least one indicator exceeds the preset threshold, an abnormal state is generated.

6. The tunnel structure prediction method based on digital twin according to claim 5, characterized in that: The method of constructing the expression curve structure and obtaining all parameter values ​​of the control curve shape by solving the constraint equation to form a continuous and smooth longitudinal deformation curve within the monitoring interval includes reconstructing the measured tunnel settlement data into the tunnel longitudinal deformation curve using a quartic B-spline fitting method, and the expression is: Among them, S(x) is the longitudinal deformation curve of the tunnel, c i To control the vertex, B i,4 (x) is the 4th-order B-spline basis function; The interpolation nodes are added at both ends of the monitoring interval and are required to pass through all interpolation nodes. The constraint equation is constructed and the longitudinal deformation curve is refitted to form a continuous and smooth longitudinal deformation curve within the monitoring interval. The expression is: Among them, a j To set up supplementary interpolation nodes at appropriate lengths at both ends of the monitoring interval, m j is the settlement constraint value of the corresponding node, x k is the coordinate of the monitoring point, y k is the measured settlement value; Derivative calculations are performed to extract structural response parameters. The longitudinal deformation curvature radius is a basic geometric feature of the tunnel. The curvature radius of the continuous and smooth longitudinal deformation curve within the tunnel monitoring section is the inverse of the curvature at each point on the longitudinal deformation curve, and the expression is: Where R is the curvature radius of the continuous and smooth longitudinal deformation curve in the monitoring section of the tunnel, S(x) and S″(x) are the first and second order derivatives of the continuous and smooth longitudinal deformation curve function in the monitoring section, respectively.

7. The tunnel structure prediction method based on digital twin according to claim 6, characterized in that: The evaluation indicators reflecting different deformation characteristics based on the structural response parameters include: The health status of the tunnel structure is evaluated and calculated based on the bending deformation mode and the staggered deformation mode according to the curvature radius of the continuous and smooth longitudinal deformation curve within the monitoring interval; Bending deformation mode, the expression is: Where Δ is the tunnel joint opening, l s is the tunnel ring width, D is the tunnel outer diameter, and R is the curvature radius of the tunnel longitudinal deformation curve; The staggered deformation mode is expressed as: Where, δ is the tunnel stagger; The tunnel joint opening and stagger evaluation indices are calculated based on the curvature radius of the continuous and smooth longitudinal deformation curve within the monitoring interval. The three data are compared with the corresponding preset thresholds respectively. If at least one indicator exceeds the preset threshold, an abnormal state is generated.

8. A tunnel structure prediction system based on digital twin, applying the tunnel structure prediction method based on digital twin according to any one of claims 1 to 7, characterized in that: It includes: data collection and verification module, structure deformation index calculation module, data comparison module and early warning module; The data collection and verification module collects time-series monitoring data of the tunnel structure and performs spatial stability verification; The module for calculating structural deformation index constructs a longitudinal deformation curve based on the inspected data and calculates several structural deformation indexes; The data comparison module generates a classification result of the structural state by comparing the structural deformation index with a preset threshold; The early warning module generates early warning information based on classification results and historical data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the processor implements the steps of a tunnel structure prediction method based on digital twins according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of a tunnel structure prediction method based on digital twins according to any one of claims 1 to 7 are implemented.

Images