Dam disease three-dimensional nondestructive detection method, storage medium and computer equipment

Abstract

The invention discloses a dam disease three-dimensional nondestructive detection method, a storage medium and computer equipment, and relates to the technical field of geophysical detection.The dam disease three-dimensional nondestructive detection method mainly comprises the steps that a seismic source and a detector are used for detecting a target dam to obtain seismic data; obtaining first arrival travel time by using an improved self-correlation constraint energy ratio method, constructing an underground medium wave impedance initial model by using a back projection algorithm, performing forward modeling calculation to obtain theoretical travel time, and correcting an average slowness value in each grid by using a joint iteration method to obtain a corrected underground medium wave impedance model; and stopping iteration when the iteration is ended to obtain a dam disease three-dimensional nondestructive detection result. By implementing the dam disease three-dimensional nondestructive detection method, the storage medium and the computer equipment provided by the invention, the position of a dam hidden danger can be quickly, efficiently and accurately detected.

Description

Technical Field

[0001] The present invention relates to the field of geophysical detection technology, and more particularly to a three-dimensional non-destructive detection method for dam defects, a storage medium, and a computer device. Background Art

[0002] In the field of water conservancy projects, dam safety is of vital importance. Three-dimensional seismic detection technology provides an efficient and accurate geophysical means for detecting hidden dangers in dams.

[0003] Currently, three main technical methods are used to detect dam hidden dangers: drilling, manual inspection, and geophysical exploration. While drilling is feasible, it is costly, inefficient, and can cause damage to the dam structure. Manual inspection relies heavily on experience, is equally inefficient, and struggles to detect hidden hazards. Geophysical exploration relies on the physical property differences between the dam and the hidden dangers. Specifically, hidden dangers such as cracks, leaks, weak bodies, and loose solids within the dam exhibit significant differences from the surrounding medium in terms of resistivity, longitudinal and shear wave velocities, and wave velocity and impedance interfaces. Geophysical exploration detects hidden dangers within earth-rockfill dams by studying the changes in dynamic and kinematic characteristics caused by these differences. Seismic tomography has made significant progress in dam inspection. Currently, seismic tomography inspections of dams mostly involve drilling holes on both sides of the dam for cross-well tomography. This method causes a certain degree of damage to the dam and does not fully realize the advantages of geophysical non-destructive detection.

[0004] One geophysical exploration method uses cross-well seismic technology for dam surveying. Two wells are drilled at the dam crest: one well serves as an "excitation well" with a seismic source, and the other, a "receiving well," houses a geophone. The seismic detectors are then deployed to survey the strata between the two wells. However, in practice, this method often fails to achieve optimal results due to damage to the dam, the diversity of dam structures, and the heterogeneity and anisotropy of dam construction materials.

[0005] The above content is only used to assist in understanding the technical solution of the present invention and does not constitute an admission that the above content is prior art. Summary of the Invention

[0006] The purpose of the present invention is to provide a three-dimensional non-destructive detection method for dam defects, a storage medium and a computer device, which can quickly, efficiently and accurately detect the location of dam hidden dangers.

[0007] The present invention provides a three-dimensional non-destructive detection method for dam diseases, comprising the following steps: S1: Use the source and detector to detect the target dam and obtain seismic data; S2: Based on the seismic data, the first arrival travel time is obtained using the improved autocorrelation constrained energy ratio method; S3: constructing an initial model of wave impedance of the underground medium according to the seismic data by using a back projection algorithm, and performing forward calculation to obtain a theoretical travel time; S4: calculating a travel time residual according to the first arrival travel time and the theoretical travel time, correcting an average slowness value in each grid by using a joint iteration method according to the initial model of wave impedance of the underground medium, and obtaining a corrected model of wave impedance of the underground medium; S5: returning to step S4 when an iteration termination condition is not met, and stopping iteration when the iteration termination condition is met, and obtaining a result of three-dimensional nondestructive detection of dam diseases according to the corrected model of wave impedance of the underground medium.

[0008] Further, step S2 specifically comprises: obtaining the first arrival travel time according to the seismic data by using an improved autocorrelation constraint energy ratio method, as shown in the following formula: , , , , , , , wherein, is a recording length; is an amplitude value of the seismic record at a time point is a short-time window autocorrelation; is a length of the short-time window; is an amplitude value of the seismic record at a time point ; is an amplitude value of the seismic record at , and constitutes a signal segment in a time window; is a long-time window autocorrelation; is a length of the long-time window; is a short-time window average energy; is a short-time window autocorrelation function value at a time point ; is a long-time window average energy; is a long-time window autocorrelation function value at a time point ; is a short-long-time window energy ratio.

[0009] Further, step S3 specifically comprises: constructing an initial model of wave impedance of the underground medium according to the seismic data by using a back projection algorithm, as shown in the following formula: , , , , wherein, is the time required to cross the grid; denotes the seismic ray travel time; denotes the propagation distance of the th seismic ray within the grid; denotes the total distance of the ray propagation path; is the total travel time required for all rays in the grid; is the total propagation distance required for all rays in the grid; is the average slowness value within the grid.

[0010] Further, step S4 specifically comprises: calculating a travel time residual according to the first arrival travel time and the theoretical travel time, correcting the average slowness value within each grid using a joint iteration method according to the initial wave impedance model of the underground medium, and obtaining a corrected wave impedance model of the underground medium, as shown in the following formula: , , wherein, denotes the ray number; denotes the grid number; is the grid numbered of the th iteration model vector; is the grid numbered of the th iteration model vector; is the total number of rays passing through the grid j; is the relaxation factor; is the current processed ray number; is the total number of rays; is the first arrival travel time; is the theoretical travel time of the th ray; is the path length of the ray within the grid ; is the grid numbered ; is the effective ray coverage of the grid ; is the average travel time residual of the grid .

[0011] Further, the iteration termination condition comprises that a travel time residual difference between the modified wave impedance model of the underground medium and a travel time residual difference of an initial wave impedance model of the underground medium is less than a preset difference value.

[0012] Further, the iteration termination condition further comprises that an updating amount of the wave impedance model of the underground medium does not substantially change with an increase in the iteration number.

[0013] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the dam disease three-dimensional nondestructive detection method.

[0014] The application further provides a computer device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the dam disease three-dimensional nondestructive detection method when executing the program.

[0015] The dam disease three-dimensional nondestructive detection method, the storage medium and the computer device provided by the application have the following beneficial effects: The application adopts the "pseudo-well interlayer tomography" technology, and through utilizing the geometric characteristics of the dam body, the seismic sources and the geophones are arranged on the dam top and the two sides of the dam slope to form a three-dimensional observation system surrounded by three sides, and the three-dimensional exploration effect similar to the well interlayer seismic exploration can be achieved without drilling, the direct wave data directly penetrating the dam body can be obtained, and various hidden structure such as cracks and leakage areas can be effectively detected; in the specific implementation, first, the seismic sources and the geophones are alternately arranged on the dam top and the two sides of the dam slope to form three kinds of excitation-receiving combinations: when the left dam slope is excited, the right dam slope and the dam top receive the signals; when the right dam slope is excited, the left dam slope and the dam top receive the signals; and when the dam top is excited, the two sides of the dam slope receive the signals at the same time. The multi-angle observation mode significantly improves the data coverage range, and makes the tomography result more accurate. The application adopts an improved autocorrelation constraint energy ratio method in a data processing stage to pick up a first arrival travel time, the method has strong anti-noise ability and can accurately extract the first arrival time of the seismic wave; wherein, the establishment of the initial model is based on the equivalent medium theory of the soil-rock mixture, combined with the physical property parameter table in the water conservancy industry standard, the equivalent parameters of the composite medium are calculated according to the proportion of the soil material, the stone material, the water and the air, and the model is ensured to be closer to the real dam structure; the inversion calculation adopts a joint iteration method (SIRT), the algorithm is not sensitive to the measurement error, has good convergence and is suitable for the imaging problem of the complex medium; in the iteration process, the model parameters are continuously corrected through the calculation of the travel time residual (the difference between the measured travel time and the theoretical travel time), until the termination condition is met, that is, the travel time residual is reduced by 1-2 orders of magnitude, and the model update amount tends to be stable; the observation system adopts a trapezoidal profile design, which is different from the traditional rectangular grid and is more consistent with the actual geometric shape of the dam; finally, through three-dimensional grid discretization and matrix operation (including travel time matrix, distance matrix and slowness matrix), the wave impedance distribution image inside the dam is output, and the spatial position and physical property abnormal characteristics (such as loose area showing low-velocity anomaly) of the hidden trouble body are accurately identified; The application constructs a three-dimensional observation system surrounded by three sides by innovatively arranging the seismic source and the geophone on the dam top and the two sides of the dam slope, realizes the three-dimensional exploration effect similar to interwell seismic, and avoids the disadvantages of needing to drill holes in the traditional method; this unique observation method significantly improves the comprehensiveness and reliability of data acquisition, can obtain key seismic wave information such as direct wave, and lays a solid foundation for subsequent accurate imaging; in the data processing and inversion calculation, the improved autocorrelation constraint energy ratio method is used to pick up the first arrival travel time, which greatly improves the accuracy and anti-interference ability of data processing; combined with the physical property parameters in the water conservancy industry standard and the equivalent medium theory of the soil-rock mixture, the initial model is closer to the actual dam structure characteristics; the SIRT inversion algorithm has the advantages of good convergence and insensitivity to error, and finally generates a high-precision three-dimensional wave impedance distribution image through iteration calculation; the method can clearly identify various hidden trouble structures inside the dam, including cracks, leakage channels, weak zones and the like, has high positioning accuracy, and can effectively distinguish different types of hidden trouble characteristics such as water-containing area (high-velocity anomaly) and loose area (low-velocity anomaly); In summary, the application uses three-dimensional seismic detection technology to detect the dam hidden trouble through the first arrival wave travel time imaging of the seismic wave, can quickly and efficiently and accurately detect the position of the dam hidden trouble, and provides technical support for disaster prevention and management of the dam. BRIEF DESCRIPTION OF DRAWINGS

[0016] The application will be further described below in combination with the drawings and examples, and the drawings are as follows: Figure 1 is the flow chart of the dam disease three-dimensional nondestructive detection method provided by the application; Figure 2The dam three-dimensional measuring line layout provided by the present application is provided; Figure 3 The dam body pseudo-well interlayer tomography observation system two-dimensional profile provided by the present application is provided; Figure 4 The rectangular grid dissection schematic diagram provided by the present application is provided; Figure 5 The structural block diagram of the computer equipment provided by the present application is provided. DETAILED DESCRIPTION

[0017] In order to have a clearer understanding of the technical features, objectives and effects of the present application, the specific embodiments of the present application will be described in detail with reference to the drawings.

[0018] Figure 1 A schematic diagram of the dam disease three-dimensional nondestructive detection method of the present embodiment is shown. In the present embodiment, the dam disease three-dimensional nondestructive detection method comprises the following steps: S1: detecting the target dam by using a seismic source and a geophone to obtain seismic data; As an exemplary embodiment, in step S1, the existing data is analyzed, and according to the actual detection needs, suitable measuring lines are arranged according to the distance length, etc. at the detection position, and the measuring points are set according to the detection accuracy requirements; in the present embodiment, the geophones are arranged to surround the dam, and when the seismic source is arranged on the left dam, the geophones on the right dam slope and the dam top receive the seismic signals; when the seismic source is arranged on the right dam slope, the geophones on the left dam slope and the dam top receive the seismic signals; when the seismic source is arranged on the dam top, the geophones on the two dam slopes receive the seismic signals; the seismic source is excited in turn to collect the seismic data by using the three-dimensional seismic technology, and the data is processed, and finally the generated seismic record is interpreted; In the present embodiment, the seismic source and the geophone are arranged on the dam top and the two dam slopes, the dam body is surrounded from two sides or even three sides, and a natural pseudo-well interlayer tomography observation system is formed. In the specific operation, first, the seismic source is excited on one side of the dam slope, and the geophones are arranged on the dam top and the opposite side of the dam slope to receive the seismic waves; then, the seismic source is excited on the other side of the dam slope, and the geophones are arranged on the dam top and the opposite side of the dam slope to receive the seismic waves; finally, the seismic source is excited on the dam top, and the geophones are arranged on the left and right sides of the dam slope to receive the seismic waves, so as to perform three-dimensional seismic exploration. This arrangement not only enriches the arrangement method and the selection of the observation system, but also can obtain the direct wave which is difficult to capture by general geophysical methods. Through the analysis of the response of different target structures (such as cracks, leakage areas, weak bodies, and not dense bodies, etc.) to the seismic waves, the tomography of the physical property difference of the dam body can be performed, and then the potential hidden danger structure in the dam body can be accurately detected.

[0019] S2: Based on the seismic data, the first arrival travel time is obtained using the improved autocorrelation constrained energy ratio method; In an exemplary embodiment, step S2 specifically includes: obtaining the first arrival travel time based on the seismic data using an improved autocorrelation constrained energy ratio method, such as the formula: , , , , , , , in, is the record length; For a certain moment Short-time window autocorrelation; is the length of the short-term window; Record earthquakes at a specific time The amplitude value at ; For earthquake records The amplitude value at Constructing signal segments within the time window; is the long-term window autocorrelation; is the length of the long time window; is the average energy in the short time window; For time point The short-time window autocorrelation function value at ; is the average energy of the long time window; For time point The long-time window autocorrelation function value at ; is the energy ratio of the short and long time windows.

[0020] S3: Based on the seismic data, the initial model of underground medium wave impedance is constructed using the back-projection algorithm; In an exemplary embodiment, step S3 specifically includes: constructing an initial model of underground medium wave impedance based on seismic data using a back-projection algorithm, such as the formula: , , , , in, is the time required to traverse the grid; It represents the time when the seismic wave ray travels; Indicates the first The propagation distance of the root seismic wave ray; Indicates the total distance of the ray propagation path; is the total travel time required for all rays in the grid; is the total propagation distance required by all rays in the grid; is the average slowness value within the grid.

[0021] S4: Calculating travel time residuals based on the first arrival travel time and the theoretical travel time, and correcting the average slowness value in each grid using a joint iteration method based on the initial underground medium wave impedance model to obtain a corrected underground medium wave impedance model; In an exemplary embodiment, step S4 specifically includes: calculating the travel time residual based on the first arrival travel time and the theoretical travel time, and correcting the average slowness value in each grid using a joint iteration method based on the initial underground medium wave impedance model to obtain a corrected underground medium wave impedance model, such as the formula: , , in, Indicates the ray number; Indicates the grid number; For the The iteration model vector is numbered Grid; For the The iteration model vector is numbered Grid; is the total number of rays passing through grid j; is the relaxation factor; Number the ray currently being processed; is the total number of rays; is the time of the first arrival; for Theoretical travel time of a ray; For rays In the grid Path length within ; Number Grid; For Grid The effective ray coverage number; For Grid The average travel time residual.

[0022] S5: When the iteration termination condition is not met, return to step S4; when the iteration termination condition is met, stop the iteration and obtain the three-dimensional non-destructive detection results of the dam disease based on the modified underground medium wave impedance model; In an exemplary embodiment, the iteration termination condition includes: the difference between the travel time residual of the modified underground medium wave impedance model and the travel time residual of the initial underground medium wave impedance model is less than a preset difference; the iteration termination condition includes: In an exemplary embodiment, the iteration termination condition further includes: the update amount of the underground medium wave impedance model does not substantially change with an increase in the number of iterations.

[0023] In some embodiments, the above-mentioned three-dimensional non-destructive detection method for dam defects can also be implemented in the following manner.

[0024] In this embodiment, the three-dimensional non-destructive detection method for dam defects includes the following steps: Step 1: First, analyze the existing data and arrange appropriate survey lines at the detection location according to the actual detection needs, such as the distance, and set the measurement points according to the detection accuracy requirements; Figure 2 The figure shows the layout of the three-dimensional survey line of the dam. In this embodiment, the dam is surrounded by detectors to perform three-dimensional seismic imaging. When the seismic source is arranged on the left side of the dam, the detectors on the right side of the dam slope and the dam top receive seismic signals; when the seismic source is arranged on the right side of the dam slope, the detectors on the left side of the dam slope and the dam top receive seismic signals; when the seismic source is arranged on the dam top, the detectors on the dam slopes on both sides receive seismic signals.

[0025] In this way, the dam is surrounded on three sides. The side where the seismic source is arranged is similar to the "excitation well" of interwell seismic, and the two sides where the detectors are arranged are similar to the "receiving wells" of interwell seismic. This method can be used for three-dimensional seismic exploration without the need for drilling wells. It is a non-destructive detection method.

[0026] Then, 3D seismic technology is used to sequentially stimulate the seismic sources to collect seismic data, perform data processing, and finally interpret the generated seismic records.

[0027] Step 2: Data processing: The travel time data T is obtained by using ray and wave equation theory for given underground media. cal This is the forward problem, through the actual measured travel time data T obs The inversion problem is to determine the wave impedance coefficient of the medium in the region. The key issues involved in the inversion process include: accurate picking of the first arrival travel time, establishment of the initial model, determination of the tomographic sensitivity kernel matrix, solution of the travel time tomographic inversion equations, determination of the iterative termination conditions, and imaging of the inversion results. The iterative inversion solution process is as follows: (1) Extract the first arrival travel time and obtain the travel time observation data set T obs ; (2) Establish an initial model of underground medium wave impedance and forward calculate T cal ; (3) Calculate the travel time residual, modify the model parameters, and iterate until the accuracy requirements are met; (4) Output the inversion results to achieve three-dimensional imaging.

[0028] As can be seen from the Rodan transform, tomography requires as many observation angles as possible to ensure more accurate imaging results. However, this is difficult to achieve in actual seismic tomography. In this embodiment, the simulated cross-well tomography will adopt a two-side observation system or a three-side observation system, converting the traditional rectangular observation model into a trapezoidal observation system, such as Figure 3 Shown is a two-dimensional cross-sectional view of the dam body's simulated interwell tomography observation system.

[0029] Based on the cross-well seismic technology, this embodiment adopts the design concept of inclined well-ground-inclined well observation system on the dam surface, and performs cross-well tomography of the dam through different excitation and reception combinations, such as Figure 2 、 Figure 3 As shown. For the proposed inter-well observation system, the theoretical travel time and path are forward calculated based on the straight line or curved ray tracing method. The BPT reconstruction result is used as the initial value for iteration, and the inversion calculation is performed based on the SIRT algorithm to test the convergence and stability of the algorithm. Assume that the propagation time of the jth ray in the target area after transmission and reception is T j , the distance through the i-th grid is L i , the velocity of the corresponding grid is V i , slowness value is S i , then the travel time of the jth ray passing through the entire grid area is:

[0030] The travel time CT is ultimately written as a Kx1 matrix using the first arrival time data T of all rays. Since the distance values ​​of the grid cells passed by all rays are different, a distance matrix A is defined. Each row of the distance matrix A represents the distance value of a ray passing through all grid cells. The number of rows is the total number of rays, K. The total number of grid cells is H, and the slowness value is defined as a slowness matrix S. Therefore:

[0031] The forward problem is to obtain the travel time data T of a given underground medium through ray or wave equation theory, and the inversion problem is to obtain the velocity field of the medium in the area through the actual measured travel time data T.

[0032] 1. Extract the first arrival time: Geometric ray traveltime tomography primarily uses the traveltime of first arrivals to invert the slowness (or velocity) distribution of the exploration target area, and is therefore also called first arrival traveltime tomography. First arrival traveltime imaging is often used in actual engineering exploration, and this embodiment will adopt this imaging method. Seismic first arrival traveltime tomography uses the first arrival time values ​​picked from seismic records to invert the distribution of physical properties within the model. Typically, when processing seismic data, the first arrival time is determined based on changes in the amplitude, frequency, and phase of the seismic wave signal. Some techniques also use image segmentation or signal comparison to determine the first arrival. First arrival picking is a relatively mature technology in seismic signal processing, and many methods have been proposed, such as energy ratio methods, fractal dimension methods, and neural grid recognition. This embodiment improves on the conventional energy ratio method and establishes an automatic energy ratio first arrival picking method based on autocorrelation constraints.

[0033] Assume that the record length is still M, and the lengths of the short and long time windows are still N s and N l At some point i Short-time window autocorrelation R s,i,j for:

[0034] Long-term autocorrelation R l,i,j for:

[0035] Short-time window average energy E rs,i It can be expressed as:

[0036] Obviously, when the window is in the random noise segment:

[0037] Long-term window average energy E rl,i It can be expressed as:

[0038] Obviously, when the window is in the random noise segment:

[0039] Short-long time window energy ratio E rsl,i for:

[0040] This method has high accuracy and high noise resistance.

[0041] 2. Establishment of the initial model: The back projection technique (BPT) is a classic initial model construction method in seismic tomography. Its core concept is based on the inverse process of the Radon transform. This method gradually reconstructs the physical property distribution (such as velocity or slowness field) of the subsurface medium by back-projecting observed seismic wave travel time data onto the imaging area. This method is computationally efficient and suitable for rapid initialization of large datasets. However, due to its assumption of straight-line rays, it has low resolution and often requires subsequent correction in conjunction with iterative optimization algorithms (such as SIRT).

[0042] The back-projection algorithm regards the density value of a certain grid in the area as the average value of the sum of all ray projections passing through the grid in the area, which often represents the contribution of a certain grid. Its implementation process is to mesh the dam body and ensure that the grid size is smaller than the minimum scale of the abnormal body. Assume that the three-dimensional space is divided into , , pixel units, the total number of spatial grids is indivual.

[0043] First, a seismic source is excited on the top of a dam, and the geophone receives all longitudinal grid units in turn. After the measurement is completed, N ray data of a three-dimensional seismic device are obtained. Then, the number of rays K in the area is completed by exciting the source on both sides of the dam slope. Figure 4 The figure shows a schematic diagram of rectangular grid division; Generally, straight ray tracing is faster than curved ray tracing. Based on the initial uniform velocity model, the straight ray tracing method is used to calculate the path length matrix of each ray in each grid, ultimately achieving high-precision tomography. A seismic wave ray passes through a grid cell. Indicates the first The propagation distance of the root seismic wave ray, Represents the total distance of the ray propagation path, It means that when the seismic wave ray travels, if the ray does not pass through a certain pixel, then Assign a value of 0. Before applying BPT, a ray path calculation must be performed. The time required to pass through the grid is:

[0044] The total travel time and total propagation distance required for all rays in the grid are:

[0045] The average slowness value within this grid is:

[0046] This formula essentially distributes the traveltime error of each ray evenly across the pixels it passes through, thereby constructing a preliminary slowness distribution. For example, if a pixel is traversed by multiple rays with long traveltimes, its initial slowness value will be significantly increased (corresponding to a low-speed anomaly). For areas with sparse ray coverage (such as near dam slopes), numerical instability can be avoided through interpolation or constraints.

[0047] Finally, the slowness field is converted into a velocity field, and the generated velocity field is used as the input of the subsequent SIRT.

[0048] 3. Joint Iterative Method (SIRT) Inversion: The Simultaneous Iterative Reconstruction Technique (SIRT) was selected to correct the average slowness within each grid. SIRT has the advantages of low memory usage and fast computational speed, but its disadvantages include sensitivity to measurement errors, high dependence on initial values, and strict selection of relaxation factors, which can lead to non-convergence when the relaxation factor exceeds a certain range. SIRT, on the other hand, is insensitive to measurement errors, has relatively low initial value requirements, and exhibits good convergence, making it widely used in tomography.

[0049] The basic idea of ​​the SIRT algorithm is to first set up an initial model, then calculate the travel time errors of all rays passing through the grid cells in order, and use the travel time errors of all rays to correct the slowness value in the grid cells until the pre-set accuracy requirements are met.

[0050] The SIRT algorithm is characterized by using the correction values ​​of all rays passing through a cell to determine the average correction value for that cell. The SIRT algorithm completes an iteration only after all rays have passed through the cell, which can suppress interference factors to a certain extent, and the calculation results are independent of the order in which the projection data is used. The iterative correction formula of the SIRT algorithm can be written as:

[0051]

[0052] Where s q is the model vector for the qth iteration, μ is the relaxation factor, 0 < μ ≤ 2, the subscript i represents the ray number, and j represents the grid number. As can be seen from the above formula, the modified increment of the model vector is actually the weighted average of the modified increments of all rays passing through the jth grid.

[0053] Although the SIRT algorithm has the disadvantage of large memory usage, it has good stability and convergence and wide adaptability. Therefore, it is currently a commonly used and mature method in the actual work of tomography.

[0054] 4. Determination of travel time residuals and iterative termination conditions: Calculate the travel time residual of each ray measured travel time and forward calculated travel time, which is The conditions for iterative termination directly determine the calculation time and final inversion effect of tomographic inversion. The conditions for iterative termination generally follow the following two principles: ① The forward fitting data of the inversion model are basically consistent with the actual observation data, and the numerical value of the inverted travel time residuals differs from that of the initial model by one to two orders of magnitude.

[0055] ② The update amount of the velocity model basically does not change with the increase of the number of iterations, and the descending curve of the iteration error tends to be flat, indicating that the inversion results of multiple consecutive iterations are basically consistent. At this time, further iterations are of little significance for model modification, and it can be determined as the termination condition.

[0056] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the aforementioned method for three-dimensional nondestructive detection of dam defects. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD); the storage medium may also include a combination of the aforementioned types of memory.

[0057] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps of the above-mentioned three-dimensional non-destructive detection method for dam defects are implemented.

[0058] like Figure 5As shown, the computer device 120 may include: at least one processor 121, such as a central processing unit (CPU), at least one communication interface 123, a memory 124, and at least one communication bus 122. The communication bus 122 is used to enable communication between these components. The communication interface 123 may include a display and a keyboard. Optionally, the communication interface 123 may also include a standard wired interface or a wireless interface. The memory 124 may be a high-speed random access memory (RAM) or a non-volatile memory, such as at least one disk drive. The memory 124 may optionally be at least one storage device located remote from the processor 121. The memory 124 stores application programs, and the processor 121 invokes program code stored in the memory 124 to execute any of the aforementioned method steps. The communication bus 122 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, for example. The communication bus 122 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5The bus 123 is used to connect the above-mentioned elements in the system 120, and only one bus is represented, but it does not mean that there is only one bus or only one type of bus. Among them, the memory 124 can include volatile memory such as random-access memory (RAM); the memory can also include non-volatile memory such as flash memory, a hard disk drive (HDD) or a solid-state drive (SSD); the memory 124 can also include a combination of the above-mentioned types of memory. Among them, the processor 121 can be a central processing unit (CPU), a network processor (NP) or a combination of CPU and NP. The processor 121 can further include a hardware chip. The above-mentioned hardware chip can be an application-specific integrated circuit (ASIC), a programmable logic device (PLD) or a combination thereof. The above-mentioned PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof. Alternatively, the memory 124 is also used to store program instructions. The processor 121 can call the program instructions to realize the dam disease three-dimensional non-destructive detection method as in the embodiment.

[0059] The embodiments of the present application are described above in combination with the drawings, but the present application is not limited to the above-mentioned specific embodiments, and the above-mentioned specific embodiments are only illustrative and not restrictive. Those skilled in the art can make many forms under the inspiration of the present application without departing from the purpose of the present application and the scope protected by the claims, and these all belong to the protection of the present application.

Claims

1. A three-dimensional non-destructive detection method for dam defects, characterized by: The following steps are involved: S1: Use the source and detector to detect the target dam and obtain seismic data; S2: Based on the seismic data, the first arrival travel time is obtained using the improved autocorrelation constrained energy ratio method; S3: Based on the seismic data, the back-projection algorithm is used to construct an initial model of the underground medium wave impedance, and forward calculation is performed to obtain the theoretical travel time; S4: Calculating travel time residuals based on the first arrival travel time and the theoretical travel time, and correcting the average slowness value in each grid using a joint iteration method based on the initial underground medium wave impedance model to obtain a corrected underground medium wave impedance model; S5: When the iteration termination condition is not met, return to step S4; when the iteration termination condition is met, stop the iteration and obtain the three-dimensional non-destructive detection results of the dam disease based on the modified underground medium wave impedance model.

2. The three-dimensional non-destructive detection method for dam defects according to claim 1 is characterized in that: Step S2 specifically includes: obtaining the first arrival travel time based on the seismic data using the improved autocorrelation constrained energy ratio method, such as the formula: ， ， ， ， ， ， ， in, is the record length; For a certain moment Short-time window autocorrelation; is the length of the short-term window; Record earthquakes at a specific time The amplitude value at ; For earthquake records The amplitude value at Constructing signal segments within the time window; is the long-term window autocorrelation; is the length of the long time window; is the average energy in the short time window; For time point The short-time window autocorrelation function value at ; is the average energy of the long time window; For time point The long-time window autocorrelation function value at ; is the energy ratio of the short and long time windows.

3. The three-dimensional non-destructive detection method for dam defects according to claim 1 is characterized in that: Step S3 specifically includes: constructing an initial model of underground medium wave impedance based on seismic data using a back-projection algorithm, such as the formula: ， ， ， ， in, is the time required to traverse the grid; It represents the time when the seismic wave ray travels; Indicates the first The propagation distance of the root seismic wave ray; Indicates the total distance of the ray propagation path; is the total travel time required for all rays in the grid; is the total propagation distance required by all rays in the grid; is the average slowness value within the grid.

4. The three-dimensional non-destructive detection method for dam defects according to claim 1 is characterized in that: Step S4 specifically includes: calculating the travel time residual based on the first arrival travel time and the theoretical travel time, and correcting the average slowness value in each grid using the joint iteration method based on the initial underground medium wave impedance model to obtain a corrected underground medium wave impedance model, such as the formula: ， ， in, Indicates the ray number; Indicates the grid number; For the The iteration model vector is numbered Grid; For the The iteration model vector is numbered Grid; is the total number of rays passing through grid j; is the relaxation factor; Number the ray currently being processed; is the total number of rays; is the time of the first arrival; for Theoretical travel time of a ray; For rays In the grid Path length within ; Number Grid; For Grid The effective ray coverage number; For Grid The average travel time residual.

5. The three-dimensional non-destructive detection method for dam defects according to claim 1 is characterized in that: The iteration termination condition includes: the difference between the travel time residual of the modified underground medium wave impedance model and the travel time residual of the underground medium wave impedance initial model is less than a preset difference; the iteration termination condition includes.

6. The three-dimensional non-destructive detection method for dam defects according to claim 5 is characterized in that: The iteration termination condition also includes: the update amount of the underground medium wave impedance model does not change substantially as the number of iterations increases.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the three-dimensional non-destructive detection method for dam defects according to any one of claims 1 to 6 are implemented.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of the three-dimensional non-destructive detection method for dam defects are implemented as described in any one of claims 1 to 6.

Images