Ground tunnel electromagnetic joint detection method based on natural field

Abstract

The invention relates to the technical field of magnetotelluric sounding methods, in particular to a ground tunnel electromagnetic joint detection method based on a natural field. According to the method, the overall information is detected by using the ground magnetotelluric method, and the detection method combining the measurement of the local information in the tunnel is used, so that the accuracy and resolution of abnormal recognition in the tunnel engineering can be improved, and the inversion precision of the geoelectric structure near the tunnel can be effectively improved. Meanwhile, the measuring points are placed in the ground, electromagnetic field data of the ground are measured point by point, impedance data of ground tunnel electromagnetic field information are constructed in combination with the electromagnetic field data measured by the measuring points in the tunnel, and the purpose that the resolution ratio of ground MT measurement to local abnormal information is insufficient is improved. Meanwhile, the method improves the geophysical prospecting accuracy and resolution, reduces the workload of subsequent drilling, and saves the cost.

Description

Technical Field

[0001] The invention relates to the technical field of magnetotelluric sounding, in particular to a natural field-based ground tunnel electromagnetic combined detection method. Background Art

[0002] The natural field source used by the magnetotelluric method has an extremely wide frequency band range. Based on the skin depth of the electromagnetic field, when the alternating electromagnetic field propagates in the underground medium, electromagnetic waves of different frequencies correspond to different detection depths, which can obtain more abundant effective information on the underground electrical structure. At present, the magnetotelluric method is widely used in engineering, but with the increasingly complex exploration environment, the single ground magnetotelluric method can only detect the geophysical information of the underground as a whole, and the detection of some local anomalies is still not accurate enough. In particular, with the increase in tunnel construction in the mountainous areas of western my country, the geological conditions are complex, involving adverse geological phenomena such as faults, folds, landslides, karst, etc., which are prone to safety accidents. Therefore, it is crucial to fully grasp the geological conditions of tunnel engineering and take corresponding countermeasures.

[0003] At present, an electromagnetic joint inversion method is disclosed in the invention patent with the publication number CN111323830A, which is named "A joint inversion method based on magnetotelluric and DC resistivity data". The main steps include: determining the inversion area according to the observed data, dividing it to obtain the initial grid, and setting the initial value of the inversion parameter vector; performing Gauss-Newton inversion iteration based on the current value of the inversion parameter vector to calculate the model change and the linear search step, and then calculating the iterative update value of the inversion parameter vector; judging whether the inversion termination condition is met, if not, performing the next iteration; if met, judging whether the progressive grid inversion termination condition is met, if met, the inversion ends; if not, the grid unit is refined and the inversion parameter vector value is updated, and then the Gauss-Newton inversion iteration is returned until the inversion ends. The invention develops a joint inversion technology with adaptive adjustment of the inversion grid, which effectively reduces the multi-solution problem of the joint inversion interpretation of magnetotelluric and DC resistivity data and improves the accuracy of the joint interpretation.

[0004] In the invention patent with the publication number CN112578470A, named "Marine Controllable Source Electromagnetic and Magnetotelluric Joint Inversion Method Based on Product Function", an electromagnetic joint inversion method is disclosed, which mainly includes the following steps: reading and converting marine controlled source electromagnetic field and marine magnetotelluric data; setting joint inversion execution parameters; establishing a joint inversion initial model; constructing a joint inversion objective function; nonlinear conversion of inversion parameters; obtaining the Jacobian matrix and Hessian matrix of the inversion data about the anisotropic resistivity and thickness of the formation; calculating the model parameter update amount based on the inversion parameters; searching for the inversion update amount step; calculating the inversion iteration fitting difference; judging whether the inversion requirements are met; and outputting the final inversion model. This method simultaneously inverts the anisotropic resistivity and thickness of the submarine formation, and has a higher inversion efficiency; at the same time, it integrates the high resolution of marine controlled source electromagnetic data for shallow formations and the detection capability of magnetotelluric data for deep structures, and the inversion effect is better than the traditional single electromagnetic data inversion; the proposed joint inversion method has a wider applicability.

[0005] However, in the above two joint inversion invention patents, one is a joint inversion method that combines shallow information of DC resistivity data with magnetotelluric information. This invention still cannot accurately detect local information. The other is a joint inversion method of marine controlled source electromagnetic and magnetotelluric, and the object of this invention is marine exploration. Both inventions cannot be applied to land tunnel projects.

[0006] Therefore, a joint inversion method with high processing accuracy, high resolution, and practicability is needed, which can be applied to the field of tunnel engineering to explore underground anomalies. Summary of the invention

[0007] The purpose of the present invention is to overcome the above-mentioned deficiencies in the prior art and to provide a natural field-based ground tunnel electromagnetic combined detection method.

[0008] In order to achieve the above-mentioned object of the invention, the present invention provides the following technical solutions:

[0009] A natural field-based ground tunnel electromagnetic combined detection method comprises the following steps:

[0010] S1: Construct a three-dimensional geoelectric model of the tunnel in the area to be detected;

[0011] S2: Setting fixed collection points on the ground and in the tunnel of the area to be detected, collecting electric field data and magnetic field data, and generating impedance data according to the electric field data and the magnetic field data;

[0012] S3: constructing an underground tunnel electromagnetic field observation file, setting forward modeling parameters according to the underground tunnel electromagnetic field observation file, and performing finite difference forward modeling on the underground tunnel three-dimensional model; the underground tunnel electromagnetic field observation file includes frequency point information and receiving point coordinate information;

[0013] S4: constructing a three-dimensional regularized inversion objective function according to the impedance data and the finite difference forward modeling results;

[0014] S5: construct the initial model of uniform half space;

[0015] S6: Invert and iterate the uniform half-space initial model through the three-dimensional regularized inversion objective function, and output the inversion result as the detection result of the area to be detected.

[0016] As a preferred solution of the present invention, the three-dimensional model of the underground tunnel adopts a geoelectric model, and the three-dimensional grid of the geoelectric model is refined and evenly divided in the central survey area in the XY direction, divided into two parts according to the target survey area and the grid extension area, and the depth Z direction is subdivided in the shallow layer. At the same time, the underground anomaly is placed within the target area, and the data collection area is in the form of a uniformly divided grid, and the interval distance of the extension grid is increased by multiples.

[0017] As a preferred solution of the present invention, the setting of fixed collection points in S2 includes: evenly setting fixed collection points at preset intervals on the ground and in tunnels within the area to be detected; the calculation formula for the impedance data is:

[0018]

[0019]

[0020] Among them, Z xx , Z xy , Z yx , Z yy are the impedance components in the xx, xy, yx, and yy directions respectively, and are the electric field data in the x direction of the ground; E x2 is the electric field data in the x direction of the tunnel; E y1 is the electric field data in the y direction of the ground; E y2 is the electric field data in the y direction of the tunnel; H x1 is the ground x-direction magnetic field data, H x2 is the magnetic field data in the x direction of the tunnel, H y1 is the ground y-direction magnetic field data, H y2 is the magnetic field data in the y direction of the tunnel.

[0021] As a preferred solution of the present invention, the finite difference forward modeling in S3 adopts staggered grid finite difference forward modeling, and its expression is:

[0022] KE=b,

[0023] K is a large sparse matrix, including the electromagnetic properties of the underground medium and the propagation law of the electromagnetic field; E is the column vector of the three components of the electric field to be determined at the sampling point; b is the discretized right-hand side term, including the information of the ground and tunnel electromagnetic fields and the boundary conditions.

[0024] As a preferred embodiment of the present invention, S4 comprises the following steps:

[0025] S41: Establish three-dimensional regularized inversion objective function;

[0026] S42: deriving the three-dimensional regularized inversion objective function to calculate the objective function gradient;

[0027] S43: solving the objective function gradient by quasi-forward modeling; the quasi-forward modeling is to repeatedly perform the finite difference forward modeling operation until the fitting error meets the preset value;

[0028] S44: Output the current three-dimensional regularized inversion objective function.

[0029] As a preferred embodiment of the present invention, the expression of the three-dimensional regularized inversion objective function is:

[0030]

[0031] Where d is the difference between the collected electromagnetic data and the electromagnetic data calculated by forward modeling, λ is the regularization factor, and are the data covariance matrix and the model covariance matrix respectively, m and m0 are the model parameters and prior model of resistivity respectively.

[0032] As a preferred solution of the present invention, the expression for calculating the gradient of the objective function by derivation in S42 is:

[0033]

[0034] in, is through (Δd) T The derivative is obtained, J is the sensitivity matrix, and m0 is the prior model vector.

[0035] As a preferred embodiment of the present invention, the uniform half-space initial model in S5 is a combination of resistivity data at different positions in the three-dimensional space, including an upper half-space and a lower half-space;

[0036] The upper half space is a resistivity model of air, and the lower half space is a resistivity model of underground medium resistivity being a set value.

[0037] As a preferred solution of the present invention, S6 uses the quasi-Newton method or NLCG to perform inversion iteration on the uniform half-space initial model; its termination condition is to meet a preset maximum number of iterations or a minimum fitting difference parameter condition.

[0038] A natural field-based ground-tunnel electromagnetic combined detection device comprises at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any of the methods described above.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] The present invention uses the ground magnetotelluric method to detect overall information, and uses a detection method that combines local information measurement in the tunnel, which can improve the accuracy and resolution of anomaly identification in tunnel engineering, and can effectively improve the inversion accuracy of detecting geoelectric structures near the tunnel. At the same time, the present invention places measuring points on the ground, measures the electromagnetic field data of the ground point by point, combines the electromagnetic field data measured by the measuring points in the tunnel, and jointly constructs the impedance data of the electromagnetic field information of the tunnel, thereby improving the purpose of insufficient resolution of local anomaly information in ground MT measurement. At the same time, this method improves the accuracy and resolution of geophysical exploration, which will reduce the workload of subsequent drilling, thereby saving costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of a method for combined electromagnetic detection of underground tunnels based on natural fields as described in Example 1 of the present invention;

[0042] Figure 2 This is a flow chart of a method for combined electromagnetic detection of underground tunnels based on natural fields as described in Example 3 of the present invention;

[0043] Figure 3 A schematic diagram of a three-dimensional initial model grid partitioning and slicing in a natural field-based ground tunnel electromagnetic joint detection method described in Example 3 of the present invention;

[0044] Figure 4 This is a schematic diagram of the observation system of Example 1 and Example 2 in a natural field-based ground tunnel electromagnetic joint detection method described in Example 3 of the present invention;

[0045] Figure 5 The forward response diagrams of Examples 1 and 2 in a ground tunnel electromagnetic joint detection method based on a natural field described in Example 3 of the present invention (the first row is Example 1, and the second row is Example 2);

[0046] Figure 6The three-dimensional inversion effect diagrams of Examples 1 and 2 in a ground tunnel electromagnetic joint detection method based on a natural field described in Example 3 of the present invention (a is Example 1, b is Example 2, and the red frame is the real model frame);

[0047] Figure 7 A three-dimensional inversion iterative fitting difference curve diagram provided in Examples 1 and 2 of a ground tunnel electromagnetic joint detection method based on a natural field described in Example 3 of the present invention;

[0048] Figure 8 A three-dimensional inversion iteration step length curve diagram provided in Examples 1 and 2 of a ground tunnel electromagnetic joint detection method based on a natural field described in Example 3 of the present invention;

[0049] Fig. 9 This is a schematic structural diagram of a natural field-based ground tunnel electromagnetic combined detection device described in Example 4 of the present invention, which utilizes the natural field-based ground tunnel electromagnetic combined detection method described in Example 1. DETAILED DESCRIPTION

[0050] The present invention is further described in detail below in conjunction with test examples and specific implementation methods. However, this should not be understood as the scope of the above subject matter of the present invention being limited to the following embodiments, and all technologies realized based on the content of the present invention belong to the scope of the present invention.

[0051] Example 1

[0052] like Figure 1 As shown, a method for electromagnetic combined detection of underground tunnels based on natural fields comprises the following steps:

[0053] S1: Construct a three-dimensional model of the underground tunnel in the area to be detected.

[0054] S2: Setting fixed collection points on the ground and in the tunnel of the area to be detected, collecting electric field data and magnetic field data, and generating impedance data according to the electric field data and the magnetic field data.

[0055] S3: constructing an underground tunnel electromagnetic field observation file, setting forward modeling parameters according to the underground tunnel electromagnetic field observation file, and performing finite difference forward modeling on the underground tunnel three-dimensional model; the underground tunnel electromagnetic field observation file includes frequency point information and receiving point coordinate information.

[0056] S4: Constructing a three-dimensional regularized inversion objective function according to the impedance data and the finite difference forward modeling results.

[0057] S5: Construct an initial uniform half-space model.

[0058] S6: Invert and iterate the uniform half-space initial model through the three-dimensional regularized inversion objective function, and output the inversion result as the detection result of the area to be detected.

[0059] Example 2

[0060] This embodiment is a specific implementation of the natural field-based ground tunnel electromagnetic joint detection method described in Embodiment 1, and includes the following steps:

[0061] S1: Construct a three-dimensional model of the underground tunnel in the area to be detected.

[0062] The three-dimensional model of the underground tunnel adopts the geoelectric model. The present invention constructs the three-dimensional model of the underground tunnel according to the terrain environment to be studied. In order to reduce the error caused by the boundary being too close to the measuring point, the three-dimensional grid of the present invention is refined and evenly divided in the central measuring area in the XY direction, divided into two parts according to the target measuring area and the grid extension area, and the depth Z direction is subdivided in the shallow layer. The underground anomaly is placed within the target area, and the data collection area is used in the form of a uniformly divided grid, and the extension grid increases the interval distance by multiples. The spatial position information of the relevant anomaly and the resistivity assignment are set to construct the initial three-dimensional model.

[0063] S2: evenly setting fixed collection points at preset intervals on the ground and in the tunnel of the area to be detected, collecting electric field data and magnetic field data, and generating impedance data according to the electric field data and the magnetic field data.

[0064] Furthermore, in this embodiment, the ground measuring points are arranged at intervals of 25m in the area to be detected, and the measuring area size is 2000m×2000m; the measuring points in the tunnel are arranged at intervals of 25m in the range of -1000m to 0m in the X direction at a depth of z=500m. The calculation formula of the impedance data is:

[0065]

[0066] Among them, Z xx , Z xy , Z yx , Z yy are the impedance components in the xx, xy, yx, and yy directions respectively, and E x1 is the electric field data in the x direction of the ground; E x2 is the electric field data in the x direction of the tunnel; E y1 is the electric field data in the y direction of the ground; E y2 is the electric field data in the y direction of the tunnel; H x1 is the ground x-direction magnetic field data, H x2 is the magnetic field data in the x direction of the tunnel, H y1 is the ground y-direction magnetic field data, H y2 is the magnetic field data in the y direction of the tunnel.

[0067] S3: constructing an underground tunnel electromagnetic field observation file, setting forward modeling parameters according to the underground tunnel electromagnetic field observation file, and performing finite difference forward modeling on the underground tunnel three-dimensional model; the underground tunnel electromagnetic field observation file includes frequency point information and receiving point coordinate information.

[0068] S31: Constructing underground tunnel electromagnetic field observation files: Selecting frequencies of 10 Hz, 100 Hz, and 500 Hz to arrange observation files, and setting up two models: pseudo-fault and pseudo-water-rich zone.

[0069] S32: Perform finite difference forward modeling on the three-dimensional model of the underground tunnel.

[0070] Furthermore, the finite difference forward modeling described in this embodiment adopts staggered grid finite difference forward modeling. Specifically, when performing finite difference forward modeling calculation on the initial model, the displacement current is ignored and the time harmonic factor is taken as e -iωt , the differential form of Maxwell's equations is:

[0071]

[0072]

[0073] Where σ is the conductivity, μ0 is the magnetic permeability of vacuum, and ω is the angular frequency.

[0074] Furthermore, by simply substituting the two equations, we can obtain the relationship based on the electric field strength E:

[0075]

[0076] Furthermore, the staggered grid is used to discretize the study area and obtain the forward equation of the electric field at the grid unit sampling point:

[0077] KE=b,

[0078] Among them, K is a large sparse matrix, which contains the electromagnetic properties of the underground medium (such as conductivity, magnetic permeability, etc.) and the propagation law of the electromagnetic field; E is the column vector of the three components of the electric field to be calculated at the sampling point, which represents the distribution of the electric field in the underground medium; b is the discretized right-hand side term, which contains information about the electromagnetic field of the ground and tunnel and boundary conditions. The above formula is solved by the QMR method.

[0079] S4: Constructing a three-dimensional regularized inversion objective function according to the impedance data and the finite difference forward modeling results.

[0080] S41: Establish three-dimensional regularized inversion objective function;

[0081] By adding model term constraints to the objective function through regularized inversion, the problem of multiple solutions of inversion is reduced; the expression of the three-dimensional regularized inversion objective function is:

[0082]

[0083] Where d is the difference between the collected electromagnetic data and the electromagnetic data calculated by forward modeling, λ is the regularization factor, and are the data covariance matrix and the model covariance matrix respectively, m and m0 are the model parameters and prior model of resistivity respectively.

[0084] S42: Deriving the three-dimensional regularized inversion objective function to calculate the objective function gradient; the expression is:

[0085]

[0086] in, is through (Δd) T Obtained by differentiation, m0 is the prior model vector, and J is the sensitivity matrix.

[0087] S43: solving the objective function gradient by quasi-forward modeling; the quasi-forward modeling is to repeatedly perform the finite difference forward modeling operation until the fitting error meets the preset value;

[0088] S44: Output the current three-dimensional regularized inversion objective function.

[0089] S5: Construct an initial uniform half-space model.

[0090] The uniform half-space initial model is a combination of resistivity data at different positions in the three-dimensional space, including an upper half-space and a lower half-space;

[0091] The upper half space is a resistivity model of air, and the lower half space is a resistivity model of underground medium resistivity being a set value.

[0092] S6: Invert and iterate the uniform half-space initial model through the three-dimensional regularized inversion objective function, and output the inversion result as the detection result of the area to be detected.

[0093] Furthermore, this embodiment uses the quasi-Newton method or the NLCG (NonLinear Conjugate Gradients) method to perform inversion iteration on the uniform half-space initial model; the termination condition is to meet the preset maximum number of iterations or minimum fitting error parameter condition.

[0094] Example 3

[0095] This embodiment is a specific experimental example of the electromagnetic combined detection method of underground tunnels based on natural fields described in Embodiment 2. Figure 2 As shown, the following steps are included:

[0096] Step 1: Create an initial 3D model:

[0097] In order to reduce the error caused by the boundary being too close to the measurement point, the three-dimensional grid of the present invention is refined and evenly divided in the central measurement area in the XY direction, divided into two parts according to the target measurement area and the grid extension area, and the depth Z direction is subdivided in the shallow layer. The underground anomaly is placed in the target area and used as the data collection area in the form of a uniformly divided grid, and the extension grid increases the interval distance by multiples. The spatial position information of the relevant anomaly and the resistivity assignment are set to construct the initial three-dimensional model.

[0098] Further, in the embodiment, Figure 3 The figure shows a partial schematic diagram of the segmentation slices. The center point of the top surface of the grid is set to the midpoint (0,0,0). The target XY measurement area is evenly segmented at intervals of 25m. The depth Z direction is segmented according to the measurement point position. The epitaxial grid increases in multiples, and the background resistivity is set to 500Ω·m. The underground simulated fault is placed at a depth of 425m-500m underground, and the resistivity value is set to 10Ω·m; and a series of strata with different resistance values ​​are set.

[0099] Step 2: Set the measuring points on the ground and in the tunnel:

[0100] The ground measuring points are arranged at intervals of 25m within the measuring area, and the measuring area size is 2000m×2000m; the measuring points in the tunnel are arranged at intervals of 25m at a depth of z=500m and in the range of -1000m to 0m in the X direction.

[0101] Step 3: Construct the underground tunnel electromagnetic field observation file:

[0102] In the embodiment, the frequencies of 10 Hz, 100 Hz, and 500 Hz are selected to arrange the observation files, and two models of pseudo-fault and pseudo-water-rich zone are set, such as Figure 4 An example observation system is shown.

[0103] Step 4: Forward modeling of the constructed three-dimensional geoelectric model using staggered grid finite difference

[0104] Furthermore, the measured magnetic field data is calculated to obtain the impedance component:

[0105]

[0106] When the method of the present invention performs finite difference forward calculation on the initial model, the displacement current is ignored and the time harmonic factor is taken as e -iωt , the differential form of Maxwell's equations is:

[0107]

[0108] Where σ is the conductivity, μ0 is the vacuum permeability, and ω is the angular frequency

[0109] Furthermore, by simply substituting the two equations, we can obtain the relationship based on the electric field strength E:

[0110]

[0111] Furthermore, the staggered grid is used to discretize the study area and obtain the forward equation of the electric field at the grid unit sampling point:

[0112] KE=b,

[0113] K is a large sparse matrix; E is the column vector of the three components of the electric field to be calculated at the sampling point; b is the discretized right-hand side term. The above equation is solved by the QMR method. Figure 5 As shown, it is the forward response diagram of an embodiment of the present invention (the first row is Example 1, and the second row is Example 2).

[0114] Step 5: Establish the regularized inversion objective function and calculate the gradient.

[0115] In Example 2, the model term constraint on the objective function is added by regularizing the inversion to reduce the problem of multiple solutions of the inversion.

[0116]

[0117] is a weighted matrix. Its purpose is to effectively reduce the weight of data with poor accuracy when the accuracy of the observed data is inconsistent. is the constraint matrix of the model

[0118] Furthermore, by taking the derivative of the above formula, the objective function gradient is calculated:

[0119]

[0120] In the formula, m0 is the prior model vector and J is the sensitivity matrix.

[0121] In Example 2, a priori model vectors are added to the initial model, and the inversion parameters are constrained in a logarithmic manner, so that the calculated geophysical parameters are all positive numbers, thereby improving the sensitivity of the inversion.

[0122] Step 6: Invert and iterate the target geological body to obtain a new result model:

[0123] Construct a uniform half-space initial model, combine it with the forward modeling data, and use the quasi-Newton method or NLCG to iterate the model until the maximum number of iterations or the minimum fitting error parameter conditions are met to obtain the final inversion result.

[0124] Further, in the embodiment, the given minimum fitting difference is 1.0, and the RMS formula is expressed as follows:

[0125]

[0126] If it is an inversion iteration of measured data, the data may be affected by many aspects, such as noise, human factors, etc. In the embodiment, a 2% random error is added to simulate the measured data.

[0127] like Figure 6 As shown, it is a three-dimensional inversion result diagram of an embodiment of the present invention. It can be seen from the figure that in terms of inversion effect, the ground-tunnel joint inversion method is more obvious in anomaly recognition than the single ground inversion method.

[0128] like Figure 7 As shown, it is a fitting difference curve diagram of the three-dimensional inversion of the embodiment of the technical solution of the present invention, Figure 8 This is a step length curve diagram of the three-dimensional inversion of the implementation example of the technical solution of the present invention. It can be seen that the inversion iterations of different implementation examples all converge stably, indicating the correctness of the technical solution of the present invention based on the joint detection method of the ground magnetotelluric field and the tunnel electromagnetic field.

[0129] Example 4

[0130] like Fig. 9 As shown, a ground-tunnel electromagnetic combined detection device based on natural field includes at least one processor, a memory connected to the at least one processor in communication, and at least one input-output interface connected to the at least one processor in communication; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the ground-tunnel electromagnetic combined detection method based on natural field described in the above embodiment. The input-output interface may include a display, a keyboard, a mouse, and a USB interface for inputting and outputting data.

[0131] Those skilled in the art can understand that: all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM), disks or optical disks, etc. Various media that can store program codes.

[0132] When the above-mentioned integrated unit of the present invention is implemented in the form of a software functional unit and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROMs, magnetic disks, or optical disks.

[0133] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for electromagnetic joint detection of underground tunnels based on natural fields, characterized in that: The following steps are involved: S1: Construct a three-dimensional geoelectric model of the tunnel in the area to be detected; S2: Setting fixed collection points on the ground and in the tunnel of the area to be detected, collecting electric field data and magnetic field data, and generating impedance data according to the electric field data and the magnetic field data; S3: constructing an underground tunnel electromagnetic field observation file, setting forward modeling parameters according to the underground tunnel electromagnetic field observation file, and performing finite difference forward modeling on the underground tunnel three-dimensional model; the underground tunnel electromagnetic field observation file includes frequency point information and receiving point coordinate information; S4: constructing a three-dimensional regularized inversion objective function according to the impedance data and the finite difference forward modeling results; S5: construct the initial model of uniform half space; S6: Invert and iterate the uniform half-space initial model through the three-dimensional regularized inversion objective function, and output the inversion result as the detection result of the area to be detected.

2. The method for electromagnetic combined detection of underground tunnels based on natural fields according to claim 1 is characterized in that: The three-dimensional model of the underground tunnel adopts a geoelectric model, and the three-dimensional grid of the geoelectric model is refined and evenly divided in the central measurement area in the XY direction, divided into two parts according to the target measurement area and the grid extension area, and the depth Z direction is subdivided in the shallow layer. At the same time, the underground anomaly is placed in the target area, and the data collection area is in the form of a uniformly divided grid, and the interval distance of the extension grid is increased by multiples.

3. The method for electromagnetic combined detection of underground tunnels based on natural fields according to claim 1 is characterized in that: The setting of fixed collection points in S2 includes: evenly setting fixed collection points at preset intervals on the ground and in tunnels within the area to be detected; the calculation formula for the impedance data is: Among them, Z xx , Z xy , Z yx , Z yy are the impedance components in the xx, xy, yx, and yy directions respectively, and E x1 is the electric field data in the x direction of the ground; E x2 is the electric field data in the x direction of the tunnel; E y1 is the electric field data in the y direction of the ground; E y2 is the electric field data in the y direction of the tunnel; H x1 is the ground x-direction magnetic field data, H x2 is the magnetic field data in the x direction of the tunnel, H y1 is the ground y-direction magnetic field data, H y2 is the magnetic field data in the y direction of the tunnel.

4. The method for electromagnetic combined detection of underground tunnels based on natural fields according to claim 1 is characterized in that: The finite difference forward modeling in S3 adopts staggered grid finite difference forward modeling, and its expression is: KE=b, K is a large sparse matrix, including the electromagnetic properties of the underground medium and the propagation law of the electromagnetic field; E is the column vector of the three components of the electric field to be determined at the sampling point; b is the discretized right-hand side term, including the information of the ground and tunnel electromagnetic fields and the boundary conditions.

5. The method for electromagnetic combined detection of underground tunnels based on natural fields according to claim 4 is characterized in that: The S4 comprises the following steps: S41: Establish three-dimensional regularized inversion objective function; S42: deriving the three-dimensional regularized inversion objective function to calculate the objective function gradient; S43: solving the objective function gradient by quasi-forward modeling; the quasi-forward modeling is to repeatedly perform the finite difference forward modeling operation until the fitting error meets the preset value; S44: Output the current three-dimensional regularized inversion objective function.

6. The method for electromagnetic combined detection of underground tunnels based on natural fields according to claim 5 is characterized in that: The expression of the three-dimensional regularized inversion objective function is: Among them, d is the difference between the collected electromagnetic data and the electromagnetic data calculated by forward modeling, λ is the regularization factor, C-d1 and C-m1 are the data covariance matrix and the model covariance matrix respectively, and m and m0 are the model parameters and prior model of resistivity respectively.

7. The method for electromagnetic combined detection of underground tunnels based on natural fields according to claim 6 is characterized in that: The expression for calculating the gradient of the objective function by derivation in S42 is: in, is through (Δd) T The derivative is obtained, J is the sensitivity matrix, and m0 is the prior model vector.

8. The method for electromagnetic combined detection of underground tunnels based on natural fields according to claim 5 is characterized in that: The uniform half-space initial model in S5 is a combination of resistivity data at different positions in the three-dimensional space, including an upper half-space and a lower half-space; The upper half space is a resistivity model of air, and the lower half space is a resistivity model of underground medium resistivity being a set value.

9. The method for electromagnetic combined detection of underground tunnels based on natural fields according to claim 8, characterized in that: The S6 uses the quasi-Newton method or NLCG to perform inversion iteration on the uniform half-space initial model; the termination condition is that a preset maximum number of iterations or a minimum fitting error parameter condition is satisfied.

10. A natural field-based ground tunnel electromagnetic combined detection device, characterized in that: It includes at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method described in any one of claims 1 to 9.

Images