A dam displacement detection method based on a sensing system

Abstract

The application relates to the technical field of dam safety monitoring, and discloses a dam displacement detection method based on a sensing system. A two-dimensional rectangular coordinate system is established on a water-facing slope, double-channel reflection FBGs are arranged according to the row and column spacing, the sensitive direction is consistent with the generatrix, the initial wavelengths of the double channels are collected in the benchmark working condition, and the strain and temperature sensitivity coefficients are calibrated, real-time wavelengths are obtained during the monitoring period, and the relative drift is calculated, the axial strain and temperature change of each node are obtained based on the double-channel linear coupling and decoupling, the axial displacement is converted according to the strain and the effective length of the pasting, the node coordinates and the displacement are used as the normalized weight interpolation, a two-dimensional displacement field is constructed, and the spatial derivative is solved, the abnormality is identified according to the gradient module and the uncertainty threshold, the node displacement, the abnormal coordinates and the interpolation parameters are packaged and uploaded according to the period, and a time sequence file is formed.

Description

Technical Field

[0001] This invention relates to the field of dam safety monitoring technology, specifically a dam displacement detection method based on a sensor system. Background Technology

[0002] As a core infrastructure in water conservancy and hydropower projects, the structural safety of dams has long been a major concern. Abnormal displacement or structural deformation of a dam not only directly threatens the stability of the project itself but may also trigger downstream disasters, causing significant losses to people, property, and even the ecological environment. Therefore, dam deformation monitoring and displacement field measurement technologies are crucial for ensuring the safe operation of hydraulic structures. Currently, dam displacement monitoring methods both domestically and internationally mainly include traditional point-based instrument methods, total station spatial measurement methods, and distributed or quasi-distributed fiber optic sensing, among others.

[0003] In existing technologies, the most common approach is to deploy discrete monitoring points such as joint gauges, strain gauges, pressure tubes, or inclinometers, and then manually read the data periodically or use wired acquisition systems to achieve semi-automated monitoring. While these methods can provide long-term historical data on dam deformation and local stress, they suffer from significant drawbacks, including limited point density, insufficient spatial resolution, strong locality of measurement points, difficulty in system deployment, and long data acquisition cycles. This is especially true for large-scale structures such as large-volume concrete dams and earth-rock dams, where it is difficult to achieve global high-resolution and real-time deformation field capture. With the development of spatial positioning technologies such as laser total stations and GNSS, some projects have adopted spatial coordinate monitoring, achieving sub-centimeter-level displacement measurements at the dam crest and key monitoring points. However, these methods are sensitive to the measurement environment, weather conditions, and lighting interference, and can only obtain spatial motion information for a limited number of points, making it difficult to cover deep strain and subtle deformations in critical areas such as the dam interior and the uplift slope. Furthermore, field measurements are limited by weather and working windows, failing to meet the demands of dam health monitoring for all-weather, long-term continuous, and high-frequency data. In recent years, fiber Bragg grating (FBG) sensing technology has been increasingly applied to dam health monitoring. FBG sensors offer advantages such as small size, corrosion resistance, electromagnetic interference resistance, and distributed deployment, enabling quasi-distributed multi-point, multi-parameter monitoring across multiple areas, including the dam's concrete interior, surface, and foundation. Traditional FBG strain or temperature measurements typically employ single-channel deployment, using a single wavelength drift to correspond to a single parameter change. While this method improves data real-time performance and monitoring accuracy, in real-world engineering environments, temperature changes and structural strain are often highly coupled, making it impossible for a single-channel FBG signal to directly distinguish between temperature and strain effects. To avoid measurement errors, additional temperature-compensating fibers are usually deployed simultaneously, or complex algorithm models are used to estimate temperature components, significantly increasing system cost and engineering complexity.

[0004] Therefore, this study aims to propose a dam displacement detection method based on a sensor system. By constructing a regularized local two-dimensional coordinate grid on the upstream slope of the dam and deploying dual-channel fiber Bragg sensor units, high-density, full-coverage monitoring of dam surface displacement is achieved. A combination of initial calibration and online real-time acquisition is employed. A linear coupling model is constructed using the differential characteristics of the dual-channel wavelength response to decouple temperature and strain signals with high precision, converting them into axial displacement. A continuous two-dimensional displacement field is then generated within the slope domain through normalized weighted interpolation. Based on the spatial gradient of the displacement field and the instrument measurement uncertainty, a threshold is set to automatically identify anomalies and upload the monitoring results in real time, forming historical data updated according to a time series. Summary of the Invention

[0005] This invention provides a dam displacement detection method based on a sensing system, which helps to solve the problems mentioned in the background art.

[0006] This invention provides the following technical solution: a dam displacement detection method based on a sensing system, comprising:

[0007] A local two-dimensional rectangular coordinate system is established on the upstream slope of the dam. A node grid is formed according to the preset row and column spacing. Fiber optic sensing units with dual-channel reflection characteristics are deployed at each node so that their sensitive direction is consistent with the generatrix of the upstream slope.

[0008] The initial reflection center wavelength of each node's dual channels was acquired under the reference operating conditions, and the strain sensitivity coefficient and temperature sensitivity coefficient of each channel were calibrated based on the reference strain and reference temperature conditions.

[0009] During the monitoring period, the real-time reflection center wavelength of each node's dual channels is acquired, and the corresponding wavelength change is calculated relative to the initial value.

[0010] Based on the combination of a dual-channel linear coupling model and calibrated sensitivity coefficients, the axial strain and temperature change of each node during the monitoring period are decoupled and obtained.

[0011] The axial displacement of each node is calculated based on the axial strain of each node and the effective measurement length of the adhesive substrate.

[0012] Using node coordinates and corresponding displacements as input, a continuous two-dimensional displacement field is constructed in the domain of the uphill slope using normalized weighted interpolation, and the spatial derivative of the displacement field with respect to the coordinates is obtained.

[0013] Based on the gradient magnitude of the displacement field and the allowable threshold determined by the measurement uncertainty and grid spacing, out-of-limit locations are identified as a set of anomalies.

[0014] According to the set time interval, the node displacement, abnormal point coordinates and displacement field interpolation parameters corresponding to each monitoring time are packaged, uploaded and updated in chronological order to update the historical monitoring sequence.

[0015] Optionally, the step of establishing a local two-dimensional rectangular coordinate system on the upstream slope of the dam, forming a node grid according to a preset row and column spacing, and deploying fiber optic sensing units with dual-channel reflection characteristics at each node, ensuring that their sensing direction is consistent with the generatrix of the upstream slope, specifically includes:

[0016] Set the origin of the coordinate system at any end of the intersection of the water-facing surface and the dam foundation. The first axis is consistent with the dam axis and is set along the dam bottom direction. The second axis is set upward along the water-facing slope.

[0017] The node array is determined according to the preset node spacing in the first and second axis directions; the total number of nodes is determined according to the unfolded length of the dam body along the first axis and the unfolded height along the second axis.

[0018] Dual-wavelength fiber Bragg grating sensors are attached at each node location, the sensing stack is attached along the second axis, and the reflection center wavelengths of the two channels are recorded. Monitoring is performed continuously over time.

[0019] Optionally, the step of acquiring the initial reflection center wavelength of each node's dual channels under reference operating conditions, and calibrating the strain sensitivity coefficient and temperature sensitivity coefficient of each channel based on reference strain and reference temperature conditions, specifically includes:

[0020] The initial reflection center wavelength of each node's two channels is acquired at the reference time;

[0021] The strain sensitivity coefficient and temperature sensitivity coefficient of the two channels were obtained through calibration experiments. The strain sensitivity coefficient was obtained by applying a reference strain and calculating the wavelength response, and the temperature sensitivity coefficient was obtained by applying a reference temperature rise and calculating the wavelength response.

[0022] The sensitivity coefficients for each node and each channel are stored in the form of a parameter table.

[0023] Optionally, the step of acquiring the real-time reflection center wavelength of each node's dual channels during the monitoring period and calculating the corresponding wavelength change relative to the initial value specifically includes:

[0024] At any given monitoring time, the real-time reflection center wavelength of each node's two channels is acquired;

[0025] The wavelength change of each node and each channel is calculated based on the initial reflection center wavelength.

[0026] Optionally, the method of decoupling and obtaining the axial strain and temperature change of each node during the monitoring period based on the combination of the dual-channel linear coupling model and the calibrated sensitivity coefficients specifically includes:

[0027] Using the wavelength changes of the two channels as the observed quantities and the axial strain and relative temperature changes as the quantities to be determined, a linear equation system is established by combining the strain and temperature sensitivity coefficients of the two channels.

[0028] Under the condition that the discriminant is non-zero, the solution is performed on each node to obtain the true axial strain and relative temperature change of each node at that moment.

[0029] Optionally, the axial displacement of each node is calculated based on the axial strain of each node and the effective measurement length of the adhesive substrate, specifically including:

[0030] Obtain the effective length of the fiber optic sensor's bonding substrate;

[0031] Based on the linear relationship between axial strain and displacement, the axial displacement of each node at that moment is calculated.

[0032] Optionally, the step of constructing a continuous two-dimensional displacement field within the domain of the upstream slope using normalized weighted interpolation with nodal coordinates and corresponding displacements as input, and obtaining the spatial derivative of the displacement field with respect to coordinates, specifically includes:

[0033] Establish a discrete set of points consisting of node coordinates and node displacements;

[0034] The node influence weights are defined according to the node spacing in the first and second axis directions, and weighted interpolation is performed on any coordinate point in the inclined plane region to obtain the continuous displacement field.

[0035] Based on the relationship between the weights and the coordinates, the partial derivatives of the displacement field with respect to the first and second axes are calculated.

[0036] Optionally, the step of identifying out-of-range locations as a set of anomalies based on the gradient magnitude of the displacement field and an allowable threshold determined by measurement uncertainty and grid spacing specifically includes:

[0037] Calculate the gradient and gradient magnitude of the displacement field;

[0038] The allowable gradient upper limit is determined based on the resolution of the two-channel instrument and its converted standard uncertainty, node spacing, and the allowable maximum displacement difference between adjacent nodes.

[0039] When the gradient magnitude of a certain coordinate point exceeds the upper limit of the gradient, it is included in the set of abnormal point coordinates as an outlier.

[0040] Optionally, the step of packaging the node displacements, anomaly point coordinates, and displacement field interpolation parameters corresponding to each monitoring time point according to a set time interval, uploading them, and updating the historical monitoring sequence in chronological order specifically includes:

[0041] A recording unit is formed for each monitoring moment, and the recording unit includes a discrete point set, an anomaly point set and displacement field interpolation coefficients for that moment;

[0042] Every 10 seconds, the recording units are packaged into data packets and uploaded to the database. The central node receives and updates the entire dam monitoring historical record sequence in chronological order.

[0043] The present invention has the following beneficial effects:

[0044] 1. A local two-dimensional rectangular coordinate system is established on the upstream slope of the dam. Based on the actual structure and monitoring requirements, a regular node grid and orderly deployment of dual-channel fiber optic sensing units are designed. Compared with the traditional method of using only one-dimensional or scattered points, this deployment method can not only achieve spatial integration of multi-point information on the structural surface, ensuring the integrity and comparability of monitoring data, but also effectively improve the response capability of the upstream slope along the main deformation direction through optimized arrangement of sensitive directions. This method enhances the spatial resolution of the displacement field, laying a solid foundation for subsequent continuous field interpolation and overall health assessment. The combination of gridded point layout, direction sensitivity, and coordinate system helps to form a multi-dimensional, multi-scale monitoring mode for complex structures, adapting to the digital monitoring requirements of large-scale hydraulic structures such as dams.

[0045] 2. Under baseline operating conditions, the initial reflection center wavelengths of multiple channels at each node were acquired, and the strain and temperature sensitivity coefficients were obtained through calibration experiments, achieving full-process calibration of the fiber optic sensor's fundamental parameters. Compared with existing technologies, this scheme employs dual-channel parallel acquisition and independent calibration, improving the accuracy and robustness of subsequent strain and temperature decoupling. Through systematic parameter calibration, the discrete errors of different sensing units, environmental disturbances, and uncertainties in the initial state can be eliminated, ensuring the consistency and traceability of the overall system data. This provides a theoretical basis for accurately identifying subtle structural deformations and separating temperature interference, laying the foundation for high-quality displacement calculation and anomaly detection.

[0046] 3. This scheme proposes to continuously acquire the real-time reflection center wavelength of each node's dual channels during actual monitoring, and dynamically calculate the wavelength change based on the initial state. This process improves the continuity and temporal resolution of data acquisition, overcoming the previous problem of relying on manual periodic sampling and being unable to dynamically reflect the deformation process. Simultaneously, all changes are standardized and quantified at the node level, facilitating subsequent batch calculations and systematic analysis. Through this innovative approach, the transient and cumulative deformation of the dam body under various loads and environmental changes can be captured, effectively supporting full-cycle monitoring of structural health.

[0047] 4. To address the complexity of fiber optic sensing signals being simultaneously affected by temperature and strain, the proposed solution introduces a dual-channel linear coupling model. Using calibrated sensitivity coefficients, a system of linear equations is established to decouple strain and temperature changes in real time. Compared to traditional single-channel methods relying on compensation plates or external temperature sensors, this approach enables independent and synchronous decoupling of each monitoring node, eliminating errors caused by uneven temperature fields in different regions and improving the spatial consistency and physical reliability of the calculations. This innovation enhances the accuracy of extracting the true displacement of the dam surface, ensuring the scientific validity and practical usability of deformation monitoring results, which is particularly crucial for monitoring major engineering structures with high safety requirements.

[0048] 5. By linearly converting nodal strain to the effective length of the bonded substrate, the scheme achieves a precise conversion of minute structural deformations into absolute axial displacement. This method not only facilitates the physical interpretation of monitoring results but also enables the system to generate a dynamic displacement history sequence for each node by combining monitoring timelines. Compared with existing point-based monitoring methods, this method can effectively quantify and track the cumulative deformation and sudden events of the dam body at different time scales, possessing extremely high practical value for engineering safety assessment and risk early warning. Furthermore, it fully considers the physical characteristics and response sections of the actual installation of the sensing units, improving the accuracy of the calculation results and their adaptability to the field.

[0049] 6. This invention innovatively integrates nodal displacement with spatial coordinates, constructing a continuous two-dimensional displacement field on the dam surface through normalized weighted interpolation technology. This method overcomes the limitations of traditional point- and line-based monitoring, which can only obtain discrete information, and realizes a digital expression of the continuous deformation distribution on the structural surface. Through further calculation of spatial derivatives, the deformation gradient and deformation trend in different regions can be systematically analyzed, providing scientific support for early identification, zonal analysis, and refined operation and maintenance of abnormal dam deformation. Compared with existing technologies, this method expands the spatial utilization efficiency and expressive power of monitoring information, and can more comprehensively reflect the overall and local health status of the structure.

[0050] 7. Based on a two-dimensional displacement field, this scheme scientifically sets anomaly detection threshold by combining measurement uncertainty and node grid spacing, and automatically identifies out-of-limit points by analyzing the displacement gradient modulus field. This innovative method incorporates statistical analysis and engineering safety boundaries, avoiding missed reports of major anomalies and effectively suppressing false alarms, thus improving the reliability and intelligence level of anomaly detection. Compared with existing manual inspection or subjective threshold judgment modes, this method has the advantages of high automation, strong objectivity, and rapid response, making it an indispensable key link in realizing intelligent structural health monitoring and risk early warning.

[0051] 8. This solution concludes by periodically packaging monitoring data, including node displacements, anomaly coordinates, and displacement field parameters, and uploading them to a central database to form a complete historical monitoring sequence stored in chronological order. This end-to-end data archiving and management model effectively ensures the integrity, security, and traceability of monitoring data, facilitating subsequent data mining, trend analysis, and risk prediction. Compared to the traditional model of scattered monitoring data and manual aggregation, this method boasts a high degree of informatization and integration, supporting the construction of intelligent dam operation and maintenance and decision support systems, and providing a solid data foundation for the lifecycle health management of large-scale engineering structures. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the process of the present invention.

[0053] Figure 2 This is a schematic diagram of the coordinate system of the present invention.

[0054] In the diagram: 1 - origin of coordinates, 2 - first axis, 3 - second axis, 4 - water-facing surface. Detailed Implementation

[0055] 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, and 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.

[0056] Example, refer to Figure 1 A method for dam displacement detection based on a sensor system, comprising:

[0057] A local two-dimensional rectangular coordinate system is established on the upstream slope of the dam. A node grid is formed according to the preset row and column spacing. Fiber optic sensing units with dual-channel reflection characteristics are deployed at each node so that their sensitive direction is consistent with the generatrix of the upstream slope.

[0058] The initial reflection center wavelength of each node's dual channels was acquired under the reference operating conditions, and the strain sensitivity coefficient and temperature sensitivity coefficient of each channel were calibrated based on the reference strain and reference temperature conditions.

[0059] During the monitoring period, the real-time reflection center wavelength of each node's dual channels is acquired, and the corresponding wavelength change is calculated relative to the initial value.

[0060] Based on the combination of a dual-channel linear coupling model and calibrated sensitivity coefficients, the axial strain and temperature change of each node during the monitoring period are decoupled and obtained.

[0061] The axial displacement of each node is calculated based on the axial strain of each node and the effective measurement length of the adhesive substrate.

[0062] Using node coordinates and corresponding displacements as input, a continuous two-dimensional displacement field is constructed in the domain of the uphill slope using normalized weighted interpolation, and the spatial derivative of the displacement field with respect to the coordinates is obtained.

[0063] Based on the gradient magnitude of the displacement field and the allowable threshold determined by the measurement uncertainty and grid spacing, out-of-limit locations are identified as a set of anomalies.

[0064] According to the set time interval, the node displacement, abnormal point coordinates and displacement field interpolation parameters corresponding to each monitoring time are packaged, uploaded and updated in chronological order to update the historical monitoring sequence.

[0065] A sensor-based dam displacement detection method is proposed, encompassing the entire process from sensor deployment, initial parameter calibration, real-time data acquisition and processing, two-dimensional spatial displacement field construction, anomaly detection, and data archiving and uploading. By scientifically establishing a two-dimensional rectangular coordinate system on the dam's upstream slope and regularly deploying fiber optic sensing units with dual-channel reflection characteristics, structural deformation can be monitored in real-time in a planar, multi-point, and high-density manner. Subsequently, sensitivity coefficient calibration and initial wavelength acquisition under baseline conditions effectively solve the data inaccuracies caused by sensor discreteness, inconsistent sensitivity, and environmental disturbances in traditional monitoring. During the monitoring phase, by acquiring node wavelength changes in real time and performing temperature-strain decoupling based on a dual-channel model, the limitations of previous methods, such as difficult temperature compensation and inaccurate data separation, are overcome. Furthermore, displacement is calculated by converting node strain and bonding length, and a continuous two-dimensional displacement field is reconstructed using a normalized weighted interpolation algorithm, enabling precise understanding of the overall deformation characteristics and spatial distribution of the dam's upstream slope. Ultimately, anomaly detection was achieved based on the spatial derivative of the displacement field and the uncertainty threshold, enabling intelligent identification and early warning of abnormal deformation of the dam body. All data was automatically archived and uploaded, ensuring information continuity and traceability. Through multi-stage innovation, the overall solution systematically improved the accuracy, spatial integrity, and automation level of dam displacement monitoring, outperforming traditional single-point, manual, or passive monitoring methods, and providing reliable technical support for dam structural health assessment and safe operation and maintenance.

[0066] Reference Figure 2 The process involves establishing a local two-dimensional rectangular coordinate system on the upstream slope of the dam, forming a node grid according to a preset row and column spacing, and deploying fiber optic sensing units with dual-channel reflection characteristics at each node, ensuring that their sensing direction is consistent with the generatrix of the upstream slope. Specifically, this includes:

[0067] Set the origin of the coordinate system at any end of the intersection of the water-facing surface and the dam foundation. The first axis is consistent with the dam axis and is set along the dam bottom direction. The second axis is set upward along the water-facing slope.

[0068] The node array is determined according to the preset node spacing in the first and second axis directions; the total number of nodes is determined according to the unfolded length of the dam body along the first axis and the unfolded height along the second axis.

[0069] Dual-wavelength fiber Bragg grating sensors are attached at each node location, the sensing stack is attached along the second axis, and the reflection center wavelengths of the two channels are recorded. Monitoring is performed continuously over time.

[0070] The specific steps are as follows:

[0071] Constructing a local two-dimensional rectangular coordinate system on the upstream slope of the dam ,origin Choose any endpoint of the intersection line between the upstream face and the dam foundation. The axis is aligned with the dam axis along the dam base. The axis runs along the upward-facing generatrix of the inclined plane.

[0072] Set node edge The spacing is ,along The spacing is ;

[0073] Obtain the dam body along The unfolded length is denoted as Obtain the dam body along The unfolded height is denoted as The total number of nodes is ;

[0074] The first The coordinates of each node are as follows ;in, For node indexing;

[0075] At the node location on the water-facing slope Attaching a dual-wavelength fiber Bragg grating, its sensitive stack edge Axial bonding;

[0076] Record the center wavelengths of the two reflections: first channel wavelength Second channel wavelength ;in, The independent variable is continuous time.

[0077] By establishing a local two-dimensional rectangular coordinate system on the upstream slope of the dam, and scientifically defining the origin and axes using the structural principal axes and generatrices, along with preset row and column spacing, the distribution and total number of node grids are strictly determined. Fiber optic sensing units are then attached and deployed along the main deformation direction of the structure. Compared to existing technologies that often employ disordered point placement or only one-dimensional acquisition methods, this method improves the spatial resolution and data comparability of the monitoring system. The coordinates of each node, the sensor installation direction, and the spacing between points all have clear physical meanings, providing a more solid foundation for spatial interpolation and field reconstruction in subsequent data analysis. Furthermore, the parallel deployment of dual-channel sensing units allows for the simultaneous acquisition of multiple channels of signals at the same location, facilitating subsequent temperature-strain decoupling and anomaly identification. This increases the coverage and density of monitoring on the upstream slope of the dam, ensuring the completeness and detail of deformation information; it also guarantees the spatial correspondence between monitoring data and the structural body, improving the reliability and engineering applicability of the monitoring results.

[0078] The process of acquiring the initial reflection center wavelength of each node's dual channels under reference operating conditions, and calibrating the strain sensitivity coefficient and temperature sensitivity coefficient of each channel based on reference strain and reference temperature conditions, specifically includes:

[0079] The initial reflection center wavelength of each node's two channels is acquired at the reference time;

[0080] The strain sensitivity coefficient and temperature sensitivity coefficient of the two channels were obtained through calibration experiments. The strain sensitivity coefficient was obtained by applying a reference strain and calculating the wavelength response, and the temperature sensitivity coefficient was obtained by applying a reference temperature rise and calculating the wavelength response.

[0081] The sensitivity coefficients for each node and each channel are stored in the form of a parameter table.

[0082] The specific steps are as follows:

[0083] Reference time under no-load, uniform temperature conditions , collect the first Initial wavelength of the node:

[0084] , ;in, For the first The initial reflection center wavelength of the node in the first channel; For the first The initial reflection center wavelength of the node in the second channel;

[0085] The strain and temperature sensitivity coefficients corresponding to the first and second channel wavelengths were obtained using calibration experiments.

[0086] For each channel, execute steps S201 to S202 sequentially:

[0087] S201. Calculate the strain sensitivity coefficient: ;in, The reference strain applied; For reference temperature; For channel indexing; For the first The strain sensitivity coefficient of the channel; Strain, unit is microstrain. ;

[0088] S202, Calculate the temperature sensitivity coefficient: ;in, For reference temperature rise; For the first Temperature sensitivity coefficient of the channel.

[0089] By acquiring the initial reflection center wavelength of the dual-channel sensing unit at each monitoring node under no-load, uniform temperature reference conditions, and determining the strain sensitivity coefficient and temperature sensitivity coefficient through calibration experiments, this invention, unlike some existing technologies that use theoretical parameters or single calibrations, performs independent and systematic experimental calibration for each channel, ensuring the accuracy and relevance of subsequent data calculations. The parameters of each node and each channel are meticulously recorded, eliminating sensitivity inconsistencies caused by differences in sensor batches, installation conditions, and environmental interference. Furthermore, calibration using actual physical quantities (such as known strain and temperature rise) enhances the physical basis and reliability of the method. This strengthens the parameter uniformity and signal response consistency of the entire system, providing a solid foundation for subsequent high-precision temperature-strain separation and displacement conversion, while avoiding reference drift caused by environmental disturbances or long-term operation, thus improving the reliability and maintainability of long-term monitoring.

[0090] The process of acquiring the real-time reflection center wavelength of each node's dual channels during the monitoring period and calculating the corresponding wavelength change relative to the initial value specifically includes:

[0091] At any given monitoring time, the real-time reflection center wavelength of each node's two channels is acquired;

[0092] The wavelength change of each node and each channel is calculated based on the initial reflection center wavelength.

[0093] The specific steps are as follows:

[0094] During monitoring time , obtain the Real-time wavelength values ​​at nodes: , ;in, For the first Node, First Channel at time The reflection center wavelength;

[0095] Calculate wavelength drift: , ;in, For the first Node, First Channel at time The amount of change in wavelength relative to the initial value.

[0096] During monitoring, the reflection center wavelength of each node's dual channels is collected periodically, and the wavelength drift is accurately calculated using the initial wavelength under baseline conditions as a reference. Unlike existing technologies that only collect absolute values ​​or perform intermittent sampling, this method continuously and periodically monitors signal changes and uses a unique baseline for each node as a standard to achieve personalized and refined data normalization. This effectively eliminates systematic offsets caused by system aging, environmental fluctuations, or changes in sensor status, ensuring that all subsequent parameters such as strain and displacement are based on the same reference standard, thus improving the dynamic response capability and spatiotemporal comparability of the monitoring data. Furthermore, the refined management of wavelength changes provides data assurance for subsequent temperature-strain decoupling and anomaly identification, laying a solid foundation for long-term stable monitoring of the dam and detection of emergencies.

[0097] The method, based on a dual-channel linear coupling model and calibrated sensitivity coefficients, decouples and calculates the axial strain and temperature changes of each node during the monitoring period. Specifically, this includes:

[0098] Using the wavelength changes of the two channels as the observed quantities and the axial strain and relative temperature changes as the quantities to be determined, a linear equation system is established by combining the strain and temperature sensitivity coefficients of the two channels.

[0099] Under the condition that the discriminant is non-zero, the solution is performed on each node to obtain the true axial strain and relative temperature change of each node at that moment.

[0100] The specific steps are as follows:

[0101] Get the Node at time The true axial strain is denoted as , No. Node at time The change in relative initial temperature is denoted as ;

[0102] Establish a linear coupling relationship: ;

[0103] remember ;in, for The system's discriminant;

[0104] From the calibration value, we can obtain Therefore, this system of equations can be solved to obtain:

[0105] , .

[0106] This invention innovates in the decoupling of temperature and strain, proposing a solution method based on a dual-channel linear coupling model and a combination of calibrated sensitivity coefficients. The wavelength change at each node is influenced by both strain and temperature; traditional methods struggle to accurately separate these two influences. This method utilizes the different sensitivities of the two channels to establish a system of linear equations, independently solving for the true axial strain and temperature change at each node at every moment. This method overcomes the limitations of traditional temperature compensation plates or external temperature field corrections, achieving high-precision, real-time separation of physical quantities. In situations where the structure is heated, environmental fluctuations are frequent, or local temperature rises are significant, it ensures that the obtained strain (i.e., the true structural deformation) is not confused by temperature changes, thereby improving the accuracy and reliability of displacement monitoring. For complex and coupled environmental conditions in actual engineering operations, this invention enhances the engineering applicability of monitoring data and the scientific rigor of subsequent decision support.

[0107] The axial displacement of each node is calculated based on the axial strain of each node and the effective measurement length of the bonding substrate, specifically including:

[0108] Obtain the effective length of the fiber optic sensor's bonding substrate;

[0109] Based on the linear relationship between axial strain and displacement, the axial displacement of each node at that moment is calculated.

[0110] The specific steps are as follows:

[0111] The effective length of the fiber optic bonding substrate is denoted as . Then the first The nodal displacement is:

[0112] ;in, For the first Node at time axial displacement.

[0113] By obtaining the effective measurement length of the fiber optic bonding substrate, the axial strain monitored at the nodes is converted into absolute displacement. This step fully considers the actual installation location of the sensor and the length characteristics of the stress area, giving the conversion results direct physical meaning. Compared to some existing methods that only provide strain or ununitized data, this solution directly outputs the actual displacement of the structural deformation, making it easier for engineers to understand and use. It retains the advantages of high-precision strain monitoring while converting the results into a more intuitive and engineering-significant displacement quantity, facilitating timely decision-making by maintenance personnel. Especially in long-term series and spatial distribution analysis, this physical quantity conversion improves the usability and scientific rigor of the data.

[0114] The process involves using node coordinates and corresponding displacements as inputs, employing normalized weighted interpolation to construct a continuous two-dimensional displacement field within the domain of the upstream slope, and obtaining the spatial derivative of the displacement field with respect to the coordinates. Specifically, this includes:

[0115] Establish a discrete set of points consisting of node coordinates and node displacements;

[0116] The node influence weights are defined according to the node spacing in the first and second axis directions, and weighted interpolation is performed on any coordinate point in the inclined plane region to obtain the continuous displacement field.

[0117] Based on the relationship between the weights and the coordinates, the partial derivatives of the displacement field with respect to the first and second axes are calculated.

[0118] The specific steps are as follows:

[0119] Constructing a discrete displacement field point set: ;in, For a moment Coordinate-displacement point set, elements are triples ;

[0120] Construct the original weight function: ;in, For the first For the first The original weights of the node's influence;

[0121] Calculate the normalized weights:

[0122] , ;in, This is the sum of the original weights of all nodes; For summation index; For the first Normalized weights of nodes;

[0123] Only At that time, construct the displacement field interpolation function:

[0124] ;in, coordinates of the inclined plane At any moment The continuous displacement field value; For the domain, satisfying sloping area;

[0125] Denote the symbolic function ;

[0126] exist Within the area, there are:

[0127] , ;

[0128] exist Within the area, , ;

[0129] set up , ,but:

[0130] , ;in, , The sum of the partial derivatives of the normalization factor;

[0131] The partial derivatives of the displacement field are:

[0132] , ;in, , This is the partial derivative of the displacement field with respect to the coordinates.

[0133] A method is proposed to construct a continuous two-dimensional displacement field using nodal coordinates and displacements, and to calculate its spatial derivative. By using normalized weighted interpolation, the displacement information of discrete nodes is weighted and distributed across the entire upstream sloping surface, realizing a digital representation of the continuous deformation field on the structural surface. Compared to traditional methods that only obtain discrete point information or one-dimensional profile changes, this approach comprehensively reflects the deformation distribution and spatial trends of the entire dam surface. The further introduction of the spatial derivative reveals in detail the deformation gradient, local anomalies, or overall deformation patterns of the dam in various regions. This enhances the monitoring system's overall control over the structural health status, facilitates early identification of potential hazardous areas and analysis of deformation mechanisms, and provides a solid technical foundation for subsequent risk warning and engineering decision-making.

[0134] The step of identifying out-of-limit locations as a set of anomalies based on the gradient magnitude of the displacement field and the allowable threshold determined by measurement uncertainty and grid spacing specifically includes:

[0135] Calculate the gradient and gradient magnitude of the displacement field;

[0136] The allowable gradient upper limit is determined based on the resolution of the two-channel instrument and its converted standard uncertainty, node spacing, and the allowable maximum displacement difference between adjacent nodes.

[0137] When the gradient magnitude of a certain coordinate point exceeds the upper limit of the gradient, it is included in the set of abnormal point coordinates as an outlier.

[0138] The specific steps are as follows:

[0139] Calculate the gradient of the displacement field ;

[0140] Calculate gradient mode field ;

[0141] structure Compared to Linear form:

[0142] ;in, , ;

[0143] Let the standard uncertainties of the two-channel wavelength measurement be respectively and And independent;

[0144] The minimum resolution of the two channels of the instrument is obtained separately and denoted as follows: , Converted to uniform distribution: ;

[0145] Then the standard uncertainty of strain standard uncertainty of displacement ;

[0146] Take 3 The standard sets the maximum allowable displacement difference between adjacent nodes. ;

[0147] The upper limit of the allowed gradient is then... ;

[0148] like a certain point inside satisfy If it is an outlier, it will be marked as an anomaly and included in the list. ; For a moment The set of outlier coordinates, where each element is a tuple of slope coordinates that satisfies the threshold condition.

[0149] A method is proposed to automatically identify out-of-limit locations as anomalies by setting thresholds based on parameters such as displacement field gradient magnitude, measurement uncertainty, and grid spacing. By calculating the spatial derivative of the displacement field at each monitoring moment and combining the accuracy of the sensing system with engineering experience to set reasonable thresholds, real-time and intelligent identification of structural deformation anomalies is achieved. Unlike traditional manual threshold setting or judgment based solely on experience, this method scientifically integrates statistical analysis and engineering practice, effectively avoiding false alarms and missed alarms. It can promptly detect and locate problem areas when abnormal deformation has just occurred or has not yet significantly expanded, saving valuable time for engineering operation and maintenance. This automated anomaly identification method is particularly suitable for large-scale structures such as dams, reducing the burden and risk of manual inspection.

[0150] The process involves packaging the node displacements, anomaly point coordinates, and displacement field interpolation parameters corresponding to each monitoring time point according to a set time interval, uploading them, and updating the historical monitoring sequence in chronological order. Specifically, this includes:

[0151] A recording unit is formed for each monitoring moment, and the recording unit includes a discrete point set, an anomaly point set and displacement field interpolation coefficients for that moment;

[0152] Every 10 seconds, the recording units are packaged into data packets and uploaded to the database. The central node receives and updates the entire dam monitoring historical record sequence in chronological order.

[0153] The specific steps are as follows:

[0154] For each moment, a recording unit is formed: ;in, For the first The moment of each package upload; Upload batch number; For included time The set of points, anomalies, and displacement field functions;

[0155] Each interval Send data packets to the database The data packet contains the displacement of all nodes, the coordinates of outliers, and the interpolation coefficients in the two-dimensional function expression.

[0156] After receiving the data, the central node updates the dam monitoring timeline in chronological order: ;in, It is a historical sequence; This represents the number of data packets that have been recorded.

[0157] This method involves archiving, uploading, and managing historical monitoring sequences of monitoring results. At each monitoring moment, node displacements, anomaly coordinates, and displacement field interpolation parameters are packaged and uploaded, with the central node managing historical records chronologically. Unlike existing decentralized records or manual archiving, this method achieves real-time, standardized, and automated data aggregation and storage, ensuring the integrity and traceability of monitoring information. It provides systematic data support for subsequent data mining, trend analysis, deformation evolution studies, and structural safety assessments. Simultaneously, this management mechanism greatly facilitates cross-cycle and cross-event comparative analysis, enabling health management throughout the entire project lifecycle and enhancing the modernization level of intelligent monitoring and operation and maintenance of dam structures.

[0158] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0159] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for dam displacement detection based on a sensor system, characterized in that, include: A local two-dimensional rectangular coordinate system is established on the upstream slope of the dam. A node grid is formed according to the preset row and column spacing. Fiber optic sensing units with dual-channel reflection characteristics are deployed at each node so that their sensitive direction is consistent with the generatrix of the upstream slope. The initial reflection center wavelength of each node's dual channels was acquired under the reference operating conditions, and the strain sensitivity coefficient and temperature sensitivity coefficient of each channel were calibrated based on the reference strain and reference temperature conditions. During the monitoring period, the real-time reflection center wavelength of each node's dual channels is acquired, and the corresponding wavelength change is calculated relative to the initial value. Based on the combination of a dual-channel linear coupling model and calibrated sensitivity coefficients, the axial strain and temperature change of each node during the monitoring period are decoupled and obtained. The axial displacement of each node is calculated based on the axial strain of each node and the effective measurement length of the adhesive substrate. Using node coordinates and corresponding displacements as input, a continuous two-dimensional displacement field is constructed in the domain of the uphill slope using normalized weighted interpolation, and the spatial derivative of the displacement field with respect to the coordinates is obtained. Based on the gradient magnitude of the displacement field and the allowable threshold determined by the measurement uncertainty and grid spacing, out-of-limit locations are identified as a set of anomalies. According to the set time interval, the node displacement, abnormal point coordinates and displacement field interpolation parameters corresponding to each monitoring time are packaged, uploaded and updated in chronological order to update the historical monitoring sequence.

2. The dam displacement detection method based on a sensor system according to claim 1, characterized in that, The process involves establishing a local two-dimensional rectangular coordinate system on the upstream slope of the dam, forming a node grid according to a preset row and column spacing, and deploying fiber optic sensing units with dual-channel reflection characteristics at each node, ensuring that their sensing direction is consistent with the generatrix of the upstream slope. Specifically, this includes: Set the origin of the coordinate system at any end of the intersection of the water-facing surface and the dam foundation. The first axis is consistent with the dam axis and is set along the dam bottom direction. The second axis is set upward along the water-facing slope. The node array is determined according to the preset node spacing in the first and second axis directions; the total number of nodes is determined according to the unfolded length of the dam body along the first axis and the unfolded height along the second axis. Dual-wavelength fiber Bragg grating sensors are attached at each node location, the sensing stack is attached along the second axis, and the reflection center wavelengths of the two channels are recorded. Monitoring is performed continuously over time.

3. The dam displacement detection method based on a sensor system according to claim 2, characterized in that, The process of acquiring the initial reflection center wavelength of each node's dual channels under reference operating conditions, and calibrating the strain sensitivity coefficient and temperature sensitivity coefficient of each channel based on reference strain and reference temperature conditions, specifically includes: The initial reflection center wavelength of each node's two channels is acquired at the reference time; The strain sensitivity coefficient and temperature sensitivity coefficient of the two channels were obtained through calibration experiments. The strain sensitivity coefficient was obtained by applying a reference strain and calculating the wavelength response, and the temperature sensitivity coefficient was obtained by applying a reference temperature rise and calculating the wavelength response. The sensitivity coefficients for each node and each channel are stored in the form of a parameter table.

4. The dam displacement detection method based on a sensor system according to claim 3, characterized in that, The process of acquiring the real-time reflection center wavelength of each node's dual channels during the monitoring period and calculating the corresponding wavelength change relative to the initial value specifically includes: At any given monitoring time, the real-time reflection center wavelength of each node's two channels is acquired; The wavelength change of each node and each channel is calculated based on the initial reflection center wavelength.

5. The dam displacement detection method based on a sensor system according to claim 4, characterized in that, The method, based on a dual-channel linear coupling model and calibrated sensitivity coefficients, decouples and calculates the axial strain and temperature changes of each node during the monitoring period. Specifically, this includes: Using the wavelength changes of the two channels as the observed quantities and the axial strain and relative temperature changes as the quantities to be determined, a linear equation system is established by combining the strain and temperature sensitivity coefficients of the two channels. Under the condition that the discriminant is non-zero, the solution is performed on each node to obtain the true axial strain and relative temperature change of each node at that moment.

6. The dam displacement detection method based on a sensor system according to claim 5, characterized in that, The axial displacement of each node is calculated based on the axial strain of each node and the effective measurement length of the bonding substrate, specifically including: Obtain the effective measurement length of the fiber optic sensor bonding substrate; Based on the linear relationship between axial strain and displacement, the axial displacement of each node at that moment is calculated.

7. The dam displacement detection method based on a sensor system according to claim 6, characterized in that, The process involves using node coordinates and corresponding displacements as inputs, employing normalized weighted interpolation to construct a continuous two-dimensional displacement field within the domain of the upstream slope, and obtaining the spatial derivative of the displacement field with respect to the coordinates. Specifically, this includes: Establish a discrete set of points consisting of node coordinates and node displacements; The node influence weights are defined according to the node spacing in the first and second axis directions, and weighted interpolation is performed on any coordinate point in the inclined plane region to obtain the continuous displacement field. Based on the relationship between the weights and the coordinates, the partial derivatives of the displacement field with respect to the first and second axes are calculated.

8. The dam displacement detection method based on a sensor system according to claim 7, characterized in that, The step of identifying out-of-limit locations as a set of anomalies based on the gradient magnitude of the displacement field and the allowable threshold determined by measurement uncertainty and grid spacing specifically includes: Calculate the gradient and gradient magnitude of the displacement field; The allowable gradient upper limit is determined based on the resolution of the two-channel instrument and its converted standard uncertainty, node spacing, and the allowable maximum displacement difference between adjacent nodes. When the gradient magnitude of a certain coordinate point exceeds the upper limit of the gradient, it is included in the set of abnormal point coordinates as an outlier.

9. A dam displacement detection method based on a sensor system according to claim 8, characterized in that, The process involves packaging the node displacements, anomaly point coordinates, and displacement field interpolation parameters corresponding to each monitoring time point according to a set time interval, uploading them, and updating the historical monitoring sequence in chronological order. Specifically, this includes: A recording unit is formed for each monitoring moment, and the recording unit includes a discrete point set, an anomaly point set and displacement field interpolation parameters for that moment; Every 10 seconds, the recording units are packaged into data packets and uploaded to the database. The central node receives and updates the entire dam monitoring historical record sequence in chronological order.

Images