A tunnel health monitoring method based on distributed optical fiber sensing

Abstract

The application provides a tunnel health monitoring method based on distributed optical fiber sensing, and relates to the technical field of tunnel engineering health monitoring.The application adopts an analytic hierarchy process to perform risk assessment on a tunnel, sets a differentiated optical fiber layout scheme according to an assessment result, and lays distributed optical fiber sensors on an inner surface of tunnel lining; obtains original strain data through Brillouin optical time domain analysis, and performs pretreatment by using a sliding average filtering algorithm; builds a finite element model based on actual physical data of the tunnel, inverses strain data into full-field node displacement, and obtains complete strain tensors of finite element units; performs unit-level strain state evaluation, time sequence trend risk evaluation and spatial clustering risk evaluation, and obtains a comprehensive risk value; sets a dynamic evaluation threshold interval, and completes accurate regional health evaluation of the tunnel.The application realizes accurate monitoring and risk evaluation of the health state of the tunnel, and can be widely applied to the safety monitoring field of highway and railway tunnels.

Description

Technical Field

[0001] This invention relates to the field of tunnel engineering health monitoring technology, specifically a tunnel health monitoring method based on distributed optical fiber sensing. Background Technology

[0002] With the acceleration of urbanization and the large-scale construction of transportation infrastructure, tunnels, as important transportation hubs, are receiving increasing attention for their structural safety and operational reliability. During long-term operation, tunnels are subject to multiple influences, including changes in geological conditions, traffic loads, and environmental factors, making them prone to structural damage such as lining cracking, deformation, and leakage, which seriously threaten the safe operation of tunnels. Therefore, establishing an effective tunnel health monitoring system is of great significance for the timely detection of structural anomalies and the prevention of safety accidents.

[0003] Traditional tunnel health monitoring methods primarily rely on point sensors and manual inspections. However, point sensors, such as strain gauges and displacement meters, can only provide monitoring information at discrete points, making it difficult to achieve continuous monitoring of the entire tunnel structure, resulting in blind spots and data discontinuities. While manual inspections can detect obvious structural defects, they are inefficient and cannot achieve real-time dynamic monitoring, failing to meet the needs of modern tunnel safety management.

[0004] In recent years, distributed optical fiber sensing technology has been widely used in tunnel health monitoring due to its advantages such as continuous monitoring, resistance to electromagnetic interference, and good durability. Prior art (CN116792155A) discloses a tunnel health status monitoring and early warning method based on distributed optical fiber sensing, which collects tunnel operation data through a fully distributed sensing optical cable and performs data analysis and safety assessment. Prior art (CN120668051A) provides a tunnel deformation monitoring system based on distributed optical fiber sensing technology, which arranges helically wound optical fibers on the tunnel lining surface and calculates stress and strain through temperature compensation. Prior art (CN105089701A) relates to an operational tunnel health monitoring and early warning system based on distributed optical fiber sensing, which performs real-time online monitoring of the tunnel through an optical fiber sensing subsystem. Prior art (CN120426895A) provides a tunnel deformation parameter inversion method based on distributed optical fiber, which extracts strain information through Brillouin scattering signals and performs three-dimensional reconstruction. The prior art disclosed in CN118366291A relates to an urban tunnel monitoring and emergency alarm system that integrates data acquisition, risk prediction and structural health assessment functions.

[0005] Although existing distributed fiber optic monitoring technologies have made some progress in tunnel health monitoring, the following technical problems still exist: First, the design of existing monitoring schemes lacks risk orientation, failing to differentiate fiber optic deployment according to the risk levels of different tunnel sections, resulting in unreasonable allocation of monitoring resources, insufficient monitoring density in high-risk areas, and wasted resources in low-risk areas; second, the accuracy of data processing needs to be improved, lacking effective methods for strain data inversion and full-field displacement reconstruction, making it difficult to accurately obtain the complete strain state of the tunnel structure; third, the existing risk assessment system is relatively simple, lacking a multi-level, multi-dimensional comprehensive evaluation mechanism, and cannot fully reflect the health status of the tunnel structure; finally, there is a lack of dynamic threshold assessment methods, and existing early warning systems mostly use fixed thresholds, which cannot adapt to the dynamic changes in the tunnel structure state and make it difficult to achieve intelligent early warning.

[0006] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0007] The purpose of this invention is to provide a tunnel health monitoring method based on distributed optical fiber sensing to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A tunnel health monitoring method based on distributed optical fiber sensing includes the following steps:

[0010] S1: The risk assessment of the tunnel is carried out using the analytic hierarchy process (AHP). Based on the risk assessment results, a differentiated fiber optic deployment scheme is set, and distributed fiber optic sensors are deployed on the inner surface of the tunnel lining according to the set fiber optic deployment scheme.

[0011] S2: Set a fixed time monitoring interval, acquire data from the fiber optic sensor, and obtain raw strain data after Brillouin optical time domain analysis. Use the moving average filtering algorithm to preprocess the acquired raw strain data to obtain strain data of the fiber optic sampling points.

[0012] S3: Based on the actual physical data of the tunnel, build an initial finite element model of the tunnel, and invert the strain data into the displacement of all nodal elements through the strain-displacement relationship to obtain the complete strain tensor of all finite element elements.

[0013] S4: Based on the complete strain tensor of the finite element element, perform element-level strain state evaluation, time-series trend risk evaluation and spatial clustering risk evaluation, and obtain the comprehensive risk value of the finite element element based on the above three-level evaluation results.

[0014] S5: Set a dynamic evaluation threshold range, and complete the accurate regional health assessment of the tunnel based on the matching result of the comprehensive risk value of the obtained finite element element and the dynamic evaluation threshold range.

[0015] Furthermore, when using the analytic hierarchy process (AHP) to assess the risk of a tunnel, the tunnel is divided into assessment sections of equal length along its extension direction. Several assessment indicators are set, and the score for each assessment indicator within each assessment section is determined. The scores of each assessment indicator are then weighted and summed to obtain the comprehensive risk score for that assessment section. The assessment indicators include: geological condition indicators, structural characteristic indicators, operating environment indicators, and historical status indicators. When setting differentiated fiber optic deployment schemes, the assessment sections are divided into high-risk, medium-risk, and low-risk sections based on the comprehensive risk score of each assessment section.

[0016] Furthermore, when determining the score for each evaluation indicator within each evaluation section, the geological condition indicators for each evaluation section are calculated using the tunnel surrounding rock quality coefficient obtained from borehole sampling and the groundwater level depth design measured from water level monitoring wells, as follows:

[0017]

[0018] in, For the first The geological condition index scores for each assessment section This is the weighting factor for the tunnel surrounding rock quality coefficient. For section The tunnel surrounding rock quality coefficient, This represents the maximum value of the surrounding rock quality coefficient for the entire tunnel. To monitor the groundwater level depth weighting coefficient, This is the highest groundwater level depth along the entire tunnel. For section The current depth of the groundwater level;

[0019] The structural characteristic indicators for each assessment section are designed using the average crack width of the lining obtained from the crack gauge and the estimated concrete strength obtained from the rebound hammer, as shown in the following formula:

[0020]

[0021] in, For the first The structural characteristic index scores of each evaluation section This is the crack width weighting coefficient. For section The average width of cracks in the lining. The crack width threshold. This is the concrete strength weighting coefficient. For section Estimated concrete strength To design the standard value of concrete strength;

[0022] The operational environment indicators for each assessment section are calculated using the annual average traffic volume obtained from traffic flow monitoring equipment and the proportion of heavy vehicles obtained from vehicle classification statistics, as follows:

[0023]

[0024] in, For the first The operating environment index scores for each assessment segment Traffic volume weighting coefficient, For section The average annual traffic volume. This represents the highest traffic volume across the entire cableway line. This is the weighting coefficient for the proportion of heavy vehicles. For section The proportion of heavy vehicles, The maximum proportion of heavy vehicles along the entire tunnel; the historical status indicators for each assessment section are obtained through maintenance frequency statistics and cumulative maintenance area statistics, as shown in the following formulas:

[0025]

[0026] in, For the first Historical status index scores for each assessment segment This is the weighting coefficient for the number of repairs. For section Number of repairs in the past 5 years This represents the highest number of annual maintenance visits for the entire tunnel line. The maintenance area weighting coefficient. For section The cumulative repair area, For section The total lining area.

[0027] Furthermore, the logic for deploying distributed fiber optic sensors is as follows:

[0028] For the assessment section of the high-risk section, longitudinal backbone sensing optical fibers are laid along the inner surface of the tunnel arch lining and the inner surface of the arch waist lining on both sides, and circumferential sensing optical fibers are laid at fixed length intervals on the inner surface of the lining.

[0029] For the assessment section of medium risk, longitudinal backbone sensing optical fibers are laid along the inner surface of the tunnel arch lining, and circumferential sensing optical fibers are laid along the inner surface of the lining at the midpoint of the section.

[0030] For the assessment section which is a low-risk section, longitudinal sensing optical fibers are only laid on the inner surface of the tunnel arch lining.

[0031] Among them, the low, medium and high risk sections are respectively set with corresponding low, medium and high risk section interval spatial resolution as fixed intervals; different risk sections are set with fiber sampling points at corresponding fixed intervals on the laid longitudinal backbone sensing fiber or circumferential sensing fiber.

[0032] Furthermore, the raw strain data of the fiber optic sampling points are obtained using Brillouin optical time-domain analysis technology. The specific calculation formula is as follows:

[0033]

[0034] in, for Time-of-flight fiber sampling points The original strain data at the location, for Time-of-flight fiber sampling points Frequency shift of Brillouin scattered light at that location denoted as the strain coefficient of the optical fiber.

[0035] Furthermore, the specific steps for preprocessing using the adaptive moving average filtering algorithm are as follows: Set the filter window width to... Several fiber optic sampling points are taken, with the fiber optic sampling point to be preprocessed as the center, and samples are taken before and after it. The original strain data from each fiber optic sampling point are averaged using the following formula:

[0036]

[0037] Calculate the mean value of the strain data of the entire tunnel at the current moment. and standard deviation , will satisfy Data points that are not considered outliers are replaced with linear interpolation of adjacent points; additionally, preprocessing requires the deployment of temperature-compensated optical fibers of the same specification to read their values. Real-time strain data was obtained after temperature compensation. The formula is as follows:

[0038] in, for Time-of-flight fiber sampling points Strain data at the location, for Time-of-flight fiber sampling points Pure temperature strain measured by temperature-compensated fiber optic cable.

[0039] Furthermore, an initial finite element model of the tunnel is constructed based on the actual physical data of the tunnel. Specifically, the actual physical data of the tunnel lining are obtained, discretized using eight-node hexahedral solid elements, and the elastic modulus and Poisson's ratio are calibrated. An initial constitutive matrix is ​​constructed based on the elastic modulus and Poisson's ratio. The strain signal is inverted into the full-field nodal displacement through the strain-displacement relationship.

[0040] The total nodal displacement at each moment is obtained by using a vector composed of strain data from all fiber optic sampling points at each time step, based on the following formula:

[0041]

[0042] in, The vector formed by the strain data of all fiber optic sampling points. This is the transformation matrix between the geometric relationship between strain and displacement. Let be the total nodal displacement vector; the optimal nodal displacement at each time step is determined using the following formula:

[0043]

[0044] in, This represents the optimal total nodal displacement. For regularization parameters, Let be the Laplacian operator matrix; calculate the complete strain tensor of the finite element at each time step using the standard finite element formula, as follows:

[0045]

[0046] in, For the complete strain tensor of the finite element element, The constitutive matrix is It is the strain-displacement transformation matrix of the element. It is the nodal displacement vector of the element.

[0047] Furthermore, based on the complete strain tensor of each finite element, the element-level strain state evaluation is specifically performed as follows: the complete strain tensor of the finite element at each moment is transformed into the equivalent strain of the finite element at that moment, according to the formula:

[0048] in, Finite element Equivalent change, Finite element The three normal strains of the complete strain tensor Poisson's ratio; based on concrete cracking strain threshold Set dual thresholds for judgment: Marked as safe Marked as a warning. Mark as an alarm;

[0049] The specific steps for conducting time-series trend risk assessment are as follows: The equivalent strains of finite element elements marked as early warning and alarm at each time point are sorted chronologically to form the equivalent strain time series for each element. This time series is then analyzed, and a short-term prediction is made using an autoregressive integral moving average model. Based on this model, future predictions are then made. The time-series trend risk index is calculated based on the strain value after a certain period, using the following formula:

[0050]

[0051] in, Finite element The time-series trend risk index Predicting the future using models Equivalent variable after time, This represents the standard deviation of the equivalent variation historical data for this unit.

[0052] The spatial clustering risk assessment specifically involves: identifying finite element elements marked as warning or alarm, and using the DBSCAN spatial clustering algorithm to perform cluster analysis on the positions of these marked finite element elements in three-dimensional space, forming a cluster of abnormal elements in the finite element model. For finite element elements If it belongs to a high-risk cluster The formula for calculating its spatial risk score is as follows:

[0053]

[0054] in, High-risk cluster Number of units included High-risk cluster The outer envelope area, This is the scaling factor.

[0055] Furthermore, based on the three-level evaluation results, the comprehensive risk value of the finite element is obtained, and the calculation formula is as follows:

[0056]

[0057] in, Finite element exist The overall risk value at any given moment. Let be the weighting coefficient, satisfying and The dynamic evaluation threshold range is set as follows: a safety threshold upper limit, a warning threshold upper limit, and an alarm threshold upper limit are set. When the comprehensive risk value is less than the safety threshold upper limit, the tunnel area where the current finite element element is located is marked as a normal element. When the comprehensive risk value is between the safety threshold upper limit and the warning threshold upper limit, the tunnel area where the current finite element element is located is marked as a warning element. When the comprehensive risk value is between the warning threshold upper limit and the alarm threshold upper limit, the tunnel area where the current finite element element is located is marked as a dangerous alarm element. When the comprehensive risk value is greater than the alarm threshold upper limit, the tunnel area where the current finite element element is located is marked as a severely dangerous element.

[0058] Compared with existing technologies, the beneficial effects of this invention are as follows: By employing a risk-oriented differentiated fiber optic deployment scheme, the efficiency of monitoring resource utilization is improved compared to the traditional uniform deployment method, allowing for focused attention on high-risk areas; by fusing multi-source data and combining geological, structural, operational, and historical data from multiple dimensions, the evaluation results are more comprehensive and accurate; high spatial resolution strain monitoring is achieved through distributed fiber optics, overcoming the limitations of point-based sensors and enabling continuous monitoring across the entire field; a multi-level evaluation system that comprehensively evaluates from three dimensions—unit status, temporal trend, and spatial distribution—is more scientific and reliable than a single evaluation method; dynamic adaptive thresholds enable dynamic classification and early warning of health status, improving the accuracy and timeliness of early warnings; and refined and intelligent monitoring and early warning of tunnel structural status are achieved, significantly enhancing the level of tunnel operational safety assurance. Attached Figure Description

[0059] Figure 1 This is a schematic diagram of the overall method flow of the present invention;

[0060] Figure 2 This is a data diagram of the risk zone in an embodiment of the present invention. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0062] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0063] Example:

[0064] Please see Figures 1-2 The present invention provides a technical solution:

[0065] A tunnel health monitoring method based on distributed optical fiber sensing includes the following steps:

[0066] S1: The risk assessment of the tunnel is carried out using the analytic hierarchy process (AHP). Based on the risk assessment results, a differentiated fiber optic deployment scheme is set, and distributed fiber optic sensors are deployed on the inner surface of the tunnel lining according to the set fiber optic deployment scheme.

[0067] In this embodiment, when using the analytic hierarchy process (AHP) to assess the risk of a tunnel, the tunnel is divided into assessment sections of equal length along its extension direction. Several assessment indicators are set, and the score for each assessment indicator within each assessment section is determined. The scores of each assessment indicator are then weighted and summed to obtain the comprehensive risk score for that assessment section. The assessment indicators include: geological condition indicators, structural characteristic indicators, operating environment indicators, and historical status indicators. The comprehensive risk score for each section is calculated using the following formula:

[0068]

[0069] in, For the first The overall risk score for each assessment section For the first The geological condition index scores for each assessment section For the first The structural characteristic index scores of each evaluation section For the first The operating environment index scores for each assessment segment For the first Historical status index scores for each assessment segment For the first The weights of the four indicators are as follows: When setting differentiated fiber optic deployment schemes, the assessment sections are divided into high-risk, medium-risk, and low-risk sections based on the comprehensive risk score of each assessment section. In this embodiment, a 100m long tunnel is selected for simulation experiments. The entire tunnel is divided into 5 sections at fixed intervals of 20m. The weights of the four assessment indicators are as follows: .

[0070] In this embodiment, the score of each evaluation index within each evaluation section is determined. The geological condition index of each evaluation section is determined by the tunnel surrounding rock quality coefficient obtained from borehole sampling and the groundwater level depth design measured by water level monitoring boreholes, as follows:

[0071]

[0072] in, For the first The geological condition index scores for each assessment section This is the weighting factor for the tunnel surrounding rock quality coefficient. For section The tunnel surrounding rock quality coefficient, This represents the maximum value of the surrounding rock quality coefficient for the entire tunnel. To monitor the groundwater level depth weighting coefficient, This is the highest groundwater level depth along the entire tunnel. For section The current groundwater level depth; in this embodiment, , , , The surrounding rock quality coefficient is a core parameter for rock mass engineering classification; the water level depth directly affects the water load of the lining; the better the rock mass quality and the lower the water level, the lower the geological risk.

[0073] The structural characteristic indicators for each assessment section are designed using the average crack width of the lining obtained from the crack gauge and the estimated concrete strength obtained from the rebound hammer, as shown in the following formula:

[0074]

[0075] in, For the first The structural characteristic index scores of each evaluation section This is the crack width weighting coefficient. For section The average width of cracks in the lining. The crack width threshold. This is the concrete strength weighting coefficient. For section Estimated concrete strength To design the standard value of concrete strength; in this embodiment, , , , Crack width directly reflects the degree of damage; concrete strength reflects material properties. The smaller the crack and the higher the strength, the safer the structure.

[0076] The operational environment indicators for each assessment section are calculated using the annual average traffic volume obtained from traffic flow monitoring equipment and the proportion of heavy vehicles obtained from vehicle classification statistics, as follows:

[0077]

[0078] in, For the first The operating environment index scores for each assessment segment Traffic volume weighting coefficient, For section The average annual traffic volume. This represents the highest traffic volume across the entire cableway line. This is the weighting coefficient for the proportion of heavy vehicles. For section The proportion of heavy vehicles, This represents the maximum proportion of loaded vehicles along the entire tunnel; in this embodiment, , , , The impact of dynamic load is reflected by traffic volume and the proportion of loaded vehicles. The greater the traffic load, the more significant the impact of dynamic load and the higher the risk.

[0079] The historical status indicators for each assessment section are calculated using the number of repairs obtained from the repair frequency statistics and the cumulative repair area obtained from the repair area statistics, as follows:

[0080]

[0081] in, For the first Historical status index scores for each assessment segment This is the weighting coefficient for the number of repairs. For section Number of repairs in the past 5 years This represents the highest number of annual maintenance visits for the entire tunnel line. The maintenance area weighting coefficient. For section The cumulative repair area, For section The total lining area; in this embodiment, , , , Maintenance records reflect the history of cumulative damage, and the frequency and scope of historical maintenance reflect the degree of cumulative structural damage.

[0082] A high-risk threshold and a low-risk threshold are set for each segment. When the overall risk score is greater than the low-risk threshold, the segment is classified as a low-risk segment. When the overall risk score is less than the low-risk threshold but greater than the high-risk threshold, the segment is classified as a medium-risk segment. When the overall risk score is less than the high-risk threshold, the segment is classified as a high-risk segment. In this embodiment, the high-risk threshold is 0.5 and the low-risk threshold is 0.75.

[0083] In this embodiment, the assessment section is a high-risk section. A longitudinal backbone sensing fiber is laid along the inner surface of the tunnel arch lining and the inner surface of the side arch linings, and a circumferential sensing fiber is laid at fixed length intervals on the inner surface of the lining.

[0084] For the assessment section of medium risk, longitudinal backbone sensing optical fibers are laid along the inner surface of the tunnel arch lining, and circumferential sensing optical fibers are laid along the inner surface of the lining at the midpoint of the section.

[0085] For the assessment section which is a low-risk section, longitudinal sensing optical fibers are only laid on the inner surface of the tunnel arch lining.

[0086] In this embodiment, corresponding spatial resolutions for low, medium, and high-risk sections are set as fixed intervals. For different risk sections, fiber sampling points are set at corresponding fixed intervals on the laid longitudinal backbone sensing fiber or circumferential sensing fiber. The data required for each evaluation index of each of the five sections in this embodiment are obtained as shown in Table 1 below:

[0087] Table 1: Data Required for Section Evaluation Indicators

[0088]

[0089] Table 1 reflects the significant differences in geotechnical environment, structural condition, operational load, and historical maintenance across different tunnel sections. High-risk section 3 exhibits a complex deterioration characterized by weak surrounding rock, shallow water level, wide cracks, low concrete strength, heavy traffic load, and frequent maintenance; all indicators point to high risk. Conversely, low-risk section 5 is characterized by solid surrounding rock, deep water level, fine cracks, high concrete strength, moderate traffic, and minimal maintenance—an ideal state. The indicators for medium-risk sections 2 and 4 fall within a transitional zone, revealing differences in the causes of risk. Section 2 focuses on structural defects and the impact of heavy vehicles, while section 4 is more affected by the combined effects of geology and operational load. The comprehensive risk score for each section is calculated, as shown in Table 2 below.

[0090] Table 2: Scores of Each Assessment Indicator and Overall Risk Rating for the Assessment Section

[0091]

[0092] Please see Figure 2 The risk assessment data, including scores for each assessment indicator and a comprehensive risk score, clearly reveals the differences in structural health status across different tunnel sections. High-risk section 3 shows significant deterioration in geological conditions and structural condition, indicating poor surrounding rock quality and well-developed lining cracks, making it a key area for safety monitoring. While medium-risk sections 2 and 4 are generally acceptable, they exhibit weaknesses such as insufficient structural characteristics and excessively high operational loads, respectively. Low-risk sections 1 and 5 show balanced and excellent performance across all indicators, with section 5 being nearly perfect in terms of geology and structure. The data clearly demonstrates the necessity of risk-oriented differentiated monitoring; resources should be prioritized for high-risk sections while closely monitoring the development of specific defects in medium-risk sections.

[0093] Taking section 3 as an example, it was determined to be a high-risk area based on comprehensive scoring data. Longitudinal backbone sensing fibers were laid along the inner surface of the tunnel arch lining and the inner surface of the side arch linings in the high-risk section. Circumferential sensing fibers were laid on the inner surface of the lining at fixed length intervals of 5m. The spatial resolution of the high-risk section was 0.5m, with fiber sampling points set at fixed intervals of 0.5m. A total of 376 sampling points were laid in section 3.

[0094] This step integrates four dimensions—geology, structure, operation, and history—to form a scientific risk quantification assessment framework, reducing the construction cost of the monitoring system. Traditional methods, such as uniform fiber optic deployment, result in significant waste of monitoring resources and potential under-monitoring of high-risk areas. This solution, based on differentiated deployment using AHP (Advanced Hierarchical Hierarchy Processing), achieves precise allocation of monitoring resources, providing a reasonable spatial sampling basis for subsequent data processing. It ensures that limited monitoring resources are concentrated in the areas requiring the most attention and provides an initial risk benchmark for setting dynamic thresholds.

[0095] S2: Set a fixed time monitoring interval, acquire data from the fiber optic sensor, and obtain raw strain data after Brillouin optical time domain analysis. Use the moving average filtering algorithm to preprocess the acquired raw strain data to obtain strain data of the fiber optic sampling points.

[0096] In this embodiment, the raw strain data of the fiber optic sampling points are obtained using Brillouin optical time-domain analysis technology. The specific calculation formula is as follows:

[0097]

[0098] in, for Time-of-flight fiber sampling points The original strain data at the location, for Time-of-flight fiber sampling points Frequency shift of Brillouin scattered light at that location denoted as the strain coefficient of the optical fiber. In this embodiment, the fixed monitoring interval is set to 1 hour.

[0099] Raw data It contains high-frequency noise. A moving average filtering algorithm is used for preprocessing. This algorithm is simple to calculate, has good real-time performance, and can effectively smooth random noise without introducing phase delay. In this embodiment, the specific steps of preprocessing using the adaptive moving average filtering algorithm are as follows: Set the filtering window width to... Several fiber optic sampling points are taken, with the fiber optic sampling point to be preprocessed as the center, and samples are taken before and after it. The original strain data from each fiber optic sampling point are averaged using the following formula:

[0100]

[0101] Calculate the mean value of the strain data of the entire tunnel at the current moment. and standard deviation , will satisfy Data points are considered outliers and replaced with linear interpolation of adjacent points.

[0102]

[0103]

[0104] In this embodiment That is, selecting fiber optic sampling points. right Rounding down, meaning the two consecutive fiber optic sampling points are considered neighboring sampling points; additionally, preprocessing requires laying out temperature-compensated fiber optic cables of the same specification and reading their data. Real-time strain data obtained after temperature compensation The formula is as follows:

[0105] in, for Time-of-flight fiber sampling points Strain data at the location, for Time-of-flight fiber sampling points Pure temperature strain measured by temperature-compensated fiber optic cable. This refers to the high-quality fiber optic sampling point strain data that has been preprocessed and is used for subsequent analysis. The following relationship must be satisfied:

[0106]

[0107] in, For temperature coefficient, for Time-of-flight fiber sampling points Temperature changes at the location. This step aims to eliminate the interference of temperature on fiber optic strain measurements, improve data reliability, and ensure that the temperature-compensated fiber and the sensing fiber are deployed in parallel, so they are affected by the same temperature field but not by mechanical strain; noise is smoothed by utilizing the correlation of spatially adjacent points; window width. The larger the value, the better the smoothing effect, but it will blur details. Compared with complex adaptive filtering algorithms, moving average filtering is sufficient to meet the noise reduction requirements in scenarios with relatively slow strain changes, such as tunnels. Moreover, the algorithm is stable, easy to implement in embedded systems, and ensures the long-term reliability of the entire system.

[0108] S3: Based on the actual physical data of the tunnel, build an initial finite element model of the tunnel, and invert the strain data into the displacement of all nodal elements through the strain-displacement relationship to obtain the complete strain tensor of all finite element elements.

[0109] In this embodiment, an initial finite element model of the tunnel is built based on the actual physical data of the tunnel. Specifically, the actual physical data of the tunnel lining is obtained, and the lining is discretized using eight-node hexahedral solid elements. The elastic modulus and Poisson's ratio are calibrated, and an initial constitutive matrix is ​​constructed based on the elastic modulus and Poisson's ratio.

[0110] The strain signal is inverted into the full-field nodal displacement through the strain-displacement relationship, specifically as follows:

[0111] The total nodal displacement at each moment is obtained by using a vector composed of strain data from all fiber optic sampling points at each time step, based on the following formula:

[0112]

[0113] in, The vector formed by the strain data of all fiber optic sampling points. This is the transformation matrix between the geometric relationship between strain and displacement; A geometric relationship between strain at each measurement point and the nodal displacement of the element containing that point was established. This relationship is deterministic, determined by the element shape function and geometric position; the strain observation equation is usually underdetermined, with infinitely many solutions; adding a regularization term can constrain the smoothness of the solution, making the inversion problem well-posed. Let U be the displacement vector of all nodes in the field. While U has a high dimension, the actual displacement field possesses spatial smoothness and physical plausibility, reducing the effective degrees of freedom. The optimal node displacement at each moment is solved using the following formula:

[0114]

[0115] in, This represents the optimal total nodal displacement. For regularization parameters, Here is the Laplace operator matrix; the Laplace smoothing operator is used to constrain the smooth changes in displacements between adjacent nodes, reflecting the assumption of continuity in structural deformation; regularization parameter. To control the balance between data fitting and smoothness, Too small: The solution is greatly affected by noise and is unstable. Too large: The solution is too smooth and may mask the true deformation; the complete strain tensor of the finite element at each time step is calculated using the standard finite element formula, as follows:

[0116]

[0117]

[0118] in, For the complete strain tensor of the finite element element, The constitutive matrix is It is a matrix that describes the physical relationship between stress and strain in a material. Based on the linear elastic assumption, it is suitable for the response of concrete in the small deformation stage. If the material enters the nonlinear stage, such as cracking, a nonlinear constitutive model needs to be used, but this will increase the computational complexity. For elastic modulus, Poisson's ratio, It is the transformation matrix between the geometric relationship between the strain and displacement of an element. It is the nodal displacement vector of the element.

[0119] This step uses only the strain data from discrete points to invert the displacement field of the entire tunnel, achieving a leap from point monitoring to field analysis; based on the fundamental geometric relationship between strain and displacement, i.e. However, there is an underdetermined problem, namely, the number of observation points is far less than the number of nodes in the finite element method. Direct solution will yield infinitely many solutions. This scheme adopts a regularization method, where the displacements of adjacent points are correlated and the overall displacement field is smooth and continuous. This prior knowledge is introduced through regularization terms, which is equivalent to adding a large number of implicit constraints, making the problem change from underdetermined to well-determined.

[0120] S4: Based on the complete strain tensor of the finite element element, perform element-level strain state evaluation, time-series trend risk evaluation and spatial clustering risk evaluation, and obtain the comprehensive risk value of the finite element element based on the above three-level evaluation results.

[0121] In this embodiment, the element-level strain state evaluation based on the complete strain tensor of each finite element is specifically performed by converting the complete strain tensor of the finite element at each moment into the equivalent strain of the finite element at that moment, according to the formula:

[0122] in, Finite element Equivalent change, Finite element The three normal strains of the complete strain tensor Poisson's ratio; as the principal strain difference increases, the equivalent strain increases, reflecting a more complex stress state; As the denominator increases, the equivalent strain decreases, reflecting that the lateral deformation shares some of the strain energy.

[0123] Based on concrete cracking strain threshold Set dual thresholds for judgment: Marked as safe Marked as a warning. It is marked as an alarm; in this embodiment, The strain data of some finite element elements in high-risk section 3 are shown in Table 3 below:

[0124] Table 3: Strain state of finite element elements in section 3

[0125]

[0126] Table 3 reveals the localized damage concentration phenomenon in the 45m arch waist region of the tunnel lining in this embodiment. All four adjacent elements exhibit significant tensile stress at principal strain 1 and compressive stress at principal strain 3, indicating that this region is under a complex combined tensile and compressive stress state. The calculated equivalent strain generally exceeds the concrete warning threshold, and the finite element elements shown enter an alarm / warning state. Spatially, the data shows a distribution pattern of attenuation towards both sides centered on the 90° arch waist, with an increasing trend along the tunnel axis; this confirms that the damage not only has circumferential locality but also carries the risk of axial expansion, providing clear abnormal element group targets for subsequent spatial clustering analysis, suggesting the need for focused attention and intensified monitoring of this area.

[0127] Under multiaxial stress, the strain in a single direction cannot accurately reflect the true damage state of concrete. The Von Mises criterion transforms the complex multiaxial strain state into a scalar index. This criterion has been successfully applied in metallic materials and, after appropriate modification, is applicable to concrete.

[0128] The specific steps for conducting time-series trend risk assessment are as follows: The equivalent strains of finite element elements marked as early warning and alarm at each time point are sorted chronologically to form the equivalent strain time series for each element. This time series is then analyzed, and a short-term prediction is made using an autoregressive integral moving average model. Based on this model, future predictions are then made. The time-series trend risk index is calculated based on the strain value after a certain period, using the following formula:

[0129]

[0130] in, Finite element The time-series trend risk index Predicting the future using models Equivalent variable after time, The standard deviation of the equivalent strain history data for this unit; numerator It indicates the magnitude of strain change during the forecast period, reflecting the speed of change rather than the direction. As the denominator, Eliminate the differences in the inherent volatility of different units. Considering the cumulative effect over time, the risk increases with the forecast period. In this embodiment, Using alarm finite element units Taking a certain monitoring moment as an example, Predicting the future using models The finite element element after Equivalent strain value, calculation of finite element elements Time-series trend risk index:

[0131]

[0132] The spatial clustering risk assessment specifically involves: identifying finite element elements marked as warning or alarm, and using the DBSCAN spatial clustering algorithm to perform cluster analysis on the positions of these marked finite element elements in three-dimensional space, forming a cluster of abnormal elements in the finite element model. For finite element elements If it belongs to a high-risk cluster The formula for calculating its spatial risk score is as follows:

[0133]

[0134] in, High-risk cluster The number of units included represents the scale of damage in this scheme; the more units, the higher the systemic risk. High-risk cluster The outer envelope area represents the density of damage in this scheme. The smaller the area, the more concentrated the damage and the more significant the local effect. The scaling factor is used to adjust the areal density value to an appropriate order of magnitude in the formula. The spatial clustering risk assessment step aims to collect all units marked as warning or alarm, run the DBSCAN algorithm to identify dense and abnormal regions, and calculate the number of units and spatial extent for each cluster; in this embodiment, alarm finite element units are used. Taking the high-risk cluster C1 as an example, the number of units in high-risk cluster C1 is 42, and the outer envelope area is... , Set to 0.1; calculate finite element elements. Spatial risk score, In this embodiment, the comprehensive risk value of the finite element is obtained based on the three-level evaluation results, and the calculation formula is as follows:

[0135]

[0136] in, Finite element exist The overall risk value at any given moment. Let be the weighting coefficient, satisfying and In this embodiment, , obtain finite element The overall risk value is 0.8913.

[0137] Existing technologies, based solely on comparing current strain values ​​with fixed thresholds, cannot identify slowly developing, progressive damage. Isolated point analysis analyzes each monitoring point independently, ignoring the spatial correlation of damage. Passive responses can only issue alarms after damage occurs, lacking predictive and early warning capabilities. Threshold setting relies on engineer experience and lacks scientific quantitative basis. This solution combines immediate status, temporal trends, and spatial distribution to comprehensively assess structural risk. Equivalent strain and dual-threshold mechanisms simplify the complex strain tensor into a single index, facilitating engineering judgment. Through temporal analysis, it predicts damage development trends, distinguishes between isolated anomalies and systemic damage, identifies damage clusters, and transforms qualitative judgments into quantitative risk values, supporting scientific decision-making.

[0138] S5: Set a dynamic evaluation threshold range, and complete the accurate regional health assessment of the tunnel based on the matching result of the comprehensive risk value of the obtained finite element element and the dynamic evaluation threshold range.

[0139] In this embodiment, the dynamic evaluation threshold range is specifically set as follows: a safety threshold upper limit, a warning threshold upper limit, and an alarm threshold upper limit are set. When the comprehensive risk value is less than the safety threshold upper limit, the tunnel area where the current finite element element is located is marked as a normal element. When the comprehensive risk value is between the safety threshold upper limit and the warning threshold upper limit, the tunnel area where the current finite element element is located is marked as a warning element. When the comprehensive risk value is between the warning threshold upper limit and the alarm threshold upper limit, the tunnel area where the current finite element element is located is marked as a dangerous alarm element. When the comprehensive risk value is greater than the alarm threshold upper limit, the tunnel area where the current finite element element is located is marked as a severely dangerous element. In this embodiment, the safety threshold upper limit is 0.52, the warning threshold upper limit is 0.65, and the alarm threshold upper limit is 0.82. Therefore, the finite element element... The overall risk value is 0.8913, marking the tunnel area where the current finite element element is located as a severely dangerous element.

[0140] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0141] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0142] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0143] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A tunnel health monitoring method based on distributed optical fiber sensing, characterized in that, The specific steps include: S1: The risk assessment of the tunnel is carried out using the analytic hierarchy process (AHP). Based on the risk assessment results, a differentiated fiber optic deployment scheme is set, and distributed fiber optic sensors are deployed on the inner surface of the tunnel lining according to the set fiber optic deployment scheme. S2: Set a fixed time monitoring interval, acquire data from the fiber optic sensor, and obtain raw strain data after Brillouin optical time domain analysis. Use the moving average filtering algorithm to preprocess the acquired raw strain data to obtain strain data of the fiber optic sampling points. S3: Based on the actual physical data of the tunnel, build an initial finite element model of the tunnel, and invert the strain data into the displacement of all nodal elements through the strain-displacement relationship to obtain the complete strain tensor of all finite element elements. S4: Based on the complete strain tensor of the finite element element, perform element-level strain state evaluation, time-series trend risk evaluation and spatial clustering risk evaluation, and obtain the comprehensive risk value of the finite element element based on the above three-level evaluation results. S5: Set a dynamic assessment threshold range, and complete the accurate regional health assessment of the tunnel based on the matching result of the comprehensive risk value of the obtained finite element element and the dynamic assessment threshold range; When using the analytic hierarchy process (AHP) to assess tunnel risk, the tunnel is divided into assessment sections of equal length along its extension direction. Several assessment indicators are set, and the score for each assessment indicator within each assessment section is determined. The scores of each assessment indicator are then weighted and summed to obtain the comprehensive risk score for that assessment section. The assessment indicators include: geological condition indicators, structural characteristic indicators, operating environment indicators, and historical status indicators. When setting differentiated fiber optic deployment schemes, the assessment sections are divided into high-risk, medium-risk, and low-risk sections based on the comprehensive risk score of each assessment section. When determining the score for each evaluation indicator within each evaluation section, the geological condition indicators for each evaluation section are calculated using the tunnel surrounding rock quality coefficient obtained from borehole sampling and the groundwater level depth design measured from water level monitoring wells, as follows: ； in, For the first The geological condition index scores for each assessment section This is the weighting factor for the tunnel surrounding rock quality coefficient. For section The tunnel surrounding rock quality coefficient, This represents the maximum value of the surrounding rock quality coefficient for the entire tunnel. To monitor the groundwater level depth weighting coefficient, This is the highest groundwater level depth along the entire tunnel. For section The current depth of the groundwater level; The structural characteristic indicators for each assessment section are designed using the average crack width of the lining obtained from the crack gauge and the estimated concrete strength obtained from the rebound hammer, as shown in the following formula: ； in, For the first The structural characteristic index scores of each evaluation section This is the crack width weighting coefficient. For section The average width of cracks in the lining. The crack width threshold. This is the concrete strength weighting coefficient. For section Estimated concrete strength To design the standard value of concrete strength; The operational environment indicators for each assessment section are calculated using the annual average traffic volume obtained from traffic flow monitoring equipment and the proportion of heavy vehicles obtained from vehicle classification statistics, as follows: ； in, For the first The operating environment index scores for each assessment segment Traffic volume weighting coefficient, For section The average annual traffic volume. This represents the highest traffic volume across the entire cableway line. This is the weighting coefficient for the proportion of heavy vehicles. For section The proportion of heavy vehicles, The maximum proportion of heavy vehicles along the entire tunnel; the historical status indicators for each assessment section are obtained through maintenance frequency statistics and cumulative maintenance area statistics, as shown in the following formulas: ； in, For the first Historical status index scores for each assessment segment This is the weighting coefficient for the number of repairs. For section Number of repairs in the past 5 years This represents the highest number of annual maintenance visits for the entire tunnel line. This is the weighting coefficient for the repair area. For section The cumulative repair area, For section The total lining area; The logic for deploying distributed fiber optic sensors is as follows: For the assessment section of the high-risk section, longitudinal backbone sensing fibers are deployed along the inner surface of the tunnel arch lining and the inner surface of the side arch linings, and circumferential sensing fibers are deployed at fixed length intervals on the inner surface of the lining. For the assessment section of medium risk, longitudinal backbone sensing optical fibers are laid along the inner surface of the tunnel arch lining, and circumferential sensing optical fibers are laid along the inner surface of the lining at the midpoint of the section. For the assessment section which is a low-risk section, longitudinal sensing optical fibers are only laid on the inner surface of the tunnel arch lining. Among them, the low, medium and high risk sections are respectively set with corresponding low, medium and high risk section interval spatial resolution as fixed intervals; different risk sections are set with fiber sampling points at corresponding fixed intervals on the laid longitudinal backbone sensing fiber or circumferential sensing fiber.

2. The tunnel health monitoring method based on distributed optical fiber sensing according to claim 1, characterized in that: The raw strain data of the fiber optic sampling points were obtained using Brillouin optical time-domain analysis. The specific calculation formula is as follows: ； in, for Time-of-flight fiber sampling points The original strain data at the location, for Time-of-flight fiber sampling points Frequency shift of Brillouin scattered light at that location denoted as the strain coefficient of the optical fiber.

3. The tunnel health monitoring method based on distributed optical fiber sensing according to claim 2, characterized in that: The specific steps for preprocessing using the adaptive moving average filtering algorithm are as follows: Set the filter window width to... Several fiber optic sampling points are taken, with the fiber optic sampling point to be preprocessed as the center, and samples are taken before and after it. The original strain data from each fiber optic sampling point are averaged using the following formula: ； Calculate the mean value of the strain data of the entire tunnel at the current moment. and standard deviation , will satisfy Data points that are not considered outliers are replaced with linear interpolation of adjacent points; additionally, preprocessing requires the deployment of temperature-compensated optical fibers of the same specification to read their values. Real-time strain data was obtained after temperature compensation. The formula is as follows: ； in, for Time-of-flight fiber sampling points Strain data at the location, for Time-of-flight fiber sampling points Pure temperature strain measured by temperature-compensated fiber optic cable.

4. The tunnel health monitoring method based on distributed optical fiber sensing according to claim 3, characterized in that: An initial finite element model of the tunnel was constructed based on actual physical data. Specifically, the actual physical data of the tunnel lining was obtained, discretized using eight-node hexahedral solid elements, and the elastic modulus and Poisson's ratio were calibrated. An initial constitutive matrix was constructed based on the elastic modulus and Poisson's ratio. The strain signal was inverted into full-field nodal displacements through the strain-displacement relationship. The total nodal displacement at each moment is obtained by using a vector composed of strain data from all fiber optic sampling points at each time step, based on the following formula: ； in, The vector formed by the strain data of all fiber optic sampling points. This is the transformation matrix between the geometric relationship between strain and displacement. Let be the total nodal displacement vector; the optimal nodal displacement at each time step is determined using the following formula: ； in, This represents the optimal total nodal displacement. For regularization parameters, Let be the Laplacian operator matrix; calculate the complete strain tensor of the finite element at each time step using the standard finite element formula, as follows: ； in, For the complete strain tensor of the finite element element, The constitutive matrix is It is the strain-displacement transformation matrix of the element. It is the nodal displacement vector of the element.

5. A tunnel health monitoring method based on distributed optical fiber sensing according to claim 4, characterized in that: The element-level strain state evaluation based on the complete strain tensor of each finite element is specifically as follows: the complete strain tensor of the finite element at each moment is transformed into the equivalent strain of the finite element at that moment, according to the formula: ； in, Finite element Equivalent change, Finite element The three normal strains of the complete strain tensor Poisson's ratio; based on concrete cracking strain threshold Set dual thresholds for judgment: Marked as safe Marked as a warning. Mark as an alarm; The specific steps for conducting time-series trend risk assessment are as follows: The equivalent strains of finite element elements marked as early warning and alarm at each time point are sorted chronologically to form the equivalent strain time series for each element. This time series is then analyzed, and a short-term prediction is made using an autoregressive integral moving average model. Based on this model, future predictions are then made. The time-series trend risk index is calculated based on the strain value after a certain period, using the following formula: ； in, Finite element The time-series trend risk index Predicting the future using models Equivalent variable after time, This represents the standard deviation of the equivalent variation historical data for this unit. The spatial clustering risk assessment specifically involves: identifying finite element elements marked as warning or alarm, and using the DBSCAN spatial clustering algorithm to perform cluster analysis on the positions of these marked finite element elements in three-dimensional space, forming a cluster of abnormal elements in the finite element model. For finite element elements If it belongs to a high-risk cluster The formula for calculating its spatial risk score is as follows: ； in, High-risk cluster Number of units included High-risk cluster The outer envelope area, This is the scaling factor.

6. The tunnel health monitoring method based on distributed optical fiber sensing according to claim 5, characterized in that: The comprehensive risk value of the finite element element is obtained based on the three-level evaluation results, and the calculation formula is as follows: ;in, Finite element exist The overall risk value at any given moment. Let be the weighting coefficient, satisfying and The dynamic evaluation threshold range is set as follows: a safety threshold upper limit, a warning threshold upper limit, and an alarm threshold upper limit are set. When the comprehensive risk value is less than the safety threshold upper limit, the tunnel area where the current finite element element is located is marked as a normal element. When the comprehensive risk value is between the safety threshold upper limit and the warning threshold upper limit, the tunnel area where the current finite element element is located is marked as a warning element. When the comprehensive risk value is between the warning threshold upper limit and the alarm threshold upper limit, the tunnel area where the current finite element element is located is marked as a dangerous alarm element. When the comprehensive risk value is greater than the alarm threshold upper limit, the tunnel area where the current finite element element is located is marked as a severely dangerous element.

Images