Ground penetrating radar based high risk hazard area detection method

Abstract

The application discloses a high-risk hidden danger area detection method based on a ground penetrating radar, and relates to the technical field of wireless detection.The method comprises the following steps: obtaining radar detection data along a target dam body measuring line; calculating water content change data at different measuring points based on the energy intensity distribution of different measuring points in the radar detection data; calculating a displacement cost matrix based on the displacement degree between different time instants of a water content change tracer point corresponding to the water content change data at each measuring point; determining a cooperative displacement cost matrix of each measuring point based on the displacement difference between parallel paths formed by different element values in the displacement cost matrix; determining a seepage channelization index based on the displacement cost matrix and the cooperative displacement cost matrix; and performing risk division processing on the area where each measuring point is located based on the displacement cost matrix corresponding index and the seepage channelization index.The application achieves the technical effect of accurately detecting high-risk areas in a dam when hidden dangers occur early.

Description

Technical Field

[0001] This application relates to the field of wireless detection technology, specifically to a method for detecting high-risk hidden danger areas based on ground-penetrating radar. Background Technology

[0002] Ground-penetrating radar (GPR) is a non-destructive testing technology that uses high-frequency electromagnetic waves to detect information about underground media. It is applied to the safety monitoring of structures such as earth-rock dams and embankments. Soil media with different moisture contents have different dielectric constants, allowing GPR to detect the interface between dry and saturated soil, i.e., the phreatic line. The location, shape, and dynamic response of the phreatic line to changes in reservoir water level are indicators for assessing the seepage stability of the dam.

[0003] Existing methods mostly rely on manual interpretation of single static profiles of the seepage line or simple threshold segmentation, focusing only on signal strength or displacement. They focus on the "position" of the seepage line while ignoring its "motion pattern." Therefore, they cannot distinguish between benign diffuse response of water or seepage in the dam and malignant piping precursors. Consequently, they cannot accurately detect high-risk areas in the dam when hidden dangers appear in the early stages, limiting the application of ground penetrating radar in the early diagnosis of seepage damage hazards. Summary of the Invention

[0004] To address the technical problem that related technologies, which only focus on signal strength or displacement, cannot accurately detect high-risk areas in dams when potential hazards appear in the early stages, this application provides a method for detecting high-risk hazard areas based on ground-penetrating radar.

[0005] The specific technical solution adopted is as follows: Acquire radar detection data within a preset time period along the survey line of the target dam body; Based on the energy intensity distribution at different measuring points in the radar detection data, the water content change data at different measuring points were calculated. Based on the displacement of the moisture change tracer points at different times corresponding to the moisture change data at each measuring point, the displacement cost matrix is ​​calculated; based on the displacement difference between the parallel paths formed by different element values ​​in the displacement cost matrix, the collaborative displacement cost matrix of each measuring point is determined. Based on the displacement cost matrix and the cooperative displacement cost matrix, the seepage channelization index is determined. The seepage channelization index is used to characterize the degree of cooperative motion of the water flow corresponding to each measuring point. Based on the displacement cost matrix corresponding indicators and the seepage channelization index, the risk classification of the area where each measuring point is located is carried out.

[0006] In one possible implementation of this application, based on the energy intensity distribution at different measuring points in radar detection data, the water content change data at different measuring points are calculated, including: Extract multiple time-series radar data frames from radar detection data; Differential calculations were performed on the time-series radar data frames at different measuring points and the preset structural reflection map of the inherent structure of the target dam to obtain the response map of water content change at different measuring points; The response diagram of moisture content change is used as the data for moisture content change.

[0007] In one possible implementation of this application, a displacement cost matrix is ​​calculated based on the displacement of the moisture change tracer point corresponding to the moisture content change data at each measuring point at different times, including: The change energy envelope of the water content change response map at each measuring point is calculated using the Hilbert transform method. Based on the changing energy envelope, the total energy at different measuring points over the entire depth range is calculated; The total energy is divided into multiple identical energy intervals. For any given energy interval, the depth of the energy centroid of the current energy interval is calculated, and the depth of the energy centroid is used as a tracer point for water change. Based on the tracer points of moisture change in each energy range, a set of tracer point depths for each measuring point is constructed. The displacement cost matrix is ​​calculated based on the displacement difference of the moisture change tracer points in the continuous time-series tracer point depth set.

[0008] In one possible implementation of this application, the collaborative displacement cost matrix for each measuring point is determined based on the displacement differences between parallel paths formed by different element values ​​in the displacement cost matrix, including: Determine the standard deviation and median of all elements in the displacement cost matrix; The cost standard deviation is calculated based on the element standard deviation, and the collaborative discount factor is determined based on the median. Based on the Gaussian kernel function, the cost standard deviation, collaborative discount coefficient, and displacement differences of parallel paths between different element values ​​are calculated to obtain the collaborative displacement cost matrix of each measuring point.

[0009] In one possible implementation of this application, the seepage channelization index is determined based on the displacement cost matrix and the cooperative displacement cost matrix, including: Based on the displacement cost matrix, the minimum total displacement cost is calculated. The total cost of collaborative displacement is calculated based on the collaborative displacement cost matrix. The seepage channelization index is obtained by comparing the total cost of minimum displacement with the total cost of cooperative displacement.

[0010] In one possible implementation of this application, if the seepage channelization index is within a preset value range, it indicates that the moisture change tracer point is in a diffuse seepage mode; if the seepage channelization index is greater than a preset threshold, it indicates that the moisture change tracer point is in a channelized seepage mode.

[0011] In one possible implementation of this application, risk classification is performed on the area where each measuring point is located based on the displacement cost matrix corresponding index and the seepage channelization index, including: Determine the thresholds for judging the degree of intensity and the thresholds for judging the degree of coordination; The minimum total displacement cost corresponding to the displacement cost matrix is ​​compared with the severity judgment threshold, and the seepage channelization index is compared with the synergy judgment threshold; When the minimum total displacement cost and the seepage channelization index meet the corresponding division conditions of each preset division area, the area where each measuring point is located will be divided into the corresponding preset division area.

[0012] In one possible implementation of this application, when the minimum total displacement cost and the seepage channelization index satisfy the corresponding division conditions for each preset division region, the region where each measuring point is located is divided into the corresponding preset division region, including: If the total cost of minimum displacement is less than or equal to the threshold for judging the severity of the flow, and the seepage channelization index is less than or equal to the threshold for judging the degree of synergy, the area where the current measuring point is located will be classified as a stable area. If the total cost of minimum displacement is greater than the threshold for judging severity, and the seepage channelization index is less than or equal to the threshold for judging synergy, the area where the current measuring point is located is classified into the diffuse response area. If the total cost of minimum displacement is less than or equal to the severity judgment threshold, and the seepage channelization index is greater than the synergy judgment threshold, the area where the current measuring point is located will be classified into the early warning area. If the total cost of minimum displacement is greater than the threshold for judging severity, and the seepage channelization index is greater than the threshold for judging synergy, the area where the current measuring point is located will be classified as a high-risk area.

[0013] In one possible implementation of this application, after risk classification of the area where each measuring point is located based on the displacement cost matrix corresponding index and the seepage channelization index, the method further includes: The various types of seepage pattern zones are visualized in the form of a distribution map.

[0014] In one possible implementation of this application, the partitioned seepage pattern zones are visualized in the form of a distribution map, including: Based on the various types of seepage pattern zoning, a seepage pattern distribution map is generated and visualized. The seepage pattern distribution map uses the location of the measuring point as the horizontal axis and uses multiple colors to represent the seepage pattern zoning to which each measuring point currently belongs. The seepage pattern distribution map is used to characterize the location, risk range, and risk nature of the risk division area. and / or; Based on the zoning of various seepage patterns, a spatiotemporal evolution map of the seepage pattern is generated and visualized. The spatiotemporal evolution map of the seepage pattern uses the location of the measuring point as the horizontal axis and time as the vertical axis, and uses different colors to represent any coordinate point corresponding to each measuring point. The spatiotemporal evolution map of the seepage pattern is used to characterize the development process and migration process of the risk-divided area.

[0015] This application has, but is not limited to, the following technical effects: By acquiring radar detection data along the target dam body within a preset time period, and calculating the water content change data at different measuring points based on the energy intensity distribution of the radar detection data, a displacement cost matrix is ​​calculated based on the displacement degree of the corresponding moisture change tracer points at different times. This displacement cost matrix reflects the intensity of seepage movement. Based on the displacement differences between different adjacent element values ​​in the displacement cost matrix, a cooperative displacement cost matrix for each measuring point is determined. Furthermore, based on the displacement cost matrix and the cooperative displacement cost matrix, a seepage channelization index is determined. This seepage channelization index reflects the degree of coordination of water flow movement at the corresponding measuring point. Then, based on the indicators corresponding to the displacement cost matrix and the seepage channelization index, risk classification is performed on the areas where each measuring point is located. By reflecting "weak but highly coordinated" early channelization signals through the indicators corresponding to the displacement cost matrix and the seepage channelization index, early and accurate early warning can be achieved even in the nascent stage of potential hazards, before energy significantly increases. This realizes the application of ground-penetrating radar in the early diagnosis of seepage damage hazards. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the first embodiment of the high-risk hazard area detection method based on ground-penetrating radar of this application. Figure 2 This is a schematic diagram of the system implementation process involved in the high-risk hazard area detection method based on ground penetrating radar in this application; Figure 3 This is a schematic diagram of the device structure of the hardware operating environment involved in the embodiments of this application. Detailed Implementation

[0017] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0018] This application provides a method for detecting high-risk hazard areas based on ground-penetrating radar. In the first embodiment of this method, referring to... Figure 1 The methods include: Step S10: Obtain radar detection data within a preset time period along the target dam body survey line; As an example, the high-risk hazard area detection method based on ground-penetrating radar can be applied to a high-risk hazard area detection device based on ground-penetrating radar. This device belongs to a high-risk hazard area detection system based on ground-penetrating radar. A schematic diagram of the overall system implementation process involved in this application is shown below. Figure 2 As shown, the specific modules included in the system and the specific processing steps are as follows: 1. Dynamic response signal extraction and discretization module; This module is responsible for converting raw, continuous time-series radar data frames containing strong structural reflections into standardized objects that reflect only moisture changes and are discretized and can be used for dynamic tracking.

[0019] Internal processing steps: Acquisition and synchronization of time-series radar data frames; By performing differential processing with the structural reflection map, a response map of moisture content change is generated; Energy envelope calculation, isoenergy interval division and centroid calculation are performed on the response map to finally extract the tracer points of moisture change and form a set of tracer point depths; Input: Timing radar data frame and That is, two adjacent time-series radar data frames, structure reflection diagram Where x represents the horizontal position and d represents the depth. This represents the k-th discrete time. Similarly.

[0020] Output: Set of tracer point depths and .

[0021] 2. Cost matrix construction and reconstruction module; This module is responsible for establishing two different migration cost models: one is an independent displacement model that ignores the mutual influence between paths, and the other is a cooperative displacement model that introduces and amplifies the path synergy effect.

[0022] Internal processing steps: Based on the depth set of tracer points at two time points, a displacement cost matrix is ​​constructed. Through a nonlinear transformation that detects the cost similarity of adjacent parallel paths, the displacement cost matrix is ​​reconstructed into a cooperative displacement cost matrix.

[0023] Input: Set of depths of tracer points and .

[0024] Output: Displacement cost matrix and cooperative displacement cost matrix.

[0025] 3. Flow pattern quantification module; This module is responsible for extracting a pair of characteristic indicators from the two cost models through optimal allocation calculations, which can decouple and quantify the "intensity" and "coordination" of seepage movement.

[0026] Internal processing steps: The Hungarian algorithm is applied to the two cost matrices respectively to calculate the minimum total displacement cost and the cooperative total displacement cost.

[0027] The seepage channelization index is calculated by comparing the two total costs mentioned above.

[0028] Input: Displacement cost matrix and collaborative displacement cost matrix.

[0029] Output: A pair of characteristic metrics (minimum displacement total cost, seepage channelization index).

[0030] 4. Seepage mode diagnosis and output module; This module is responsible for using quantitative indicators to perform the final engineering diagnosis and outputting the results in an intuitive way.

[0031] Internal processing steps: The input feature index pairs are compared with the preset diagnostic zoning thresholds, and the current measurement point is classified into one of the four predefined seepage pattern zonings.

[0032] The diagnostic zoning results at all locations along the survey line are combined to generate a seepage pattern distribution map or a spatiotemporal evolution map.

[0033] Input: Feature index pair (minimum displacement total cost, seepage channelization index); diagnostic zoning threshold.

[0034] Output: Diagnostic results (such as a seepage pattern distribution map).

[0035] As an example, a fixed ground-penetrating radar survey line is laid out on the key monitoring section of the dam, which is the target dam body survey line. The preset time period can be 1 hour or 5 hours, and the specific time is not limited, depending on the real-time monitoring of the reservoir water level. The system adaptively adjusts the data acquisition frequency based on the rate of change of reservoir water level, setting a threshold for the rate of change, for example, 1 cm / hour. When the absolute value of the rate of change is less than this threshold, the system acquires data along the entire survey line at a lower frequency (e.g., every 6 hours). When the absolute value of the rate of change is greater than or equal to this threshold, the system automatically switches to a higher frequency (e.g., every 15 minutes) until the rate of change of reservoir water level falls below the threshold. Finally, while acquiring each radar data frame, the system synchronously acquires timestamped reservoir water level data from the dam safety monitoring system. Timestamp alignment ensures that each time-series radar data frame... Each value corresponds to a unique reservoir water level, forming the time series data required for subsequent analysis, resulting in radar detection data, where x represents the horizontal position and d represents the depth. This represents the k-th discrete time point, and the meanings of the following letters are similar.

[0036] Step S20: Based on the energy intensity distribution of different measuring points in the radar detection data, calculate the water content change data at different measuring points.

[0037] As an example, radar detection data includes time-series radar data frames at multiple moments. The following analysis focuses on a time-series radar data frame at any given moment. For different structures, the energy intensity of the signals received by the radar is different. Therefore, the water content change data at different measuring points can be calculated based on the energy intensity distribution at different measuring points. There are multiple measuring points on a measuring line, and for any given measuring point, there will be multiple measurement depths.

[0038] As an example, since the time-series radar data frames contain a mixture of weak signals caused by changes in water content and strong static reflections generated by the inherent structure of different materials and compaction layers inside the dam, the latter will seriously interfere with the identification of changes in the former. Therefore, differential processing must be used to separate this time-invariant structural reflection from the time-series data.

[0039] Specifically, the first step is to establish a structural reflection map. The establishment process is a one-time initialization step in this embodiment. A period in history when the reservoir water level is relatively low and has remained stable for a long time (e.g., more than a week) without significant water level fluctuations is selected. Radar data is collected on the same measuring line. After preprocessing methods known in the art, such as zero-point correction and bandpass filtering, a preset structure reflection map is obtained.

[0040] Step S20, which involves detecting high-risk hazard areas based on ground-penetrating radar, includes: Extract multiple time-series radar data frames from the radar detection data.

[0041] Differential calculations were performed on the time-series radar data frames at different measuring points and the preset structural reflection map of the inherent structure of the target dam to obtain the response map of water content change at different measuring points.

[0042] The response diagram of moisture content change is used as the data for moisture content change.

[0043] As an example, during continuous monitoring, for each newly acquired time-series radar data frame... First, execute and establish the preset structure reflection map. The preprocessing procedure is exactly the same as before. However, due to the "zero-point drift" of radar signals, high-frequency artifacts can occur, masking the true signal. Therefore, a channel alignment step is added to the preprocessing process. Using the peak value of the direct wave or surface reflection signal as a reference, the time-series radar data frames are aligned with the preset structure reflection map on the time axis, and then differential calculation is performed. Finally, the preprocessed time-series radar data frames are... Reflection diagram of the preset structure Perform differential calculations at each measurement point and each depth sampling point to generate a response map of moisture content changes. The calculation method is as follows: Response diagram of moisture content change The signal energy in the signal mainly reflects the time. The moisture content change response diagram, which shows the change in moisture content relative to the dry and stable state, is the data on the change in moisture content.

[0044] Step S30: Based on the displacement degree of the moisture change tracer point corresponding to the moisture change data at each measuring point at different times, calculate the displacement cost matrix; based on the displacement difference between the parallel paths formed by different element values ​​in the displacement cost matrix, determine the collaborative displacement cost matrix of each measuring point.

[0045] As an example, a moisture change tracer point can be an energy concentration point within a depth range obtained by quantifying the radar signal in the water content change response map. When this energy concentration point migrates at different time steps / time intervals, there will be a migration path cost, which is quantified by a displacement cost matrix.

[0046] As an example, due to the formation process of dominant seepage channels, there is a collective migration of water regions at similar depths in terms of direction and amplitude. This manifests as multiple "parallel" displacement paths in the movement of tracer points (e.g., tracer points). The nearby tracer points They have similar displacement costs, and their displacement cost matrices are... This characteristic cannot be captured, so a nonlinear transformation must be introduced that can automatically detect and reduce the cost of path combinations that exhibit strong synergy, thereby making the model sensitive to the emerging channelization patterns. Based on this, the synergistic displacement cost matrix of each measuring point is calculated.

[0047] In step S30, based on the displacement of the moisture change tracer points corresponding to the moisture content change data at each measuring point at different times, a displacement cost matrix is ​​calculated, including: The change energy envelope of the water content change response map at each measuring point is calculated using the Hilbert transform method. As an example, due to the response diagram of water content change In this context, the energy reflecting changes in moisture is continuously distributed along the depth direction, and its form and intensity vary greatly at different locations and times. In order to make quantitative tracking and comparison between different times, this continuous energy distribution needs to be transformed into discrete objects with fixed numbers and consistent weights.

[0048] Specifically, for the response diagram of moisture content change Any fixed position / measuring point position For a single-channel signal, the energy envelope is calculated using the Hilbert transform, resulting in a varying energy envelope. The Hilbert transform is a well-known technique for extracting signal envelopes, and will not be elaborated upon here.

[0049] Based on the changing energy envelope, the total energy at different measuring points over the entire depth range is calculated.

[0050] As an example, the change in energy envelope quantifies the position ,time Different depths The energy intensity of the change in water content at a given location can be used to determine the total energy intensity at different measuring points over the entire depth range (e.g., 100 meters).

[0051] The total energy is divided into multiple identical energy intervals. For any given energy interval, the depth of the energy centroid of the current energy interval is calculated, and the depth of the energy centroid is used as a tracer point for water change.

[0052] As an example, consider the changing energy envelope. Considered as along depth The energy density function is used to calculate the total energy along the entire depth. ,in, and The effective depth range of the signal is defined, and then the total energy is divided into... The energy is divided into several equal intervals, and the energy value of each interval is... , The value can be set according to the required resolution, for example, taking... .

[0053] As an example, for the first Energy ranges (where, From 1 to Its energy distribution spans the depth range. Inside, For the first The starting depth of each energy range, For the first The depth of the energy centroid of an energy interval is calculated by determining the endpoint depth of the interval. This energy centroid depth is defined as the depth of the th energy interval. Moisture change tracer points : The calculation result of this formula This represents the location of the energy concentration at depth for the change in the i-th equal amount of water; that is, this depth is the energy concentration point within the current energy range, where the energy is concentrated. Represents the changing energy envelope. This represents the energy value of the current energy range, where the total energy... If the depth of the tracer point is less than the preset noise threshold, the depth of the tracer point at that location is set to an invalid value (NaN) or the depth of the energy centroid at the previous moment, and is not included in subsequent calculations.

[0054] Based on the tracer points of moisture change in each energy range, a set of tracer point depths for each measuring point is constructed.

[0055] As an example, for any time any horizontal measuring point position Its response to changes in moisture content is uniformly represented as a function containing An ordered set of moisture change tracer points, i.e., the depth set of tracer points. : The elements in the set are ordered by index. Sort by shallowest to deepest This indicates the water change tracer point for the first energy range, and so on.

[0056] As an example, the dynamic analysis in this application is based on the analysis of two consecutive time points ( and The comparison of data states, therefore, the complete calculation process begins upon receiving the second timing radar data frame (i.e., ... ) will begin execution after that, representing a certain time period. The output of the analysis results is all timestamped to the end time of the specified time period. .

[0057] The displacement cost matrix is ​​calculated based on the displacement difference of the moisture change tracer points in the continuous time-series tracer point depth set.

[0058] As an example, in this step, the time intervals will be considered consecutive. and Two tracer depth sets and As input, the output is a Displacement cost matrix .

[0059] As an example, to quantify moisture change tracer points at a time step The intensity of migration within a given area necessitates a fundamental metric to measure the cost of all possible migration paths. A straightforward and effective approach is to use the square of the displacement distance, which assigns a higher cost weight to larger displacements.

[0060] Specifically, the displacement cost matrix The line, number Column elements Defined as in position At this location, a moisture change tracer point from time [time missing] of Depth-independent migration to time step of The cost corresponding to depth, where the index Corresponding to the start time The first in the depth set of tracer points Each tracer point, index Corresponding to the end time The first in the depth set of tracer points Each tracer point, displacement cost matrix The calculation formula is: Wherein, the displacement cost matrix Complete record of the location of the measuring point All The geometric displacement cost of the possible tracer point pairing paths is calculated, but the mutual influence between different displacement paths has not yet been considered.

[0061] Among them, step S30, which determines the collaborative displacement cost matrix of each measuring point based on the displacement differences between parallel paths formed by different element values ​​in the displacement cost matrix, includes: Determine the standard deviation and median of all elements in the displacement cost matrix; The cost standard deviation is calculated based on the element standard deviation, and the collaborative discount factor is determined based on the median.

[0062] As an example, the Gaussian kernel function was chosen as the similarity detector because its nonlinear characteristics are highly compatible with the process: when the cost difference between two parallel paths is small, the function value is close to 1, providing a significant cost discount; however, once the cost difference increases slightly, the function value quickly decays to close to 0. This characteristic can effectively distinguish between "highly coordinated" and "accidentally similar" paths. Because the displacement cost matrix varies under different operating conditions... The magnitude and distribution range of the elements in the data vary greatly. If fixed parameters are used for subsequent calculations, the results will be unstable. Therefore, the key parameters must be able to adaptively adjust according to the inherent scale of the current data.

[0063] As an example, first, calculate the displacement cost matrix. All Statistical characteristics of an element: elemental standard deviation and median Secondly, calculate the two adaptive parameters required for this step: 1. Calculate the cost standard deviation based on the element standard deviation. , specifically, This parameter serves as the bandwidth of the Gaussian kernel function, and its value is proportional to the dispersion of the data itself, ensuring that the judgment criterion of "similar costs" is relative rather than absolute. It is a dimensionless global constant, and in one implementation, it is taken as... .

[0064] Calculate the collaborative discount factor based on the median. : This parameter determines the magnitude of the cost discount applied to detected synergies, and it is inversely proportional to the median cost, ensuring that the magnitude of the discount matches the magnitude of the base cost. It is a dimensionless global constant. It is a very small positive number set to prevent the denominator from being zero.

[0065] Based on the Gaussian kernel function, the cost standard deviation, collaborative discount coefficient, and displacement differences of parallel paths between different element values ​​are calculated to obtain the collaborative displacement cost matrix of each measuring point.

[0066] As an example, take measurement point x, time... For example, the displacement cost matrix is ​​generated through the following process. Reconstructed into a collaborative displacement cost matrix Collaborative displacement cost matrix The calculation method can be: In the formula: It is a small integer that defines the neighborhood radius of adjacent parallel paths that need to be considered in the depth direction. In one implementation, it is taken as... Summation terms for neighborhood index Perform the traversal, where, from arrive But excluding This summation only applies to conditions satisfying... and index Perform calculations to ensure that index out-of-bounds errors do not occur. The term refers to the Gaussian kernel function, used to quantize the current path. Its parallel neighbor paths Cost similarity. The entire subtraction represents the path. The total cost discount that should be obtained due to its synergy with neighboring paths; if the path If the motion is isolated, the reduction term approaches zero; if it is part of a bundle of coordinated motion paths, its cost will be significantly reduced. The number of effective neighborhood points is calculated when... When the value is less than 0, the output result is 0 to avoid negative values.

[0067] Step S40: Based on the displacement cost matrix and the cooperative displacement cost matrix, determine the seepage channelization index. The seepage channelization index is used to characterize the degree of cooperative motion of the water flow corresponding to each measuring point.

[0068] Step S40 includes: Based on the displacement cost matrix, the minimum total displacement cost is calculated.

[0069] The total cost of collaborative displacement is calculated based on the collaborative displacement cost matrix.

[0070] As an example, to find the migration scheme with the lowest total cost from the two cost matrices mentioned above, this embodiment of the application chooses to use the Hungarian algorithm (a well-known algorithm for solving optimal allocation problems, a classic algorithm for solving the maximum matching problem and assignment problem in bipartite graphs, the implementation of which will not be elaborated here). This is because the collective migration of moisture change tracer points is a globally optimal allocation problem. A simple greedy matching algorithm cannot guarantee that the final total cost is globally minimum. The Hungarian algorithm, as a standard deterministic algorithm for solving assignment problems, can guarantee finding a one-to-one, globally optimal matching scheme, thereby ensuring that the calculated total cost has a clear and unique meaning. For the displacement cost matrix, a "maximum tolerable displacement distance" threshold is introduced. Matrix elements exceeding the threshold are directly set as maximum penalty values, and physical displacement is no longer forcibly calculated.

[0071] As an example, the Hungarian algorithm is applied to two cost matrices to calculate the minimum total cost: 1. Calculate the minimum total cost of displacement. This value represents the theoretical minimum total cost required by the system to complete the tracer point migration without considering path cooperability. It is calculated as follows: in, ( ) represents the Hungarian algorithm.

[0072] Among them, the total cost of coordinated displacement The calculation method is as follows: In the formula, the total cost of cooperative displacement represents the theoretical minimum total cost required for the system to complete the migration after considering and rewarding path cooperation.

[0073] The seepage channelization index is obtained by comparing the total cost of minimum displacement with the total cost of cooperative displacement.

[0074] As an example, the seepage channelization index is calculated by comparing the two total costs mentioned above. : Among them, the numerator of the seepage channelization index This represents the total cost reduction achieved due to the detection of coordinated motion, calculated by dividing by... Normalize it to make it a relative indicator that is independent of the absolute value of the cost. It is a very small positive number set to prevent the denominator from being zero.

[0075] If the seepage channelization index is within the preset range, it indicates that the moisture change tracer point is in the diffuse seepage mode; if the seepage channelization index is greater than the preset threshold, it indicates that the moisture change tracer point is in the channelized seepage mode.

[0076] As an example, when When the value approaches 0, it indicates that the movement of the tracer point is disordered, corresponding to the diffuse seepage mode. The preset value range can be the interval (0, 1). When the value is significantly greater than 0, it indicates that there is strong cooperative motion at the tracer points, corresponding to the channelized seepage mode. The preset threshold here can be 3, 5, etc., and there is no specific limitation.

[0077] Step S50: Based on the displacement cost matrix corresponding indicators and the seepage channelization index, risk classification is performed on the area where each measuring point is located.

[0078] As an example, after determining these two indicators, each measuring point is assigned to a corresponding risk area based on the displacement cost matrix and the seepage channelization index. This embodiment aims to identify high-risk hazard areas using ground-penetrating radar data. Its basis lies in diagnosing the seepage "pattern" within the dam body, rather than simply locating water. The essence of seepage risk lies in whether its movement pattern changes from a disordered, diffuse state to a concentrated, channelized state. Therefore, the final diagnosis of risk requires considering both the "intensity" and "coordination" of the seepage movement. Based on the displacement cost matrix and the seepage channelization index, the seepage state is mapped into a two-dimensional diagnostic space for classification, achieving a qualitative diagnosis of the seepage pattern.

[0079] Step S50 includes: Determine the thresholds for judging the severity and the degree of synergy.

[0080] As an example, the threshold for judging severity and the threshold for judging synergy are both diagnostic zoning thresholds. These thresholds should be determined through statistical analysis of monitoring data of a specific dam under long-term, normal operation conditions in order to establish a benchmark for the "healthy" behavior of the dam.

[0081] One implementation method is as follows: First, time-series radar data frames are collected from the dam during multiple complete reservoir water level fluctuation cycles in history, under conditions where no known anomalies have occurred. Steps S10-S30 of this embodiment are then performed on these historical data to calculate a "healthy" dataset containing a large number of historical data points, where each data point represents a pair of feature indicators. Then, statistical analysis was performed on the health dataset to determine two judgment thresholds: Intensity judgment threshold : Retrieve all history The 95th percentile of the value is used as .

[0082] Threshold for judging the degree of collaboration : Retrieve all history The 95th percentile of the value is used as Alternatively, it can be determined based on the simulation results of the finite element seepage simulation model, or by using empirical default values ​​(e.g., TC=100). =0.3) Run first, and update dynamically as data accumulates.

[0083] The minimum total displacement cost corresponding to the displacement cost matrix is ​​compared with the severity judgment threshold, and the seepage channelization index is compared with the synergy judgment threshold.

[0084] As an example, the minimum displacement total cost is used to reflect the intensity of motion in the seepage mode, while the seepage channelization index is used to reflect the degree of coordination of seepage motion in the seepage mode. These two indicators are compared with their corresponding thresholds to achieve the purpose of risk zoning.

[0085] When the minimum total displacement cost and the seepage channelization index meet the corresponding division conditions of each preset division area, the area where each measuring point is located will be divided into the corresponding preset division area.

[0086] As an example, the preset division areas include Zone I (stable area), Zone II (diffusion response area), Zone III (early warning area), and Zone IV (high-risk area). For each division area, there are corresponding division conditions. When the two indicators of any measuring point meet the corresponding division conditions, the area where each measuring point is located is divided into the corresponding preset division area.

[0087] The step of assigning the area where each measuring point is located to the corresponding preset division area when the minimum total displacement cost and the seepage channelization index meet the division conditions of each preset division area includes: If the total cost of minimum displacement is less than or equal to the threshold for judging the severity of the flow, and the seepage channelization index is less than or equal to the threshold for judging the degree of synergy, the area where the current measuring point is located will be classified as a stable region.

[0088] If the total cost of minimum displacement is greater than the severity judgment threshold, and the seepage channelization index is less than or equal to the synergy judgment threshold, the area where the current measuring point is located is classified into the diffuse response area.

[0089] As an example, if the total cost of minimum displacement If the value is less than the preset "silent threshold" Tsilent (using the statistical average of the total displacement cost during historical periods without anomalies), a mandatory judgment will be made. =0, and is classified as Zone I (stable region), skipping the ratio calculation to avoid minor noise causing random fluctuations in PCI values ​​and generating false alarms.

[0090] As an example, the above two thresholds are used to determine the severity of the situation. And the threshold for judging the degree of cooperation , will be For the horizontal axis, The two-dimensional diagnostic space along the vertical axis is divided into four zones, each corresponding to a specific flow pattern: Zone I (Stable Region): and This area represents weak and disordered water movement, corresponding to a healthy and stable seepage state of the dam.

[0091] Zone II (Diffusion Response Area): and This area represents a region of intense but disordered water movement, typically corresponding to the normal, large-scale diffuse seepage response of the dam body when the reservoir water level changes rapidly.

[0092] If the total cost of minimum displacement is less than or equal to the severity judgment threshold, and the seepage channelization index is greater than the synergy judgment threshold, the area where the current measuring point is located will be classified as an early warning area.

[0093] If the total cost of minimum displacement is greater than the threshold for judging severity, and the seepage channelization index is greater than the threshold for judging synergy, the area where the current measuring point is located will be classified as a high-risk area.

[0094] As an example, Zone III (early warning zone): and This area represents a region where overall water movement is weak, but highly coordinated displacement paths have formed locally. This is a key characteristic of dominant seepage channels (such as piping and fissures) in their nascent or slow-developing stages.

[0095] Zone IV (High-risk area): and This area represents intense and highly coordinated water movement, indicating that the dominant seepage channels have developed or are active, and the risk level is high.

[0096] As an example, after comparing the indicators of each measuring point with the corresponding thresholds, the measuring points are... The condition is categorized into one of the four diagnostic zones I, II, III, and IV mentioned above. Measurement points categorized into zone III or IV are... This refers to areas identified as high-risk or potentially hazardous areas. Water level data should be based on the time the radar antenna moves to a specific measuring point x. Instead of using the start time of the entire frame of data, linear interpolation is used to eliminate spatiotemporal mismatches caused by rapid water level changes.

[0097] After step S50, the following is also included: The various types of seepage pattern zones are visualized in the form of a distribution map.

[0098] Specifically, the steps for visually representing the completed partitioning of various seepage pattern zones in the form of a distribution map include: Based on the various types of seepage pattern zoning, a seepage pattern distribution map is generated and visualized. The seepage pattern distribution map uses the location of the measuring point as the horizontal axis and uses multiple colors to represent the seepage pattern zoning to which each measuring point currently belongs. The seepage pattern distribution map is used to characterize the location, risk range, and risk nature of the risk-defined area.

[0099] As an example, the distribution map includes a seepage pattern distribution map and a seepage pattern spatiotemporal evolution map. The purpose of this embodiment is to transform abstract diagnostic results into intuitive graphics that are easy for engineers to interpret. The input is the location of all measuring points along the survey line. At any moment The diagnostic partitioning results.

[0100] As an example, the diagnostic zoning result is the seepage pattern zoning to which the measuring point belongs. After generating the seepage pattern distribution map, the map uses the measuring point location as the horizontal axis and different colors to represent the seepage pattern zoning to which each measuring point location currently belongs. For example, green represents Zone I (stable zone), blue represents Zone II (diffuse response zone), yellow represents Zone III (early warning zone), and red represents Zone IV (high-risk zone). Through this map, the location, extent, and risk nature of high-risk areas can be clearly observed at a glance.

[0101] and / or; Based on the zoning of various seepage patterns, a spatiotemporal evolution map of the seepage pattern is generated and visualized. The spatiotemporal evolution map of the seepage pattern uses the location of the measuring point as the horizontal axis and time as the vertical axis, and uses different colors to represent any coordinate point corresponding to each measuring point. The spatiotemporal evolution map of the seepage pattern is used to characterize the development process and migration process of the risk-divided area.

[0102] As an example, a spatiotemporal evolution diagram of the seepage pattern can also be generated. This diagram is a two-dimensional graph with the measurement point location on the horizontal axis and time on the vertical axis. Any coordinate point in the graph... The color is determined by the diagnostic zone to which the point belongs. Through this map, the formation, development and migration process of high-risk areas can be fully traced. For example, the complete evolutionary path of an area from blue (diffusion response zone) to yellow (early warning zone) and then to red (high-risk zone) can be observed.

[0103] This application provides a method for detecting high-risk hazard areas based on ground-penetrating radar (GPR). By acquiring radar detection data within a preset time period along the target dam's survey line, and calculating the water content change data at different measuring points based on the energy intensity distribution of the radar data, a displacement cost matrix is ​​calculated. This displacement cost matrix reflects the intensity of seepage movement. Based on the displacement differences between adjacent elements in the displacement cost matrix, a collaborative displacement cost matrix for each measuring point is determined. Furthermore, a seepage channelization index is determined based on the displacement cost matrix and the collaborative displacement cost matrix. This seepage channelization index reflects the degree of coordination in the water flow movement corresponding to the measuring point. Finally, based on the displacement cost matrix and the seepage channelization index, the area where each measuring point is located is classified for risk assessment. By reflecting "weak but highly coordinated" early channelization signals through the displacement cost matrix and the seepage channelization index, early and accurate warnings can be achieved even in the nascent stage of a hazard, before the energy has significantly increased. This realizes the application of GPR in the early diagnosis of seepage damage hazards.

[0104] Reference Figure 3 , Figure 3 This is a schematic diagram of the device structure of the hardware operating environment involved in the embodiments of this application.

[0105] like Figure 3 As shown, the high-risk hazard area detection device based on ground-penetrating radar may include: a processor 1001, a memory 1003, and a communication bus 1002. The communication bus 1002 is used to realize the connection and communication between the processor 1001 and the memory 1003.

[0106] Optionally, the high-risk hazard area detection equipment based on ground-penetrating radar may also include a user interface, a network interface, a camera, RF (Radio Frequency) circuitry, sensors, a WiFi module, etc. The user interface may include a display screen and an input submodule such as a keyboard; optional user interfaces may also include standard wired or wireless interfaces. The network interface may include standard wired or wireless interfaces (such as a Wi-Fi interface).

[0107] Those skilled in the art will understand that Figure 3 The structure of the high-risk hazard area detection equipment based on ground penetrating radar shown in the figure does not constitute a limitation on the high-risk hazard area detection equipment based on ground penetrating radar. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0108] like Figure 3As shown, the memory 1003, serving as a storage medium, may include an operating system, a network communication module, and a ground-penetrating radar-based high-risk hazard area detection program. The operating system is a program that manages and controls the hardware and software resources of the ground-penetrating radar-based high-risk hazard area detection equipment, supporting the operation of the ground-penetrating radar-based high-risk hazard area detection program and other software and / or programs. The network communication module is used to enable communication between the various components within the memory 1003, as well as communication with other hardware and software in the ground-penetrating radar-based high-risk hazard area detection system.

[0109] exist Figure 3 In the high-risk hazard area detection device based on ground penetrating radar shown, the processor 1001 is used to execute the high-risk hazard area detection program based on ground penetrating radar stored in the memory 1003 to implement the steps of the high-risk hazard area detection method based on ground penetrating radar described above.

[0110] The specific implementation method of the high-risk hidden danger area detection equipment based on ground penetrating radar in this application is basically the same as the various embodiments of the high-risk hidden danger area detection method based on ground penetrating radar described above, and will not be repeated here.

[0111] It should be noted that, in this document, the terms “comprising,” “including,” or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system.

[0112] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0113] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0114] The above are merely preferred embodiments of this application and do not limit the scope of this application. Any equivalent structural or procedural transformations made based on the description and drawings of this application, or direct or indirect applications in other related technical fields, are similarly included within the scope of protection of this application.

[0115] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0116] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for detecting high-risk hidden danger areas based on ground-penetrating radar, characterized in that, The method includes: Acquire radar detection data within a preset time period along the survey line of the target dam body; Based on the energy intensity distribution at different measuring points in the radar detection data, the water content variation data at different measuring points are calculated, specifically including: Extract multiple time-series radar data frames from the radar detection data; Differential calculations are performed on the time-series radar data frames at different measuring points and the preset structural reflection diagrams of the inherent structure of the target dam to obtain the water content change response diagrams at different measuring points. The moisture content change response map is used as the moisture content change data; Based on the displacement of the moisture change tracer points corresponding to the moisture change data at each of the aforementioned measuring points at different times, a displacement cost matrix is ​​calculated, which specifically includes: The change energy envelope of the water content change response map at each of the aforementioned measuring points is calculated using the Hilbert transform method. Based on the changing energy envelope, the total energy at different measuring points over the entire depth range is calculated; The total energy is divided into multiple identical energy intervals. For any energy interval, the depth of the energy centroid of the current energy interval is calculated, and the depth of the energy centroid is used as a tracer point for water change. Based on the moisture change tracer points in each of the energy ranges, a set of tracer point depths for each measuring point is constructed. The displacement cost matrix is ​​calculated based on the displacement difference of moisture change tracer points in the continuous time-series tracer point depth set. Based on the displacement differences between parallel paths formed by different element values ​​in the displacement cost matrix, the collaborative displacement cost matrix of each measuring point is determined. Based on the displacement cost matrix and the cooperative displacement cost matrix, a seepage channelization index is determined, which is used to characterize the degree of motion coordination of the water flow corresponding to each measuring point. Based on the displacement cost matrix corresponding index and the seepage channelization index, the risk classification is performed on the area where each measuring point is located.

2. The method for detecting high-risk hidden danger areas based on ground-penetrating radar as described in claim 1, characterized in that, The determination of the collaborative displacement cost matrix for each measuring point based on the displacement differences between parallel paths formed by different element values ​​in the displacement cost matrix includes: Determine the standard deviation and median of all elements in the displacement cost matrix; Based on the standard deviation of the elements, the standard deviation of the cost is calculated, and based on the median, the collaborative discount factor is determined. Based on the Gaussian kernel function, the cost standard deviation, the collaborative discount coefficient, and the displacement difference of parallel paths between different element values ​​are calculated to obtain the collaborative displacement cost matrix of each measuring point.

3. The method for detecting high-risk hidden danger areas based on ground-penetrating radar as described in claim 1, characterized in that, The determination of the seepage channelization index based on the displacement cost matrix and the cooperative displacement cost matrix includes: Based on the displacement cost matrix, the minimum total displacement cost is calculated. Based on the aforementioned collaborative displacement cost matrix, the total collaborative displacement cost is calculated. The minimum displacement total cost is compared with the cooperative displacement total cost to obtain the seepage channelization index.

4. The method for detecting high-risk hidden danger areas based on ground-penetrating radar as described in claim 3, characterized in that, If the seepage channelization index is within a preset value range, it indicates that the moisture change tracer point is in a diffuse seepage mode. If the seepage channelization index is greater than a preset threshold, it indicates that the moisture change tracer point is in a channelized seepage mode.

5. The method for detecting high-risk hidden danger areas based on ground-penetrating radar as described in claim 1, characterized in that, The risk classification process for the area where each measuring point is located, based on the displacement cost matrix corresponding index and the seepage channelization index, includes: Determine the thresholds for judging the degree of intensity and the thresholds for judging the degree of coordination; The minimum total displacement cost corresponding to the displacement cost matrix is ​​compared with the severity judgment threshold, and the seepage channelization index is compared with the synergy judgment threshold. When the minimum total displacement cost and the seepage channelization index satisfy the corresponding division conditions of each preset division area, the area where each measuring point is located is divided into the corresponding preset division area.

6. The method for detecting high-risk hidden danger areas based on ground-penetrating radar as described in claim 5, characterized in that, When the minimum total displacement cost and the seepage channelization index satisfy the corresponding division conditions of each preset division region, the region where each measuring point is located is divided into the corresponding preset division region, including: If the total cost of the minimum displacement is less than or equal to the severity judgment threshold, and the seepage channelization index is less than or equal to the synergy judgment threshold, the area where the current measuring point is located is classified into a stable area. If the minimum displacement total cost is greater than the severity judgment threshold, and the seepage channelization index is less than or equal to the synergy judgment threshold, the area where the current measuring point is located is classified into the diffuse response area. If the total cost of the minimum displacement is less than or equal to the severity judgment threshold, and the seepage channelization index is greater than the synergy judgment threshold, the area where the current measuring point is located will be classified into the early warning area. If the total cost of the minimum displacement is greater than the severity judgment threshold, and the seepage channelization index is greater than the synergy judgment threshold, the area where the current measuring point is located will be classified as a high-risk area.

7. The method for detecting high-risk hidden danger areas based on ground-penetrating radar as described in claim 1, characterized in that, After classifying the areas where each measuring point is located based on the displacement cost matrix corresponding index and the seepage channelization index, the process further includes: The various types of seepage pattern zones are visualized in the form of a distribution map.

8. The method for detecting high-risk hidden danger areas based on ground-penetrating radar as described in claim 7, characterized in that, The visualization of the various seepage pattern zones, presented in the form of a distribution map, includes: Based on the various types of seepage pattern zoning, a seepage pattern distribution map is generated and visualized. The seepage pattern distribution map uses the location of the measuring point as the horizontal axis and uses multiple colors to represent the seepage pattern zoning to which each measuring point currently belongs. The seepage pattern distribution map is used to characterize the location, risk range, and risk nature of the risk division area. and / or; Based on the zoning of various seepage patterns, a spatiotemporal evolution map of the seepage pattern is generated and visualized. The spatiotemporal evolution map of the seepage pattern uses the position of the measuring point as the horizontal axis and time as the vertical axis, and uses different colors to represent any coordinate point corresponding to each measuring point. The spatiotemporal evolution map of the seepage pattern is used to characterize the development process and migration process of the risk-divided area.

Images