Tunnel lining health state analysis method and system based on field monitoring data

Abstract

The application discloses a tunnel lining health state analysis method and system based on field monitoring data, and the method comprises the following steps: obtaining net displacement and net acoustic emission data through temperature deformation correction and thermal noise elimination; constructing a damage evolution constitutive model with a damage variable as an internal variable, performing finite element inversion on the net displacement and net acoustic emission data as boundary conditions and energy constraints, and solving a damage field distribution; identifying a continuous region with a damage variable exceeding a threshold value as an active damage area, extracting geometric center coordinates and an area of the active damage area; fitting a migration trajectory and an area expansion curve of a damage core along a time axis; extrapolating the trajectory and the curve to predict a future damage position and range, and dividing three stages of stable expansion, accelerated expansion and instability damage precursor according to a change rate of a slope of the expansion curve, and correspondingly generating a graded warning. The application realizes accurate quantification and space-time evolution prediction of internal damage of a tunnel lining, and significantly improves an intelligent level of health state evaluation.

Description

Technical Field

[0001] This invention belongs to the field of tunnel engineering structural health monitoring technology, and particularly relates to a method and system for analyzing the health status of tunnel lining based on on-site monitoring data. Background Technology

[0002] As the main load-bearing and protective component of a tunnel structure, the health status of the tunnel lining directly affects the tunnel's operational safety and service life. During long-term service, the lining structure is affected by multiple factors such as surrounding rock pressure, temperature changes, and material aging, leading to accumulated damage and even failure. Therefore, real-time health monitoring and condition assessment of tunnel linings are of significant engineering importance.

[0003] Currently, tunnel lining health monitoring mainly relies on sensors such as displacement gauges, strain gauges, and crack gauges to acquire structural deformation data and trigger alarms by setting thresholds. However, existing methods have the following shortcomings: First, monitoring data is easily affected by changes in ambient temperature; thermal expansion and contraction caused by temperature often mask the actual damage deformation, leading to misjudgments or missed detections. Second, relying solely on deformation data is insufficient to detect the initiation and propagation process of microcracks within the lining, making early warning impossible. Third, the lack of systematic analysis of the spatiotemporal laws governing damage evolution makes it difficult to predict damage development trends and potential failure modes. Therefore, there is an urgent need for a health status analysis method that can integrate multi-source monitoring data, eliminate environmental interference, reveal the internal damage evolution laws, and achieve state prediction. Summary of the Invention

[0004] This invention provides a method and system for analyzing the health status of tunnel linings based on field monitoring data, aiming to solve the technical problems in the prior art, such as monitoring data being affected by temperature, inability to detect internal damage evolution, and lack of trend prediction capabilities.

[0005] In a first aspect, the present invention provides a method for analyzing the health status of tunnel lining based on field monitoring data, comprising:

[0006] Acquire multi-point displacement gauge monitoring data sequence, multi-channel acoustic emission monitoring data sequence, and temperature data sequence of internal temperature measuring points of the tunnel lining within a preset time period;

[0007] The temperature change at each moment is calculated based on the temperature data sequence. Based on the preset thermal expansion coefficient of the lining material, the temperature deformation correction is applied to the multi-point displacement gauge monitoring data sequence to obtain the net displacement data sequence after removing the temperature influence.

[0008] Based on the temperature data sequence, identify the time periods where the temperature change rate exceeds a preset threshold, and remove the acoustic emission signals that are interfered with by thermal noise in the corresponding time periods to obtain the net acoustic emission data sequence.

[0009] A constitutive model of damage evolution of lining material is constructed with damage variables as internal variables. The constitutive model of damage evolution includes the nonlinear relationship between damage driving force and damage variable increment.

[0010] The net displacement data sequence is used as the displacement boundary condition, and the net acoustic emission data sequence is used as the energy dissipation constraint. These are input into the damage evolution constitutive model for finite element inversion calculation to obtain the damage variable values ​​of each element inside the lining at the current moment, thus forming the damage field distribution.

[0011] Based on the damage field distribution, continuous spatial regions where the damage variable values ​​exceed a preset damage threshold are identified as active damage regions, and the geometric center coordinates and area of ​​the active damage regions at each time point are extracted.

[0012] By fitting the geometric center coordinates and area of ​​the active damage region along the time axis, the spatial migration trajectory of the center of the active damage region and the expansion curve of the area over time are obtained.

[0013] Based on the migration trajectory and the expansion curve, the health status assessment results of the active damage area within a future preset time window are predicted.

[0014] Secondly, the present invention provides a tunnel lining health status analysis system based on on-site monitoring data, comprising:

[0015] The acquisition module is configured to acquire multi-point displacement gauge monitoring data sequences, multi-channel acoustic emission monitoring data sequences, and temperature data sequences of internal temperature measuring points of the tunnel lining within a preset time period.

[0016] The correction module is configured to calculate the temperature change at each moment based on the temperature data sequence, and perform temperature deformation correction on the multi-point displacement gauge monitoring data sequence based on the preset thermal expansion coefficient of the lining material to obtain the net displacement data sequence after removing the temperature influence.

[0017] The elimination module is configured to identify time periods where the rate of temperature change exceeds a preset threshold based on the temperature data sequence, and eliminate acoustic emission signals that are interfered with by thermal noise in the corresponding time period to obtain a net acoustic emission data sequence.

[0018] The construction module is configured to construct a constitutive model of lining material damage evolution with damage variables as internal variables. The constitutive model of damage evolution includes a nonlinear relationship between damage driving force and damage variable increment.

[0019] The output module is configured to use the net displacement data sequence as the displacement boundary condition and the net acoustic emission data sequence as the energy dissipation constraint, and input them into the damage evolution constitutive model for finite element inversion calculation to solve for the damage variable values ​​of each unit inside the lining at the current moment, thus forming the damage field distribution.

[0020] The extraction module is configured to identify continuous spatial regions where the damage variable value exceeds a preset damage threshold based on the damage field distribution, as active damage regions, and extract the geometric center coordinates and area of ​​the active damage regions at each time point.

[0021] The fitting module is configured to fit the geometric center coordinates and area of ​​the active damage region along the time axis to obtain the spatial migration trajectory of the center of the active damage region and the expansion curve of the area over time.

[0022] The prediction module is configured to predict the health status assessment results of the active damage area within a future preset time window based on the migration trajectory and the expansion curve.

[0023] Thirdly, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the tunnel lining health status analysis method based on field monitoring data according to any embodiment of the present invention.

[0024] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the steps of the tunnel lining health status analysis method based on field monitoring data according to any embodiment of the present invention.

[0025] This application presents a method and system for analyzing the health status of tunnel linings based on field monitoring data. It obtains net displacement and net acoustic emission data through temperature deformation correction and thermal noise removal; constructs a damage evolution constitutive model with damage variables as internal variables; uses net displacement and net acoustic emission data as boundary conditions and energy constraints for finite element inversion to solve for the damage field distribution; identifies continuous regions where damage variables exceed thresholds as active damage areas, and extracts their geometric center coordinates and area; fits the migration trajectory and area expansion curve of the damage core along the time axis; extrapolates the trajectory and curve to predict future damage location and extent; and divides the expansion curve into three stages—stable expansion, accelerated expansion, and instability precursor—based on the slope change rate, generating corresponding graded early warnings. This achieves accurate quantification and spatiotemporal evolution prediction of internal damage in tunnel linings, significantly improving the intelligence level of health status assessment. Attached Figure Description

[0026] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 A flowchart illustrating a method for analyzing the health status of tunnel lining based on field monitoring data, provided in an embodiment of the present invention;

[0028] Figure 2 This is a structural block diagram of a tunnel lining health status analysis system based on field monitoring data provided in an embodiment of the present invention;

[0029] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Please see Figure 1 The flowchart illustrates the tunnel lining health status analysis method based on field monitoring data according to this application.

[0032] like Figure 1 As shown, the method for analyzing the health status of tunnel lining based on field monitoring data specifically includes the following steps:

[0033] Step S101: Obtain the multi-point displacement gauge monitoring data sequence, the multi-channel acoustic emission monitoring data sequence, and the temperature data sequence of the internal temperature measuring points of the tunnel lining within a preset time period.

[0034] In this step, multi-point displacement gauges are used to monitor the relative displacement changes at different depths of the tunnel lining, reflecting the internal deformation characteristics of the lining.

[0035] Sensor type: Use vibrating wire or inductive multi-point displacement gauge. The range is determined according to the lining thickness and expected deformation (usually 50-100mm). The accuracy is not less than 0.1%FS.

[0036] Location of installation: A monitoring section is set up every 20-30 meters along the longitudinal direction of the tunnel; multiple displacement gauges are installed at key locations such as the arch crown, left and right arch waists, and left and right sidewalls at each monitoring section; each multi-point displacement gauge contains 3-5 measuring points, located on the lining surface, the middle of the lining, and near the contact surface with the surrounding rock.

[0037] Acoustic emission sensors are used to capture elastic wave signals released during the initiation and propagation of microcracks inside the lining, enabling real-time monitoring of the damage process.

[0038] Sensor type: Wideband resonant acoustic emission sensor with a frequency response range of 50kHz-400kHz and built-in preamplifier (gain 40dB).

[0039] Deployment location: Deployed in the same cross section as the multi-point displacement gauge, with 4-8 acoustic emission sensors deployed in each cross section; the sensors should cover key areas such as the arch crown, arch waist, and side walls, and be deployed more densely in areas with known defects; the spacing between adjacent sensors is determined according to the material attenuation characteristics (usually 1-3 meters) to ensure effective signal coverage.

[0040] Temperature sensors are used to monitor changes in the internal temperature field of the lining, providing basic data for temperature deformation correction.

[0041] Sensor type: High-precision platinum resistance temperature sensor (PT100 / PT1000) is selected, with an accuracy of ±0.1℃ and a measurement range of -20℃ to +80℃.

[0042] Installation location: Installed in the same hole as the multi-point displacement gauges, and temperature sensors are installed near the measuring points of each multi-point displacement gauge; they are installed at different depths on the lining surface and inside to form a temperature gradient monitoring profile.

[0043] Step S102: Calculate the temperature change at each moment based on the temperature data sequence, and perform temperature deformation correction on the multi-point displacement gauge monitoring data sequence based on the preset thermal expansion coefficient of the lining material to obtain the net displacement data sequence after removing the temperature influence.

[0044] In this step, the measured displacement values ​​at each moment in the multi-point displacement gauge monitoring data sequence, and the measured temperature values ​​at the corresponding temperature measurement points inside the lining, are obtained. Based on the difference between the measured temperature values ​​and the preset reference temperature values, the temperature change at each moment is calculated. According to the product of the preset thermal expansion coefficient of the lining material and the temperature change, the thermal deformation component caused by the temperature change at each moment is calculated. The measured displacement value at each moment is subtracted from the thermal deformation component at the corresponding moment to obtain the net displacement value at each moment. The net displacement values ​​at each moment are sorted in chronological order to obtain the net displacement data sequence.

[0045] In one specific embodiment, tunnel lining materials (concrete, reinforced concrete, etc.) exhibit the physical property of thermal expansion and contraction. When the ambient temperature or the internal temperature of the lining changes, the material experiences thermal strain, causing changes in displacement gauge readings. This thermal deformation is elastically reversible and does not reflect actual damage or stress deformation of the structure. According to the theory of linear thermal expansion, the thermal deformation component can be expressed as:

[0046] ,

[0047] In the formula, This refers to the amount of thermal deformation caused by temperature changes. The linear thermal expansion coefficient of the lining material (unit: / ℃). This refers to the initial gauge length of the displacement gauge or the initial distance between measuring points. This represents the temperature change relative to a reference temperature.

[0048] The temperature change at each moment is calculated by measuring temperature monitoring data, and the thermal deformation component is estimated by combining the material's thermal expansion coefficient. This component is then subtracted from the measured displacement to obtain the net displacement that reflects the true mechanical behavior of the structure.

[0049] Specifically, firstly, the displacement and temperature values ​​at the same time are extracted from the monitoring data sequence generated in step S101.

[0050] For each sampling time Obtain: Measured displacement value Data is from a multi-point displacement gauge monitoring sequence, in millimeters (mm).

[0051] Measured temperature values: derived from temperature data sequences, in degrees Celsius (°C). If multiple temperature measurement points exist at the same time, the temperature sensor data closest in space is selected based on the displacement gauge measurement point location; if no data is available at exactly the same time, a linear interpolation method can be used to obtain the temperature value at the corresponding moment.

[0052] Reference temperature It serves as a reference point for calculating temperature changes, using the average of all temperature data during the monitoring period as the baseline temperature. This method can reduce systematic biases caused by the selection of the baseline.

[0053] Based on the established reference temperature, calculate the temperature change at each moment:

[0054] ,

[0055] In the formula, Let be the measured temperature value at the i-th sampling time. The temperature change at the i-th sampling time can be positive (heating up) or negative (cooling down), and the unit is ℃.

[0056] The key to calculating the thermal deformation components caused by temperature changes at various times lies in determining the initial gauge length of the displacement gauge. and the coefficient of thermal expansion of materials .

[0057] Coefficient of thermal expansion of material The value of directly affects the correction accuracy. It can be obtained through the following methods:

[0058] Indoor test calibration: Thermal expansion coefficient tests were conducted on concrete specimens cast simultaneously on-site. Following the specifications of the "Standard for Test Methods of Long-Term Performance and Durability of Ordinary Concrete" (GB / T 50082), the length changes of the specimens at different temperatures were measured, and the linear expansion coefficient was calculated. (General concrete...) The value range is .

[0059] Empirical values: If no testing conditions are available, empirical values ​​can be obtained by referring to relevant specifications. For ordinary concrete linings, Desirable For reinforced concrete, considering the coefficient of thermal expansion of steel (approximately...), (This can be adjusted appropriately.)

[0060] Inversion calibration: Select a period of significant temperature change but no load change (such as nighttime when there is no vehicle traffic), and invert the calibration by analyzing the correlation between displacement change and temperature change. value.

[0061] For different types of multi-point displacement gauges The methods for determining them are different:

[0062] Single-point displacement meter: The initial distance between the measuring point and the reference point during sensor installation is usually determined by the installation record.

[0063] Multi-point displacement gauge: Each measuring point corresponds to a different gauge length. For example, at a depth of... The measuring points, This refers to the distance from the measuring point to the orifice (for absolute displacement measurement) or the distance between adjacent measuring points (for relative displacement measurement). The specific value should be determined based on the type of displacement gauge and the measurement principle.

[0064] In this embodiment, for each measuring point of the multi-point displacement meter Its gauge length is denoted as .

[0065] For time measuring points The heat distortion component is:

[0066] ,

[0067] In the formula, The calculation is performed using the temperature sensor data corresponding to the measurement point. If there are multiple temperature measurement points on the same cross-section, spatial interpolation can be used to obtain the temperature change at the measurement point location.

[0068] Subtract the thermal deformation component from the measured displacement value to obtain the net displacement value:

[0069] ,

[0070] In the formula, For a moment measuring point The measured displacement value, This is the corrected net displacement value;

[0071] It should be noted that if the direction of the thermal deformation component is inconsistent with the direction of the displacement gauge measurement, the vector projection relationship needs to be considered. For displacement gauges installed radially along the lining, thermal deformation mainly occurs in the radial direction, and can be directly subtracted algebraically.

[0072] Repeat the above calculation for all times at each measuring point to obtain the net displacement data sequence for that measuring point.

[0073] Step S103: Identify the time period in which the temperature change rate exceeds a preset threshold based on the temperature data sequence, and remove the acoustic emission signals that are interfered with by thermal noise in the corresponding time period to obtain the net acoustic emission data sequence.

[0074] In this step, the measured temperature values ​​of adjacent moments in the temperature data sequence are obtained, and the temperature change rate at each moment is calculated. The temperature change rate at each moment is compared with a preset temperature change rate threshold, and continuous time periods where the temperature change rate exceeds the preset threshold are identified as thermal noise interference periods. In the multi-channel acoustic emission monitoring data sequence, acoustic emission signal segments corresponding to the thermal noise interference periods are searched. The found acoustic emission signal segments are removed from the multi-channel acoustic emission monitoring data sequence, and the acoustic emission signals of the remaining time periods are retained. The retained acoustic emission signals are subjected to time-series splicing and resampling processing to obtain the net acoustic emission data sequence.

[0075] In one specific embodiment, the measured temperature values ​​at each time point are obtained from the temperature data sequence generated in step S101. For discrete sampling points, the rate of temperature change is defined as the ratio of the temperature difference between adjacent time points to the time interval.

[0076] Let the temperature data sequence be The time interval between adjacent moments is For time rate of temperature change The calculation is as follows:

[0077] ,

[0078] In the formula, For a moment Temperature value, For a moment Temperature value;

[0079] For the first moment Defineable Alternatively, backward difference can be used;

[0080] In practical engineering, the sampling interval may not be constant, so it needs to be calculated based on the actual time difference. The unit of temperature change rate is ℃ / min or ℃ / s, which can be selected according to the monitoring requirements;

[0081] Set a temperature change rate threshold This is used to determine whether temperature fluctuations are drastic. Methods for determining the threshold include:

[0082] Statistical method: Based on historical temperature data, calculate the statistical distribution of the rate of temperature change, and take a certain quantile (such as the 95th quantile) as the threshold.

[0083] Empirical method: Settings are based on sensor performance and field experience. For common tunnel environments, the following settings can be configured: =0.5℃ / min or 0.1℃ / 30s.

[0084] Experimental calibration method: During periods without actual acoustic emission events (such as nighttime shutdown periods), record the relationship between the rate of temperature change and the background noise of acoustic emission, and determine a reasonable threshold through correlation analysis.

[0085] The recognition process is as follows:

[0086] Traverse all moments Mark all that meet the criteria. At that moment.

[0087] By merging consecutively marked moments, several thermal noise interference periods are formed. Each period is defined by its starting moment. and end time definition.

[0088] To avoid edge effects, the identified time period can be appropriately extended outward by a buffer window (e.g., adding 10 seconds before and after) to completely cover the acoustic emission signals that may be affected.

[0089] Obtain the multi-channel acoustic emission monitoring data sequence generated in step S101. This sequence contains the timestamp, channel number, characteristic parameters (amplitude, energy, ring count, etc.) and optional waveform data for each acoustic emission event.

[0090] For each identified thermal noise interference period Perform the following operations:

[0091] Traverse the acoustic emission data sequence and find all events whose timestamps fall within that time period.

[0092] Record the timestamps, channel numbers, and corresponding feature parameters of these events as candidate signals to be eliminated.

[0093] If the system also stores waveform data, it can further verify whether the waveform characteristics (such as frequency distribution and attenuation characteristics) of these signals conform to the typical characteristics of thermal noise, so as to reduce the risk of false rejection.

[0094] Remove all acoustic emission events belonging to periods of thermal noise interference from the original multi-channel acoustic emission monitoring data sequence. This can be done using one of the following two methods:

[0095] Physical deletion: Directly removes the marked event records and generates a new data sequence that does not contain these events.

[0096] Ignore flag: Add a "validity" flag to the original data (e.g., 0 indicates invalid, 1 indicates valid), and only events with a valid flag of 1 will be processed during subsequent analysis.

[0097] Regardless of the method used, it is essential to ensure that the temporal order of the preserved acoustic emission events remains unchanged, and that the data from each channel is processed independently without aliasing.

[0098] After removing periods of thermal noise, discontinuities may appear in the acoustic emission data on the time axis. To facilitate subsequent analysis (such as energy accumulation and spectrum analysis), the retained signals need to be time-series spliced ​​and resampled.

[0099] Time-series splicing: The retained acoustic emission events are rearranged in their original chronological order to form a continuous event sequence. No events are inserted for missing time periods to maintain data sparsity.

[0100] In this embodiment, drastic temperature changes in the tunnel environment (such as ventilation system startup and shutdown, seasonal changes, and diurnal temperature variations) can trigger thermal noise in the sensors. The amplitude of this noise is often comparable to that of actual microcrack signals, severely contaminating the data. This step utilizes the rate of temperature change as a sensitive indicator of thermal noise, accurately identifying and eliminating interfering periods. This significantly increases the proportion of genuine damage events in the retained acoustic emission signals. Furthermore, traditional acoustic emission monitoring systems often generate numerous false alarms due to their inability to distinguish between thermal noise and real signals, misleading operational decisions. This step, through physical mechanism-driven thermal noise elimination, eliminates temperature interference at its source, allowing subsequent damage identification algorithms to focus solely on genuine microcrack activity, significantly reducing the system's false alarm rate.

[0101] Step S104: Construct a constitutive model of lining material damage evolution with damage variables as internal variables. The constitutive model of damage evolution includes the nonlinear relationship between damage driving force and damage variable increment.

[0102] In this step, a damage variable D is defined to describe the degree of deterioration of the lining material under load. The value of the damage variable D ranges from 0 to 1, where D=0 indicates that the material is intact and D=1 indicates that the material has completely failed. A nonlinear evolution equation is established between the damage driving force and the increment of the damage variable. The damage driving force is calculated based on the current stress state and the cumulative damage degree. The nonlinear evolution equation is in the form of dD / dt=f(σ,D,T), where σ is the current stress, D is the current damage variable value, T is the current temperature, and f is a preset nonlinear function. A correlation function is established between the damage variable D and the cumulative acoustic emission energy E. The correlation function is in the form of D=g(E) or dD=h(dE), where g and h are monotonically increasing functions obtained by fitting material test data. The nonlinear evolution equation and the correlation function are coupled to form the control equation set of the damage evolution constitutive model.

[0103] In one specific embodiment, tunnel lining materials (mainly concrete or reinforced concrete) experience microcrack initiation, propagation, and convergence under load, ultimately leading to macroscopic failure. This process is accompanied by the deterioration of material stiffness and irreversible energy dissipation. Damage mechanics, by introducing damage variables to describe the degree of material degradation, maps the evolution of microscopic defects to macroscopic constitutive relations. The core idea of ​​this step is to establish a nonlinear constitutive model with damage variables as internal variables. This model can predict the rate of damage evolution based on the current stress state and the cumulative damage level, and establish a correlation between external monitoring data and internal damage state through acoustic emission accumulated energy, providing a theoretical basis for subsequent inversion calculations.

[0104] Specifically, damage variables This is an internal state variable describing the degree of material degradation. There are various ways to define it. This embodiment uses a scalar damage variable based on stiffness degradation (applicable to isotropic damage), defined as follows:

[0105] ,

[0106] In the formula, This represents the initial elastic modulus of the material (in its undamaged state). The equivalent elastic modulus under the current damage state. The range of values ​​is , This indicates that the materials are intact and undamaged. This indicates that the material has completely lost its load-bearing capacity.

[0107] For anisotropic damage, a second-order damage tensor can be used, but to simplify the problem, this embodiment will first use scalar damage as an example.

[0108] Damage driving force is a thermodynamic generalized force that drives damage evolution, typically taken as the strain energy release rate conjugate with the damage variable. According to continuum damage mechanics, the damage driving force... Defined as:

[0109] ,

[0110] In the formula, Let be the Helmholtz free energy of the material. For a linear elastic material, the elastic strain energy considering damage is:

[0111] ,

[0112] In the formula, For stress tensor, For strain tensor, Let be the elastic stiffness tensor after damage. For scalar damage, we have: ,in, Let be the initial elastic stiffness tensor. Substituting this, we get:

[0113] ,

[0114] In the formula, This is the equivalent stress (which can be selected according to different strength criteria, such as Von Mises equivalent stress). Therefore, the damage driving force is related to the current stress level and the degree of damage.

[0115] Damage evolution equations describe the increments of damage variables With damage driving force The relationship between them. This embodiment uses a nonlinear evolution equation based on the energy release rate, the general form of which is:

[0116] ,

[0117] In the formula, The damage threshold is related to the current degree of damage and temperature. This is a nonlinear function. The specific function form needs to be determined based on the material properties. This embodiment preferably uses a power-law evolution equation because it is simple in form and can better describe the damage evolution of concrete-like materials.

[0118] ,

[0119] in, These are material parameters and need to be calibrated through indoor testing. The initial damage threshold (corresponding to the damage initiation point) can be regarded as a material constant or temperature-related. For temperature, if the effect of temperature is considered, we can let , wait.

[0120] Acoustic emission signals originate from the elastic wave energy released during microcrack propagation. Theoretically, the cumulative acoustic emission energy per unit time is proportional to the increment of the damage variable. Therefore, the following correlation function can be established:

[0121] ,

[0122] in, Accumulated energy of acoustic emission (unit: joules or relative energy). This is the proportionality coefficient. The initial damage is assumed to be 0. This correlation function indicates that the damage increment is proportional to the acoustic emission energy rate, with a proportionality coefficient... It reflects the damage increment corresponding to a unit of acoustic emission energy.

[0123] To more accurately describe nonlinear relationships, a more general functional form can be used:

[0124] or ,

[0125] in, and This is a monotonically increasing function obtained by fitting material test data. For example, a power function can be used. .

[0126] The above material parameters ( (etc.) need to be calibrated through indoor testing. The specific steps are as follows:

[0127] Specimen preparation: Concrete specimens (standard cylinders or prisms) with the same mix proportions as the on-site lining were prepared, and acoustic emission sensors were pre-embedded.

[0128] Loading test: Uniaxial compression or splitting test is performed on a universal testing machine, and stress-strain data and acoustic emission signals are collected simultaneously.

[0129] Data post-processing:

[0130] Based on the stress-strain curve, calculate the damage variable at each time step. (Through the unloading stiffness method or the strain equivalence assumption).

[0131] Synchronous calculation of cumulative acoustic emission energy .

[0132] Establish and Relationship, fitting correlation function or proportionality coefficient .

[0133] Based on the relationship between damage evolution rate and current stress and damage degree, a fitting was performed. .

[0134] For parameters at different temperatures, experiments under different temperature conditions are required to establish the temperature effect function.

[0135] By coupling the damage evolution equation with the acoustic emission correlation function, a complete set of governing equations for the damage evolution constitutive model is formed:

[0136] ,

[0137] This set of equations establishes the intrinsic relationship between stress, strain, damage, and acoustic emission energy, providing a physical basis for subsequent inversion calculations. In practical applications, incremental numerical solutions are typically used.

[0138] In summary, the method implemented in this paper links energy release at the microscale with damage variables at the macroscale through acoustic emission correlation functions, enabling previously isolated acoustic emission monitoring data to be integrated into structural analysis and fully leveraging the advantage of acoustic emission's sensitivity to microcracks.

[0139] Step S105: The net displacement data sequence is used as the displacement boundary condition, and the net acoustic emission data sequence is used as the energy dissipation constraint. These are input into the damage evolution constitutive model for finite element inversion calculation to obtain the damage variable values ​​of each unit inside the lining at the current moment, thus forming the damage field distribution.

[0140] In this step, a finite element discrete model of the tunnel lining structure is established, dividing the lining into multiple finite elements and assigning initial damage variable values ​​to each element. The displacement values ​​at each time step in the net displacement data sequence are used as displacement boundary conditions and applied to the corresponding nodes of the finite element model. The cumulative acoustic emission energy values ​​at each time step in the net acoustic emission data sequence are used as energy dissipation constraints to establish an objective function, which is the sum of squared residuals between the theoretical acoustic emission energy calculated by the finite element model and the actual net acoustic emission energy. An optimization algorithm is used to iteratively adjust the damage variable values ​​of each element to minimize the value of the objective function. When the objective function value converges to a preset accuracy range, the iteration stops, and the damage variable values ​​of each element at the current time are output to form the damage field distribution.

[0141] In one specific embodiment, a three-dimensional geometric model is established based on the design drawings and actual dimensions of the tunnel lining. For long tunnels, a representative cross-section (e.g., one cross-section every 20-30 meters) can be selected to establish a two-dimensional plane strain model, or a three-dimensional segmental model including the longitudinal length can be established. The geometric model should accurately reflect the thickness, outline, and possible structural joints, deformation joints, and other details of the lining.

[0142] The geometric model is meshed using finite element preprocessing software (such as ANSYS, ABAQUS, COMSOL, or the open-source tool Gmsh). The meshing principles are as follows:

[0143] Element type: For 2D models, use quadrilateral or triangular elements (such as CPE4, CPE3); for 3D models, use hexahedral or tetrahedral elements (such as C3D8, C3D4). Higher-order elements are recommended to improve computational accuracy.

[0144] Mesh density: The mesh should be denser in stress concentration areas and sensor deployment areas; it can be appropriately sparser in areas far from monitoring points. Generally, the element feature size should not exceed 1 / 5 of the lining thickness to ensure calculation accuracy.

[0145] Number of elements: Determined based on model size and computing resources. A typical cross-section two-dimensional model has approximately 1,000-5,000 elements, while a three-dimensional model can have tens of thousands to hundreds of thousands of elements.

[0146] Assign initial material properties to each element:

[0147] initial elastic modulus The value is determined based on the concrete strength grade, such as 30 GPa for C30 concrete.

[0148] Poisson's ratio : Concrete is generally taken as 0.2.

[0149] density The concrete concentration is 2400-2500 kg / m³.

[0150] Initial damage variables Normally set to 0 (intact state). If there is historical test data, a non-zero initial value can be set according to the test results.

[0151] Based on the actual stress conditions of the tunnel, the model boundary conditions are set as follows:

[0152] Displacement boundary: Normal constraints or spring boundaries can be set at the contact surface between the lining and the surrounding rock to simulate the support of the surrounding rock; fixed constraints or elastic supports can be set at the bottom of the tunnel.

[0153] Load boundary: External loads such as surrounding rock pressure and water pressure can be applied. If there is no clear load data, the displacement monitoring value can be used as the driving boundary, and no additional load is applied during the inversion process.

[0154] Net displacement data sequence It contains displacement values ​​at multiple times. Inversion calculations are usually targeted at a specific time (such as the current time), but time-series inversions can also be performed.

[0155] Match the displacement measurement points with the nodes of the finite element model:

[0156] If the measuring point happens to be located on a node, then the displacement value of that node is directly set as a constraint.

[0157] If the measuring point is located inside the element, the relationship between the displacement of the measuring point and the nodal displacement can be established through shape function interpolation: ,in For the shape function matrix, The element node displacement vector. This is the displacement vector of the unit measurement point.

[0158] For nodes with displacement monitoring or nodes associated with monitoring points, apply displacement constraints:

[0159] ,

[0160] in, Number the measurement points. For the target time.

[0161] For nodes without displacement monitoring, set them as free boundaries (i.e., unconstrained).

[0162] The net acoustic emission data sequence provides the cumulative acoustic emission energy at each time step. In the inversion process, it is necessary to establish an error function between the theoretical acoustic emission energy calculated by the finite element model and the actual monitored value.

[0163] (1) Calculation of theoretical acoustic emission energy

[0164] Based on the damage evolution constitutive model established in step S104, the element damage increment This is related to the release of acoustic emission energy. In the finite element method, for a given damage variable D, the damage driving force of each element can be calculated. The damage increment is then obtained through the damage evolution equation. According to the acoustic emission correlation function The theoretical acoustic emission energy of the unit is:

[0165] ,

[0166] in, The theoretical acoustic emission energy of the unit. The unit volume (in a two-dimensional model, it is the area multiplied by the unit thickness). The scaling factor is the one calibrated in step S104.

[0167] The theoretical cumulative acoustic emission energy of the entire model is the sum of the contributions from all elements:

[0168] ,

[0169] (2) Definition of objective function

[0170] Define the objective function The sum of squared weighted residuals between the theoretical acoustic emission energy and the actual net acoustic emission energy:

[0171] ,

[0172] in, The number of times to be used in the inversion (you can take only the current time or take multiple times for time series inversion). The weighting coefficients for each time point can be set according to the data quality (time points with higher signal-to-noise ratios have higher weights).

[0173] This is a regularization term used to ensure the well-posedness of the inversion problem. Commonly used regularization terms include: total variation regularization: Encourages segmented smooth damage fields (suitable for identifying damage region boundaries);

[0174] objective function It is a nonlinear function of the damage variable D, and requires an iterative optimization algorithm to solve.

[0175] Set the initial value of each unit's damage variable to... (In good condition), or set according to historical test results.

[0176] To solve efficiently, it is necessary to calculate the gradient of the objective function with respect to the damage variable. Two methods can be used:

[0177] The adjoint method solves the adjoint equation, efficiently calculates the gradient, and is suitable for large-scale problems.

[0178] Numerical differentiation: Apply a small perturbation to each damage variable and calculate the gradient using finite differences. It is suitable for small-scale problems.

[0179] This embodiment preferably uses the accompanying method, and its calculation process is as follows:

[0180] Solving the finite element equilibrium equations yields the displacement field. ;

[0181] Damage driving force of each element is calculated based on displacement field. ;

[0182] Solve the adjoint equation to obtain the adjoint variables. ;

[0183] Calculate the gradient: , Let the global stiffness matrix K be the damage variable of the e-th element. The partial derivatives, due to the influence of damage variables on material stiffness, and the element stiffness matrix Related to injury: ,in For lossless element stiffness; For the acoustic emission energy constraint part of the objective function For damage variables The direct partial derivatives, , Through unit damage increment calculate; For regularization terms For damage variables The partial derivatives;

[0184] Depending on the problem size and degree of nonlinearity, a quasi-Newton method (such as L-BFGS) is selected for iterative solution. The iterative process is as follows:

[0185] Let the number of iterations Set the initial damage field ;

[0186] Solving the finite element equilibrium equations yields the displacement field. ;

[0187] Calculate the damage driving force of each element Theoretical acoustic emission energy ;

[0188] Calculate the objective function value ;

[0189] like If the damage field is less than the preset threshold or the number of iterations exceeds the maximum number of iterations, output the current damage field as the inversion result; otherwise, calculate the gradient, update the damage field using L-BFGS, and continue to solve the finite element equilibrium equations.

[0190] Output the damage variable values ​​of each element and organize them according to the element coordinates to form a damage field distribution.

[0191] Step S106: Based on the damage field distribution, identify continuous spatial regions where the damage variable value exceeds a preset damage threshold as active damage regions, and extract the geometric center coordinates and area of ​​the active damage regions at each time point.

[0192] In one specific embodiment, in step S105, the damage variable values ​​of each element inside the lining are obtained through finite element inversion, forming a damage field distribution. This damage field reflects the spatial distribution of the material degradation degree. However, directly observing the damage values ​​of all elements makes it difficult to focus on key problem areas. Therefore, this step aims to automatically identify areas with high damage degree and spatial continuity from the damage field—i.e., "damage active areas"—and extract their geometric features (center coordinates, area) to provide a basis for subsequent spatiotemporal trajectory analysis.

[0193] The identification of active damage areas is based on two core criteria:

[0194] Damage severity criterion: If the value of the unit damage variable exceeds the preset threshold, it indicates that the region has undergone significant degradation.

[0195] Spatial continuity criterion: Units that meet the damage degree criterion are spatially interconnected, forming a continuous region, representing an independent damaged body.

[0196] For each remaining active damage area Calculate the following geometric features:

[0197] Geometric center coordinates: weighted by the element damage variable value, highlighting the impact of severely damaged areas.

[0198] ,

[0199] In the formula, Active damage area The geometric center, Let be the coordinates of the geometric center of element e;

[0200] For a two-dimensional model, the area of ​​the active damage region is the sum of the areas of all elements (if the element shapes are regular, the sum can be calculated directly):

[0201] ,

[0202] in, This is the area of ​​the unit (for quadrilateral units, the actual area needs to be calculated based on the node coordinates).

[0203] For a 3D model, the region volume is the sum of the volumes of all individual elements:

[0204] ,

[0205] in, For the unit volume, if the unit shape is irregular (such as triangle or tetrahedron), the accurate area / volume needs to be calculated based on the coordinates of the unit nodes.

[0206] In this embodiment, the geometric center coordinates and area of ​​the active damage zone are core parameters describing its spatial location and scale. By extracting these features at multiple time points, the migration trajectory and expansion curve of the active damage zone can be constructed, laying the data foundation for the spatiotemporal evolution analysis in step S107. The weighted center calculation method better reflects the location of the most severely damaged area, making the trajectory more accurately reflect the movement of the damage core.

[0207] Step S107: Fit the geometric center coordinates and area of ​​the active damage region along the time axis to obtain the spatial migration trajectory of the center of the active damage region and the expansion curve of the area over time.

[0208] In this step, the geometric center coordinates of the active damage area at multiple consecutive time points are obtained to form a center point coordinate sequence. The geometric center coordinates are the weighted average of the coordinates of all units in the active damage area, and the weight is the damage variable value of each unit.

[0209] By performing polynomial fitting or spline interpolation on the center point coordinate sequence, the migration trajectory of the center of the active damage area in the spatial coordinate system that changes continuously with time can be obtained.

[0210] The area of ​​the active damage region at multiple consecutive time points is obtained to form an area sequence, wherein the area of ​​the active damage region is the sum of the areas of all units within the active damage region;

[0211] An exponential function is fitted to the area sequence to obtain an expansion curve of the area of ​​the active damage zone that changes continuously over time.

[0212] In one specific embodiment, step S106 extracts the geometric features (geometric center coordinates and area of ​​the active damage region) of the damage field at each time step. These discrete time-time feature values ​​constitute the original observation sequence of the active damage region's evolution. However, discrete observations cannot directly reveal the continuous law of damage evolution and are difficult to use for prediction of future time steps. Therefore, this step aims to transform these discrete feature points into continuous functions through mathematical fitting methods, thereby obtaining the migration trajectory and expansion curve, specifically:

[0213] Multiple time points are obtained from step S106. The feature data of the active damage region. For each active damage region It is necessary to ensure that the same physical region is identified at different times (the corresponding relationship has been established through the matching algorithm in step S106). If the region is not identified at a certain time (e.g., due to damage threshold adjustment or temporary disappearance of the region), interpolation or removal of that time can be performed.

[0214] Outlier detection is performed on both the center coordinate sequence and the area sequence. The following methods can be used:

[0215] 3σ principle: Calculate the mean and standard deviation of the series, and remove points that deviate from the mean by more than 3 times the standard deviation;

[0216] Local Outlier Factor (LOF): Identifies outliers based on local density;

[0217] Manual review: For suspicious points, the decision on whether to retain them can be made by combining the original monitoring data or engineering experience.

[0218] If outliers are found, the cause needs to be analyzed (such as sensor failure, misidentification, etc.). If necessary, they should be removed or corrected. If the data noise is high, smoothing can be performed first, such as moving average, Savitzky-Golay filtering, etc., before fitting.

[0219] The coordinates of the center point are multi-dimensional vectors (two-dimensional) or three-dimensional (This can be used to fit each coordinate component separately and then combine them into a vector function).

[0220] Furthermore, the area of ​​the active damage zone typically grows non-linearly over time. Common patterns include exponential growth (such as accelerated damage expansion) or S-shaped growth (such as finite growth constrained by boundaries). Therefore, the natural logarithm of the area data is taken and transformed into a linear regression problem. The parameters are then estimated using the least squares method. If the offset is considered, non-linear least squares (such as the Levenberg-Marquardt algorithm) should be used to directly fit the original data.

[0221] Step S108: Based on the migration trajectory and the expansion curve, predict the health status assessment results of the active damage area within a future preset time window.

[0222] In this step, the migration trajectory is extrapolated towards future time to predict the spatial location that the center of the active damage area will reach at various times within a preset future time window.

[0223] Extrapolate the expansion curve towards future time to predict the area that the damage active region will reach at each time point within a preset future time window;

[0224] Based on the rate of change of the slope of the expansion curve, the stage of damage evolution is determined: when the rate of change of the slope is less than the first threshold, it is determined to be in the stable expansion stage; when the rate of change of the slope is greater than or equal to the first threshold and less than the second threshold, it is determined to be in the accelerated expansion stage; when the rate of change of the slope is greater than or equal to the second threshold, it is determined to be in the instability and damage precursor stage.

[0225] Based on the predicted future spatial location, regional area, and current stage of damage evolution, a health status assessment result is generated, including the coordinates of key areas of concern, the expected impact range, and the warning level.

[0226] In one specific embodiment, step S107 has obtained a functional representation of the migration trajectory of the active damage region. Functional representation of the spread curve These functions describe the historical patterns of damage evolution. The core of this step lies in using the mathematical form of these functions to extrapolate into the future, predicting the evolution trend of the active damage area within a preset time window, and combining this with the dynamic characteristics of the expansion curve to determine the current damage stage, ultimately generating a health status assessment result with engineering guidance significance.

[0227] Extrapolation prediction is based on a fundamental assumption: damage evolution follows its historical pattern in the short term, without any abrupt changes. By analyzing the rate of change of the slope of the expansion curve (i.e., the second derivative or curvature), the acceleration of damage expansion can be quantified, thereby dividing the evolutionary stages and providing a scientific basis for early warning. Specifically:

[0228] Prediction Time Window The choice of a prediction timeframe needs to balance predictive reliability with engineering requirements. Too short a timeframe limits predictive significance, while too long a timeframe introduces excessive uncertainty. Determination methods include:

[0229] Fixed window method: set according to the inspection cycle or early warning needs, such as the next 7 days or 30 days.

[0230] Dynamic window method: Based on the growth rate of the expansion curve, the window is dynamically adjusted. For example, when the expansion rate is high, the window is shortened to avoid over-extrapolation; when the expansion rate is low, the window can be appropriately extended.

[0231] Uncertainty threshold method: The prediction window is set to the time when the prediction confidence interval width of the migration trajectory or expansion curve reaches a certain limit.

[0232] This embodiment recommends a combination of fixed window and dynamic adjustment: First, a base window (e.g., 30 days) is set. If the confidence interval of the predicted value exceeds the allowable range within this window (e.g., the coefficient of variation of the predicted area > 0.3), the window is automatically shortened until the confidence interval meets the requirements.

[0233] Extrapolation prediction of migration trajectory:

[0234] Migration trajectory Usually time Polynomial or spline functions. Extrapolation is the calculation of... The function value at that time.

[0235] If spline fitting is used, the extension method at the endpoints needs to be specified. Common methods include:

[0236] Natural splines: Assuming the second derivative is zero at the endpoints, the extrapolated segments are linear.

[0237] Linear extrapolation based on the derivative of the last segment: keeping the slope of the last segment unchanged.

[0238] For the future preset time window At that moment Calculate the corresponding prediction center coordinates:

[0239] ,

[0240] In the formula, For the first The active damage area at future time The predicted center coordinate vector, For the first The active damage area at future time The X-axis coordinate vector of the prediction center. For the first The active damage area at future time The predicted center Y-axis coordinate vector, For the first The active damage area at future time The Z-axis coordinate vector of the prediction center;

[0241] These coordinates constitute the possible future movement path of the damaged core.

[0242] If the fitting method can provide the parameter covariance (such as the parameter covariance matrix of polynomial fitting), the confidence interval of the predicted value can be calculated through error propagation. Otherwise, an empirical method can be used: the standard deviation of the historical fitting residuals can be used as an estimate of the prediction error to construct the prediction band.

[0243] Extrapolation prediction of extended curves:

[0244] Obtain model parameters: Read the exponential model parameters of the current active damage area from the results of step S107. , and its covariance matrix (used for uncertainty estimation).

[0245] Set the forecast time window: Set the forecast duration ΔT (e.g., the next 30 days) according to engineering requirements, and discretize the window into a series of forecast times. Where k = 1, 2, ..., m, The time step (e.g., 1 day). This refers to the current moment.

[0246] Calculate the predicted area: For each prediction time, substitute the time into the exponential model to obtain the estimated area point value.

[0247] ,

[0248] In the formula, The estimated scale parameters of the fitted exponential model represent the amount of damage area related at the initial time (or baseline time). The estimated growth rate parameter of the fitted exponential model reflects how quickly the damaged area grows over time. For the k-th prediction time;

[0249] To quantify the uncertainty of the forecast, the Delta method is used to calculate the standard error of the forecast value;

[0250] Applying physical constraints: If the predicted area exceeds the maximum allowable area of ​​the structure. (If the predicted value is 1.5 times the total surface area of ​​the lining or the historical maximum damage area), then the predicted value will be truncated to... And indicate this in the results.

[0251] Generate prediction sequence: Compile the point prediction values ​​and confidence intervals at each time point into a table, and output it as the extrapolation result of the expansion curve for subsequent health status assessment (such as calculating the expansion rate and determining the evolution stage).

[0252] It should be noted that in step S107, a continuous function of the area of ​​the active damage zone changing with time has been obtained by fitting historical data. The slope change rate is defined as the rate of change of the area expansion rate with respect to time, i.e., acceleration.

[0253] In summary, the method of this application employs temperature deformation correction technology, calculates the thermal expansion component based on real-time temperature data and subtracts it from the displacement monitoring value, eliminating pseudo-deformation caused by diurnal temperature differences and seasonal changes due to thermal expansion and contraction. Simultaneously, it identifies periods of thermal noise interference by using a temperature change rate threshold, eliminating pseudo-acoustic emission signals generated by the sensor during periods of drastic temperature fluctuations. This fundamentally solves the problem of "mixed true and false" monitoring data in tunnel environments, providing a clean sequence of net displacement and net acoustic emission data for subsequent analysis. Furthermore, it constructs a nonlinear damage evolution constitutive model with damage variables as internal variables, establishes a power-law evolution equation between damage driving force and damage increment, and a correlation function between damage variables and cumulative acoustic emission energy. Net displacement data is used as displacement boundary conditions, and net acoustic emission data as energy dissipation constraints, input into the finite element model for inversion calculation, and solved using the adjoint method. Gradient and iterative optimization are used to output the damage variable values ​​of each unit inside the lining, forming a three-dimensional damage field distribution. This achieves a quantitative mapping from the external macroscopic response to the internal microscopic damage state, enabling spatial visualization of the invisible internal deterioration process. Based on the damage field, continuous regions where damage variables exceed a threshold are identified as active damage areas, and their geometric center coordinates and area are calculated using damage values ​​as weights. Polynomial or spline fitting is performed on the geometric center coordinates at multiple time points to obtain the continuous migration trajectory of the damage core in space. An exponential function fitting is performed on the area sequence to obtain the expansion curve of the damage range. Discrete monitoring time data are transformed into continuous evolution functions, revealing the movement path and expansion law of the damage core. This elevates the analysis of tunnel lining health status from qualitative description relying on experience judgment to quantitative assessment and intelligent early warning based on deep integration of physical models and measured data.

[0254] Please see Figure 2 The diagram shows the structural block diagram of the tunnel lining health status analysis system based on field monitoring data of this application.

[0255] like Figure 2 As shown, the tunnel lining health status analysis system 200 includes an acquisition module 210, a correction module 220, a rejection module 230, a construction module 240, an output module 250, an extraction module 260, a fitting module 270, and a prediction module 280.

[0256] The acquisition module 210 is configured to acquire a multi-point displacement gauge monitoring data sequence, a multi-channel acoustic emission monitoring data sequence, and a temperature data sequence of internal temperature measuring points of the tunnel lining within a preset time period. The correction module 220 is configured to calculate the temperature change at each moment based on the temperature data sequence, and perform temperature deformation correction on the multi-point displacement gauge monitoring data sequence based on a preset thermal expansion coefficient of the lining material to obtain a net displacement data sequence after removing the temperature influence. The removal module 230 is configured to identify time periods where the temperature change rate exceeds a preset threshold based on the temperature data sequence, and remove acoustic emission signals interfered with by thermal noise in the corresponding time periods to obtain a net acoustic emission data sequence. The construction module 240 is configured to construct a lining material damage evolution constitutive model with damage variables as internal variables, wherein the damage evolution constitutive model includes a nonlinear relationship between the damage driving force and the increment of the damage variable. The output module... Block 250 is configured to use the net displacement data sequence as displacement boundary conditions and the net acoustic emission data sequence as energy dissipation constraints, inputting them into the damage evolution constitutive model for finite element inversion calculation to obtain the damage variable values ​​of each element inside the lining at the current moment, forming a damage field distribution; Extraction module 260 is configured to identify continuous spatial regions where the damage variable values ​​exceed a preset damage threshold based on the damage field distribution, as damage active regions, and extract the geometric center coordinates and area of ​​the damage active regions at each moment; Fitting module 270 is configured to fit the geometric center coordinates and area of ​​the damage active regions along the time axis to obtain the spatial migration trajectory of the center of the damage active regions and the expansion curve of the area over time; Prediction module 280 is configured to predict the health status assessment results of the damage active regions within a preset time window based on the migration trajectory and the expansion curve.

[0257] It should be understood that Figure 2 The modules and references described in the document Figure 1 The steps described in the text correspond to those in the method described above. Therefore, the operations, features, and corresponding technical effects described above also apply to the method described in the text. Figure 2 The various modules in the document will not be described in detail here.

[0258] In other embodiments, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the tunnel lining health status analysis method based on field monitoring data in any of the above method embodiments.

[0259] In one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions, which are configured as follows:

[0260] Acquire multi-point displacement gauge monitoring data sequence, multi-channel acoustic emission monitoring data sequence, and temperature data sequence of internal temperature measuring points of the tunnel lining within a preset time period;

[0261] The temperature change at each moment is calculated based on the temperature data sequence. Based on the preset thermal expansion coefficient of the lining material, the temperature deformation correction is applied to the multi-point displacement gauge monitoring data sequence to obtain the net displacement data sequence after removing the temperature influence.

[0262] Based on the temperature data sequence, identify the time periods where the temperature change rate exceeds a preset threshold, and remove the acoustic emission signals that are interfered with by thermal noise in the corresponding time periods to obtain the net acoustic emission data sequence.

[0263] A constitutive model of damage evolution of lining material is constructed with damage variables as internal variables. The constitutive model of damage evolution includes the nonlinear relationship between damage driving force and damage variable increment.

[0264] The net displacement data sequence is used as the displacement boundary condition, and the net acoustic emission data sequence is used as the energy dissipation constraint. These are input into the damage evolution constitutive model for finite element inversion calculation to obtain the damage variable values ​​of each element inside the lining at the current moment, thus forming the damage field distribution.

[0265] Based on the damage field distribution, continuous spatial regions where the damage variable values ​​exceed a preset damage threshold are identified as active damage regions, and the geometric center coordinates and area of ​​the active damage regions at each time point are extracted.

[0266] By fitting the geometric center coordinates and area of ​​the active damage region along the time axis, the spatial migration trajectory of the center of the active damage region and the expansion curve of the area over time are obtained.

[0267] Based on the migration trajectory and the expansion curve, the health status assessment results of the active damage area within a future preset time window are predicted.

[0268] Computer-readable storage media may include a stored program area and a stored data area, wherein the stored program area may store an operating system and an application program required for at least one function; the stored data area may store data created based on the use of the tunnel lining health status analysis system based on field monitoring data, etc. Furthermore, the computer-readable storage medium may include high-speed random access memory, and may also include memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the computer-readable storage medium may optionally include memory remotely configured relative to a processor, which can be connected via a network to the tunnel lining health status analysis system based on field monitoring data. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0269] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present invention, such as... Figure 3 As shown, the device includes a processor 310 and a memory 320. The electronic device may also include an input device 330 and an output device 340. The processor 310, memory 320, input device 330, and output device 340 can be connected via a bus or other means. Figure 3 Taking a bus connection as an example, the memory 320 is the computer-readable storage medium described above. The processor 310 executes various server functions and data processing by running non-volatile software programs, instructions, and modules stored in the memory 320, thereby implementing the tunnel lining health status analysis method based on field monitoring data described in the above embodiment. The input device 330 can receive input digital or character information and generate key signal inputs related to user settings and function control of the tunnel lining health status analysis system based on field monitoring data. The output device 340 may include a display screen or other display device.

[0270] The aforementioned electronic device can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in the embodiments of the present invention.

[0271] In one implementation, the above-described electronic device is applied to a tunnel lining health status analysis system based on field monitoring data, and is used as a client. It includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to:

[0272] Acquire multi-point displacement gauge monitoring data sequence, multi-channel acoustic emission monitoring data sequence, and temperature data sequence of internal temperature measuring points of the tunnel lining within a preset time period;

[0273] The temperature change at each moment is calculated based on the temperature data sequence. Based on the preset thermal expansion coefficient of the lining material, the temperature deformation correction is applied to the multi-point displacement gauge monitoring data sequence to obtain the net displacement data sequence after removing the temperature influence.

[0274] Based on the temperature data sequence, identify the time periods where the temperature change rate exceeds a preset threshold, and remove the acoustic emission signals that are interfered with by thermal noise in the corresponding time periods to obtain the net acoustic emission data sequence.

[0275] A constitutive model of damage evolution of lining material is constructed with damage variables as internal variables. The constitutive model of damage evolution includes the nonlinear relationship between damage driving force and damage variable increment.

[0276] The net displacement data sequence is used as the displacement boundary condition, and the net acoustic emission data sequence is used as the energy dissipation constraint. These are input into the damage evolution constitutive model for finite element inversion calculation to obtain the damage variable values ​​of each element inside the lining at the current moment, thus forming the damage field distribution.

[0277] Based on the damage field distribution, continuous spatial regions where the damage variable value exceeds a preset damage threshold are identified as active damage regions, and the geometric center coordinates and area of ​​the active damage regions at each time point are extracted.

[0278] By fitting the geometric center coordinates and area of ​​the active damage region along the time axis, the spatial migration trajectory of the center of the active damage region and the expansion curve of the area over time are obtained.

[0279] Based on the migration trajectory and the expansion curve, the health status assessment results of the active damage area within a future preset time window are predicted.

[0280] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0281] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for analyzing the health status of tunnel lining based on field monitoring data, characterized in that, include: Acquire multi-point displacement gauge monitoring data sequence, multi-channel acoustic emission monitoring data sequence, and temperature data sequence of internal temperature measuring points of the tunnel lining within a preset time period; The temperature change at each moment is calculated based on the temperature data sequence. Based on the preset thermal expansion coefficient of the lining material, the temperature deformation correction is applied to the multi-point displacement gauge monitoring data sequence to obtain the net displacement data sequence after removing the temperature influence. Based on the temperature data sequence, identify the time periods where the temperature change rate exceeds a preset threshold, and remove the acoustic emission signals that are interfered with by thermal noise in the corresponding time periods to obtain the net acoustic emission data sequence. A constitutive model for the damage evolution of lining materials is constructed, with damage variables as internal variables. This constitutive model includes the nonlinear relationship between the damage driving force and the increment of the damage variable. The construction of this constitutive model for the damage evolution of lining materials with damage variables as internal variables includes: The damage variable D is defined to describe the degree of deterioration of the lining material under load. The value of the damage variable D ranges from 0 to 1, where D=0 indicates that the material is intact and D=1 indicates that the material has completely failed. A nonlinear evolution equation is established between the damage driving force and the damage variable increment. The damage driving force is calculated based on the current stress state and the cumulative damage degree. The nonlinear evolution equation is in the form of: dD / dt=f(σ,D,T), where σ is the current stress, D is the current damage variable value, T is the current temperature, and f is a preset nonlinear function. Establish a correlation function between the damage variable D and the cumulative acoustic emission energy E, wherein the correlation function is in the form of: or ,in, and This is a monotonically increasing function obtained by fitting material test data; The nonlinear evolution equation and the correlation function are coupled to form the control equation set of the damage evolution constitutive model; The net displacement data sequence is used as the displacement boundary condition, and the net acoustic emission data sequence is used as the energy dissipation constraint. These are input into the damage evolution constitutive model for finite element inversion calculation to obtain the damage variable values ​​of each element inside the lining at the current moment, thus forming the damage field distribution. Based on the damage field distribution, continuous spatial regions where the damage variable values ​​exceed a preset damage threshold are identified as active damage regions, and the geometric center coordinates and area of ​​the active damage regions at each time point are extracted. By fitting the geometric center coordinates and area of ​​the active damage region along the time axis, the spatial migration trajectory of the center of the active damage region and the expansion curve of the area over time are obtained. Based on the migration trajectory and the expansion curve, the health status assessment results of the active damage area within a future preset time window are predicted.

2. The method for analyzing the health status of tunnel lining based on field monitoring data according to claim 1, characterized in that, The step of calculating the temperature change at each moment based on the temperature data sequence, and correcting the temperature deformation of the multi-point displacement gauge monitoring data sequence based on the preset thermal expansion coefficient of the lining material to obtain the net displacement data sequence after removing the temperature influence includes: The measured displacement values ​​at each moment in the multi-point displacement gauge monitoring data sequence, and the measured temperature values ​​at the corresponding moment of the lining internal temperature measuring point; Based on the difference between the measured temperature value and the preset reference temperature value, the temperature change at each moment is calculated. The thermal deformation component caused by temperature change at each moment is calculated based on the product of the preset thermal expansion coefficient of the lining material and the temperature change. Subtract the thermal deformation component at the corresponding time from the measured displacement value at each time to obtain the net displacement value at each time. The net displacement values ​​at each time point are sorted in chronological order to obtain the net displacement data sequence.

3. The method for analyzing the health status of tunnel lining based on field monitoring data according to claim 1, characterized in that, The step of identifying time periods where the temperature change rate exceeds a preset threshold based on the temperature data sequence, and removing acoustic emission signals interfered with by thermal noise within the corresponding time periods to obtain a net acoustic emission data sequence includes: Obtain the measured temperature values ​​of adjacent moments in the temperature data sequence and calculate the rate of temperature change at each moment; The temperature change rate at each moment is compared with a preset temperature change rate threshold, and continuous time periods in which the temperature change rate exceeds the preset threshold are identified as thermal noise interference periods. In the multi-channel acoustic emission monitoring data sequence, find the acoustic emission signal segment corresponding to the thermal noise interference period; The found acoustic emission signal segments are removed from the multi-channel acoustic emission monitoring data sequence, and the acoustic emission signals of the remaining time periods are retained; The retained acoustic emission signals are time-series spliced ​​and resampled to obtain the net acoustic emission data sequence.

4. The method for analyzing the health status of tunnel lining based on field monitoring data according to claim 1, characterized in that, The process involves using the net displacement data sequence as displacement boundary conditions and the net acoustic emission data sequence as energy dissipation constraints, inputting them into the damage evolution constitutive model for finite element inversion calculation, and solving for the damage variable values ​​of each element inside the lining at the current moment, including: A finite element discrete model of the tunnel lining structure is established, the lining is divided into multiple finite elements, and initial damage variable values ​​are assigned to each element; The displacement values ​​at each time point in the net displacement data sequence are used as displacement boundary conditions and applied to the corresponding nodes of the finite element model. The cumulative acoustic emission energy value at each moment in the net acoustic emission data sequence is used as an energy dissipation constraint to establish an objective function, which is the sum of squared residuals between the theoretical acoustic emission energy calculated by the finite element model and the actual net acoustic emission energy. An optimization algorithm is used to iteratively adjust the damage variable values ​​of each unit to minimize the value of the objective function; When the objective function value converges to the preset accuracy range, the iteration stops, and the damage variable values ​​of each unit at the current time are output to form the damage field distribution.

5. The method for analyzing the health status of tunnel lining based on field monitoring data according to claim 1, characterized in that, The process of fitting the geometric center coordinates and area of ​​the active damage region along the time axis to obtain the spatial migration trajectory of the center of the active damage region and the expansion curve of its area over time includes: Obtain the geometric center coordinates of the active damage zone at multiple consecutive time points to form a center point coordinate sequence. The geometric center coordinates are the weighted average of the coordinates of all units within the active damage zone, with the weight being the damage variable value of each unit. By performing polynomial fitting or spline interpolation on the center point coordinate sequence, the migration trajectory of the center of the active damage area in the spatial coordinate system that changes continuously with time is obtained. The area of ​​the active damage region at multiple consecutive time points is obtained to form an area sequence, wherein the area of ​​the active damage region is the sum of the areas of all units within the active damage region; An exponential function is fitted to the area sequence to obtain an expansion curve of the area of ​​the active damage zone that changes continuously over time.

6. The method for analyzing the health status of tunnel lining based on field monitoring data according to claim 1, characterized in that, The assessment result of predicting the health status of the active damage area within a future preset time window based on the migration trajectory and the expansion curve includes: Extrapolate the migration trajectory towards future time to predict the spatial location that the center of the active damage area will reach at various times within a preset future time window. Extrapolate the expansion curve towards future time to predict the area that the damage active region will reach at each time point within a preset future time window; Based on the rate of change of the slope of the expansion curve, the stage of damage evolution is determined: when the rate of change of the slope is less than the first threshold, it is determined to be in the stable expansion stage; when the rate of change of the slope is greater than or equal to the first threshold and less than the second threshold, it is determined to be in the accelerated expansion stage; when the rate of change of the slope is greater than or equal to the second threshold, it is determined to be in the instability and damage precursor stage. Based on the predicted future spatial location, regional area, and current stage of damage evolution, a health status assessment result is generated, including the coordinates of key areas of concern, the expected impact range, and the warning level.

7. A tunnel lining health status analysis system based on field monitoring data, characterized in that, include: The acquisition module is configured to acquire multi-point displacement gauge monitoring data sequences, multi-channel acoustic emission monitoring data sequences, and temperature data sequences of internal temperature measuring points of the tunnel lining within a preset time period. The correction module is configured to calculate the temperature change at each moment based on the temperature data sequence, and perform temperature deformation correction on the multi-point displacement gauge monitoring data sequence based on the preset thermal expansion coefficient of the lining material to obtain the net displacement data sequence after removing the temperature influence. The elimination module is configured to identify time periods where the rate of temperature change exceeds a preset threshold based on the temperature data sequence, and eliminate acoustic emission signals that are interfered with by thermal noise in the corresponding time period to obtain a net acoustic emission data sequence. The construction module is configured to construct a constitutive model of lining material damage evolution with damage variables as internal variables. The constitutive model includes a nonlinear relationship between damage driving force and the increment of damage variables. The construction of the constitutive model of lining material damage evolution with damage variables as internal variables includes: The damage variable D is defined to describe the degree of deterioration of the lining material under load. The value of the damage variable D ranges from 0 to 1, where D=0 indicates that the material is intact and D=1 indicates that the material has completely failed. A nonlinear evolution equation is established between the damage driving force and the damage variable increment. The damage driving force is calculated based on the current stress state and the cumulative damage degree. The nonlinear evolution equation is in the form of: dD / dt=f(σ,D,T), where σ is the current stress, D is the current damage variable value, T is the current temperature, and f is a preset nonlinear function. Establish a correlation function between the damage variable D and the cumulative acoustic emission energy E, wherein the correlation function is in the form of: or ,in, and This is a monotonically increasing function obtained by fitting material test data; The nonlinear evolution equation and the correlation function are coupled to form the control equation set of the damage evolution constitutive model; The output module is configured to use the net displacement data sequence as the displacement boundary condition and the net acoustic emission data sequence as the energy dissipation constraint, and input them into the damage evolution constitutive model for finite element inversion calculation to solve for the damage variable values ​​of each unit inside the lining at the current moment, thus forming the damage field distribution. The extraction module is configured to identify continuous spatial regions where the damage variable value exceeds a preset damage threshold based on the damage field distribution, as active damage regions, and extract the geometric center coordinates and area of ​​the active damage regions at each time point. The fitting module is configured to fit the geometric center coordinates and area of ​​the active damage region along the time axis to obtain the spatial migration trajectory of the center of the active damage region and the expansion curve of the area over time. The prediction module is configured to predict the health status assessment results of the active damage area within a future preset time window based on the migration trajectory and the expansion curve.

8. An electronic device, characterized in that, include: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, it implements the method described in any one of claims 1 to 6.

Images