CN109946370A - A method and device for detecting leakage channel of embankment based on magnetic field measurement - Google Patents

Abstract

The invention discloses a kind of dam leakage passage detection method and its device based on magnetic-field measurement, this method comprises: S1: obtaining total magnetic flux density vector when background magnetic flux density vector and the power supply when signal transmitter is not powered on dykes and dams in each magnetic observation point in each magnetic observation point；Emission electrode A and B are laid in the two sides of dykes and dams, the emission electrode A is located in reservoir and contacts with water, connection signal transmitter between the emission electrode A and B, mentions on dam according to default line-spacing and point described away from arrangement magnetic observation point；S2: the difference for calculating separately total magnetic flux density vector and background magnetic flux density vector in each magnetic observation point obtains the magnetic anomaly vector of each magnetic observation point；S3: the magnetic anomaly vector based on magnetic observation point each in magnetic anomaly field deduces out current density distribution, wherein current density distribution is not zero place for leakage passage.Completely new method of the invention a kind of realizes that leakage passage detects, and uses Magnetic testi for the first time.

Description

A method and device for detecting leakage channel of embankment based on magnetic field measurement

technical field

The invention belongs to the technical field of exploration geophysics, and in particular relates to a method and a device for detecting a leakage channel of a dam based on magnetic field measurement.

Background technique

my country is a country prone to floods, and the losses caused by flood disasters are huge. Dam collapse or dyke burst is an important cause of flood disasters, and dam leakage is the direct cause of dam collapse or dyke burst. In the 1990s, Central South University took the lead in developing a leakage detector for dams and dams based on the theory of the flow field method, and produced stereotyped products for promotion and application, and achieved good results. The current dam piping leak meters are used to establish an AC electric field between the outlet of the leakage channel and the water in front of the dam, and use the location where the local current density or the electric field strength increases singularly to determine the location of the dam leakage inlet. Existing methods all measure the current density or electric field strength in water to find the location of the leakage inlet of the dam. Besides determining the location of the leakage inlet, how to determine the leakage channel is an important aspect of dam detection. However, so far, no one has used the magnetic field measurement method to determine the location of the dam piping channel, and the magnetic field measurement is fast, easy to operate, and low cost. The rapid detection of leakage channels provides a means, which is worthy of further in-depth study.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method and device for detecting the leakage channel of a dam based on magnetic field measurement, which for the first time realizes the use of the magnetic field measurement method to determine the position of the dam piping channel, provides a brand-new detection method, and efficiently locates the interior of the dam Leakage channels provide technical means for the protection and monitoring of dams.

On the one hand, the present invention provides a method and device for detecting leakage channel of a dam based on magnetic field measurement, comprising the following steps:

S1: Obtain the background magnetic induction intensity vector at each magnetic observation point on the dam when the signal transmitter is not powered and the total magnetic induction intensity vector at each magnetic observation point when the power is supplied;

Wherein, transmitting electrodes A and B are arranged on both sides of the dam, the transmitting electrode A is located in the reservoir and is in contact with the water, and a signal transmitter is connected between the transmitting electrodes A and B, on the dam according to the preset Line spacing and point spacing to arrange magnetic observation points;

S2: Calculate the difference between the total magnetic induction intensity vector and the background magnetic induction intensity vector at each magnetic observation point to obtain the magnetic anomaly vector of each magnetic observation point;

Among them, the magnetic anomaly vector of each magnetic observation point constitutes the magnetic anomaly field of Tiba;

S3: The current density distribution is deduced based on the magnetic anomaly vector of each magnetic observation point in the magnetic anomaly field;

Among them, the location where the current density distribution is not zero corresponds to the location of the leakage channel on the dam.

When a current signal is injected through the emitter electrodes A and B, when the current is guided along the preferential conductive path generated by the leakage channel of the dam, the current in the leakage channel will generate a magnetic anomaly. The location of the leak channel. It is not difficult to understand that a water flow path will definitely be formed in the seepage channel due to the flow of water, and its conductivity will be significantly higher than that of the background (non-aqueous substances such as rock, clay or cement). The water flow is associated with the flow path, while the conductive current path is determined by the magnetic field distribution. Moreover, the present invention also verifies the gradient sensitivity of the magnetic induction intensity vector to the electrical conductivity through formula reasoning, so the characteristics of the flow field generated by the charge flow in the leakage channel of the dam have a strong decrease relative to the electrical conductivity of the surrounding uniform material, can be detected by magnetic measurements.

Therefore, the present invention obtains the magnetic anomaly vector by first obtaining the background magnetic induction intensity vector and the total magnetic induction intensity vector obtained after injecting current signals through the A and B electrodes, and then obtains the current density distribution based on the magnetic anomaly vector, that is, at the location of the leakage channel, The current density distribution obtained based on the magnetic anomaly vector will not be zero, that is, the current density distribution generated by the shrinkage of the leakage channel can be obtained. At the location of the non-leakage channel (the background of non-aqueous materials such as rock, clay or cement), the current density distribution obtained based on the magnetic anomaly vector is zero. It should be understood that the current density distribution obtained in the present invention does not refer to the current density distribution of the total magnetic anomaly vector on the dam, but to the current density distribution generated by the magnetic anomaly vector on the dam.

Further preferably, step S3 is to deduce the current density distribution based on the relationship between the magnetic anomaly field and the current density field;

Among them, the three-dimensional space of Tiba is discretized into M units, each unit has the same volume and there is a constant current density in each unit, the relationship between the magnetic anomaly field and the current density field is as follows:

B _s = KJ

In the formula, B _s is the magnetic anomaly field matrix, the size of the matrix is 3N×1, are the magnetic anomaly vectors at the first and Nth magnetic observation points respectively, N is the number of magnetic observation points; K is the position matrix, and the size of the matrix is 3N×3M, are the distance matrices between the first magnetic observation point and the first and Mth units, respectively, are the distance matrices between the Nth magnetic observation point and the first unit and the Mth unit respectively; J is the current density field matrix, the size of the matrix is 3M×1, J ¹ , J ^M are the first unit, The current density on the Mth cell;

Among them, the magnetic anomaly vector of any i-th magnetic observation point Distance matrix between the i-th magnetic observation point and the k-th cell The current density Jk on the ^kth cell is as follows:

In the formula, magnetic anomaly vector The components in the x, y, and z directions in the spatial Cartesian coordinate system xyz, μ is the magnetic permeability of free space, ν is the unit volume, r _ik is the space between the i-th magnetic observation point and the k-th unit distance, are the spatial distance between the i-th magnetic observation point and the k-th unit, respectively The components in the x, y, and z directions in the space Cartesian coordinate system xyz; are the components of the current density J ^k on the kth unit in the x, y, and z directions in the spatial Cartesian coordinate system xyz, respectively.

Further preferably, in step S3, based on the relationship between the magnetic anomaly field and the current density field, the characteristic point method of the magnetic anomaly curve, the tangent method, the two-dimensional inversion method or the linear three-dimensional inversion method is used to deduce the current density distribution.

Based on the relationship between the magnetic anomaly field and the current density field, various algorithms are used to solve the current density distribution, and the inversion method also includes but is not limited to Gauss-Newton method, nonlinear conjugate gradient method, and genetic algorithm.

Further preferably, the process of deriving the current density distribution using the linear three-dimensional inversion method is as follows:

First, establish the objective function Φ(J) according to the principle of the least multiplication algorithm and obtain the solution of the linear inverse problem of the objective function;

Among them, the objective function is as follows:

Φ(J)=[W _d (d _obs -KJ)] ^T [W _d (d _obs -KJ)]+λ[W _m (JJ ₀ )] ^T [W _m (JJ ₀ )]

The formula for the solution is as follows:

In the formula, λ is the regularization factor, J ₀ is the prior model of the current density, d _obs is the observation data, W _d and W _m are weight matrices, and the weight matrix W _d is the error covariance matrix of the observation data d _obs . A diagonal matrix composed of reciprocals, the weight matrix W _m is a smoothness matrix, also known as a weight coefficient matrix, usually a unit matrix I _m ;

Then, the magnetic anomaly field matrix and the position matrix on the dam are substituted into the formula of the solution to calculate the current density field matrix to obtain the current density distribution on the dam;

Among them, the magnetic anomaly field matrix B _s is the observation data d _obs .

Further preferably, a magnetic probe is used to collect the magnetic induction intensity vector at the observation point.

On the other hand, a device based on the above method provided by the present invention includes: transmitting electrodes A and B, a signal transmitter, a receiver, and an analysis processor;

Wherein, the transmitting electrodes A and B are respectively arranged on both sides of the dam, the transmitting electrode A is located in the reservoir and is in contact with the water, and the signal transmitter is connected between the transmitting electrodes A and B; on the dam The magnetic observation points are arranged according to the preset line distance and point distance, and the receiver is arranged at the magnetic observation point to collect the magnetic induction intensity vector;

The analysis processor is connected with the receiver, and is used for acquiring the background magnetic induction intensity vector at each magnetic observation point collected by the receiver when the signal transmitter is not powered and each magnetic observation collected when the signal transmitter is powered on The total magnetic induction intensity vector at the point, and identify the location of the leakage channel on the dam.

Further preferably, the receiver is a magnetic probe or a magnetometer.

Among them, when the receiver is a magnetic probe, it does not need to be grounded, which overcomes the difficulty of poor grounding conditions of the dam.

Further preferably, the transmitting electrode B is located on land.

Further preferably, the transmission signal of the signal transmitter is a DC signal or an AC signal with a frequency of 0.1-380 Hz.

beneficial effect

The invention provides that the gradient sensitivity of the magnetic induction intensity vector to the conductivity is obtained through theoretical reasoning, so the characteristics of the flow field generated by the charge flow in the leakage channel of the dam have a strong decrease relative to the conductivity of the surrounding uniform material, can be detected by magnetic measurements. Based on this finding, the present invention arranges the emitter electrodes A and B, and obtains the magnetic anomaly vector through the magnetic induction intensity vector at each observation point under the injected current signal and the non-injected current signal, and then determines the current density distribution according to the magnetic anomaly vector, wherein there is a The non-zero current density distribution is considered to be generated by the seepage channel on the dam. Furthermore, the present invention provides a brand-new method for determining the location of the leakage channel of the dam lifting. For the first time, the method of magnetic field measurement is used to realize the location of the leakage channel, and based on the magnetic field measurement, it has the advantages of fast, convenient operation and low cost. Therefore, the present invention can determine the leakage channel of the dam lifting more quickly, efficiently and at low cost.

Description of drawings

FIG. 1 is a schematic diagram of the leakage channel detection of a dam based on magnetic field measurement according to the present invention. Among them, A and B are the emitter electrodes, A is located in the water on one side of the reservoir, and B is located in the water or underground on the other side of the dam; I(ω, t) is the power supply current; the solid circle is the magnetic observation point of the magnetic induction intensity vector. on the dam.

Figure 2 is an anomaly diagram of the vertical component B _sz of the magnetic field.

Figure 3 shows the results of the current density distribution in the leakage channel obtained by linear three-dimensional inversion.

Detailed ways

The present invention will be further described below with reference to the embodiments.

The present invention obtains the gradient sensitivity of the magnetic induction intensity vector to the electrical conductivity by deriving the formula of the magnetic anomaly vector, and the reasoning process is as follows:

The present invention involves injecting a current between two electrodes (ground or in a borehole) and measuring the magnetic flux density in three orthogonal directions.

This method involves Maxwell's equations under quasi-static conditions (that is, ignoring the time derivatives in the conservation equations), as follows:

Where E is the electric field strength, B=μH, B is the magnetic induction vector (or called magnetic flux density); H is the magnetization field, μ is the permeability of free space, and J _c is the conduction current density.

According to equation (1), the electric field can be passed through from scalar potential export. Whereas in a medium with a current generator, the total current can be divided into two parts: the primary current (source current density) associated with the current generator _Js and the volume current due to the electric field in the volume, so the conduction current density is given by J _c :

where J _s represents the source current density (also called the primary current density) within the source region, and σE is the volume current density outside the source region.

According to the law of conservation of charge, we can get:

It follows from equation (5) that the electric potential follows the following elliptic partial differential equation:

Therefore, the volume source current density ξ can be expressed as:

Regarding the partial differential equation of the magnetic induction vector B and its integral form solution, using Ampere's law, there is:

Applying curl to both sides of equation (8), we obtain:

Among them, when hour, because So there exists a vector A such that (A represents the magnetic vector potential). We use classical canonical conditions to avoid the uncertainty caused by the definition of A. Maxwell's equations It becomes A=μJ _c , which is the vector Poisson equation. If we apply A(|r|→∞)=0 (there is no magnetic induction B at infinity), it has a general solution at the magnetic observation point P(r) and in 3-D space:

Among them, r represents the position of the pre-arranged magnetic observation point P(r) on the dam, r' represents the position of the integration point around the source point M, A(r) represents the magnetic field vector potential at the magnetic observation point, and dτ' represents the source Volume element around point M. This method calculates A from the current density, and then calculates the curl of A to obtain the magnetic induction vector field B. Using curl, we use Biot's and Savart's laws to obtain:

Equation (12) can be written in the following form:

At the observation point P(r) (magnetometer location), the magnetic induction vector decomposes the background field magnetic induction vector B ₀ (r) and the magnetic induction vector B _s (r) produced by the current in the leakage channel:

B(r)=B ₀ (r)+B _s (r) (14)

Therefore, according to formula (13) and formula (14) there is:

The present invention also has another magnetic induction due to the fact that the wires are on the ground and are used to connect the current electrodes A and B to the generator. We discuss how to eliminate this spurious contribution below. Using this formula, B ₀ (r) is called the background magnetic induction, and the second term B _s (r) is called the abnormal magnetic induction, and the source of the current control J _s is known in the present invention. Therefore, we can easily remove the background field to get the anomalous field: B _s (r)=B(r)-B ₀ (r). Therefore there is:

use Rewriting equation (17) (based on the quasi-static constraints discussed above), we have:

because Then formula (18) can be written as:

evenly underground The magnetic field is reduced to the background field B ₀ . From Equation (18) and Equation (19), we can see that the magnetic induction vector is sensitive to the gradient of conductivity. That is to say, the characteristics of the flow field generated by the flow of charge in the leakage channel of the dam have a strong drop relative to the conductivity of the surrounding homogeneous material, which can be detected by magnetic measurements.

For anomalous fields, according to the relationship between current density and electromagnetic field, we can write formula (17) as follows:

In the formula, J(r')=σE is the current density in the leakage channel.

Based on the above theoretical basis, the idea of the present invention is to obtain the magnetic anomalous field first, and then use the magnetic anomalous field to obtain the current density distribution in the seepage channel, and then determine the seepage channel on the dam. As shown in Figure 1, the present invention arranges power supply electrodes A and B on both sides of the dam, wherein the power supply electrode A is located in the reservoir and is in direct contact with the water, and the power supply electrode B is located on the other side of the reservoir, which can be in the water or in the water. The land is connected to the signal transmitter; the observation points are arranged on the dam according to the designed line spacing and point spacing, and the receiver is used to collect the magnetic induction intensity vector of the observation point. In this embodiment, the receiver adopts a magnetic probe or a magnetometer. Point measurement magnetic flux density vector.

The present invention provides a method and device for detecting leakage channel of a dam based on magnetic field measurement, comprising the following steps:

S1: Obtain the background magnetic induction intensity vector at each magnetic observation point on the dam when the signal transmitter is not powered and the total magnetic induction intensity vector at each magnetic observation point when the power is supplied;

For example, the background magnetic induction intensity vector at the ith magnetic observation point is and the total magnetic induction vector B ⁱ (r).

S2: Calculate the difference between the total magnetic induction intensity vector and the background magnetic induction intensity vector at each magnetic observation point to obtain the magnetic anomaly vector of each magnetic observation point.

at the i-th magnetic observation point Equal to the total magnetic induction intensity vector B ⁱ (r) and the background magnetic induction intensity vector difference value. The magnetic anomaly field of _Tiba is constructed based on the magnetic anomaly vector of each magnetic observation point.

S3: The current density distribution is deduced based on the magnetic anomaly vector of each magnetic observation point in the magnetic anomaly field;

Among them, the location where the current density distribution is not zero corresponds to the location of the leakage channel on the dam. As shown in Fig. 3, the non-zero current density distribution is the location of the leakage channel on the dam.

In this embodiment, the linear three-dimensional inversion method is used to deduce the current density distribution, and the process is as follows:

Discrete the three-dimensional space of the dam into M units, assuming that each unit underground has the same volume v and that there is a constant current density in each unit For a single unit k and a single observation point i, the magnitude of the magnetic field can be expressed as:

For N measuring points, the relationship between the magnetic anomaly field and the current density field is as follows:

B _s = KJ (22)

In the formula, B _s is the magnetic anomaly field matrix, the size of the matrix is 3N×1, are the magnetic anomaly vectors at the 1st, ith, and Nth magnetic observation points respectively, N is the number of magnetic observation points; K is the position matrix, and the size of the matrix is 3N×3M, are the distance matrices between the first magnetic observation point and the first and Mth units, respectively, are the distance matrices between the Nth magnetic observation point and the first unit and the Mth unit, respectively, is the distance matrix between the i-th magnetic observation point and the k-th unit; J is the current density field matrix, the size of the matrix is 3M×1, and J ¹ , J ^k , and J ^M are the first unit and the k-th unit, respectively. unit, the current density on the Mth unit;

In the formula, magnetic anomaly vector The components in the x, y, and z directions in the spatial Cartesian coordinate system xyz, μ is the magnetic permeability of free space, ν is the unit volume, r _ik is the space between the i-th magnetic observation point and the k-th unit distance, are the spatial distance between the i-th magnetic observation point and the k-th unit, respectively The components in the x, y, and z directions in the space Cartesian coordinate system xyz; are the components of the current density J ^k on the kth unit in the x, y, and z directions in the spatial Cartesian coordinate system xyz, respectively.

In the present invention, the magnetic anomalous field matrix B _s is obtained by monitoring the magnetic induction intensity vector at the observation point, and the position matrix K is known. The next step is how to solve the current density field matrix. In the embodiment of the present invention, the objective function Φ is established according to the principle of the least multiplication algorithm. (J) and get the solution of the linear inverse problem of the objective function;

Among them, the objective function is as follows:

Φ(J)=[W _d (d _obs -KJ)] ^T [W _d (d _obs -KJ)]+λ[W _m (JJ ₀ )] ^T [W _m (JJ ₀ )]

The formula for the solution is as follows:

In the formula, λ is the regularization factor, J ₀ is the prior model of the current density, d _obs is the observation data, W _d and W _m are weight matrices, and the weight matrix W _d is the error covariance matrix of the observation data d _obs . A diagonal matrix composed of reciprocals, and the weight matrix W _m is a smoothness matrix. Among them, the prior model is derived based on the historical data of current density distribution, leakage channel and magnetic anomaly field in historical experimental data or historical monitoring data.

Substitute the magnetic anomaly field matrix and position matrix on the dam into the formula of the solution to calculate the current density field matrix to obtain the current density distribution on the dam, where the magnetic anomaly field matrix B _s(3N×1) is the observation data d _obs .

It should be noted that, in other feasible embodiments, it is further preferred that in step S3, based on the relationship between the magnetic anomaly field and the current density field, the characteristic point method of the magnetic anomaly curve, the tangent method, and the two-dimensional inversion method can also be used to deduce the current. Density distribution, wherein the inversion method also includes but is not limited to Gauss-Newton method, nonlinear conjugate gradient method, and genetic algorithm.

On the other hand, a device based on the above method provided by the present invention includes: transmitting electrodes A and B, a signal transmitter, a receiver, and an analysis processor;

Among them, the transmitting electrodes A and B are respectively arranged on both sides of the dam, the transmitting electrode A is located in the reservoir and is in contact with the water, and the signal transmitter is connected between the transmitting electrodes A and B; The magnetic observation points are arranged at the distance and point distance, and the receiver is arranged at the magnetic observation point to collect the magnetic induction intensity vector;

The analysis processor is connected with the receiver, and is used for acquiring the background magnetic induction intensity vector at each magnetic observation point collected by the receiver when the signal transmitter is not powered and at each magnetic observation point collected when the signal transmitter is powered on The total magnetic induction intensity vector, and identify the location of the leakage channel on the dam. The analysis processor may be a computer including a processing chip or other terminal equipment, which stores a program corresponding to the above method, and invokes and runs the stored program to determine the location of the leakage channel.

Among them, preferably the receiver is a magnetic probe or a magnetometer. The transmitting signal of the signal transmitter is a DC signal or an AC signal with a frequency of 0.1 to 380 Hz.

The present invention assumes that there is a leakage channel of 12m×6m×4m at z=25m in the underground space, its resistivity is 10Ω·m, the resistivity of the background space is 1000Ω·m, and the coordinates of the center point are (50m, 50m, 25m ). Electrodes A, B are located 5 meters underground and 25 meters apart. Using the above principle, the magnetic anomaly distribution map generated by the leakage channel can be obtained. The magnetic anomaly map of the vertical component is shown in Figure 2, and the unit is pt. The obtained current density distribution map is shown in Figure 3. From the figure It can be seen that the algorithm in this paper can obtain the location and current distribution of the underground leakage channel very well.

It should be emphasized that the examples described in the present invention are illustrative rather than restrictive, so the present invention is not limited to the examples described in the specific implementation manner, and all the examples obtained by those skilled in the art according to the technical solutions of the present invention Other embodiments that do not depart from the spirit and scope of the present invention, whether modified or replaced, also belong to the protection scope of the present invention.

Claims (9)

1. a dam leakage channel detection method based on magnetic field measurement, is characterized in that: comprise the steps: S1: Obtain the background magnetic induction intensity vector at each magnetic observation point on the dam when the signal transmitter is not powered and the total magnetic induction intensity vector at each magnetic observation point when the power is supplied; Wherein, transmitting electrodes A and B are arranged on both sides of the dam, the transmitting electrode A is located in the reservoir and is in contact with the water, and a signal transmitter is connected between the transmitting electrodes A and B, on the dam according to the preset Line spacing and point spacing to arrange magnetic observation points; S2: Calculate the difference between the total magnetic induction intensity vector and the background magnetic induction intensity vector at each magnetic observation point to obtain the magnetic anomaly vector of each magnetic observation point; Among them, the magnetic anomaly vector of each magnetic observation point constitutes the magnetic anomaly field of Tiba; S3: The current density distribution is deduced based on the magnetic anomaly vector of each magnetic observation point in the magnetic anomaly field; Among them, the location where the current density distribution is not zero corresponds to the location of the leakage channel on the dam. 2. The method according to claim 1, wherein in step S3, the current density distribution is deduced based on the relationship between the magnetic anomalous field and the current density field; Among them, the three-dimensional space of Tiba is discretized into M units, each unit has the same volume and there is a constant current density in each unit, the relationship between the magnetic anomaly field and the current density field is as follows: B _s = KJ In the formula, B _s is the magnetic anomaly field matrix, the size of the matrix is 3N×1, are the magnetic anomaly vectors at the first and Nth magnetic observation points respectively, N is the number of magnetic observation points; K is the position matrix, and the size of the matrix is 3N×3M, are the distance matrices between the first magnetic observation point and the first and Mth units, respectively, are the distance matrices between the Nth magnetic observation point and the first unit and the Mth unit; J is the current density field matrix, the size of the matrix is 3M×1, J ¹ , J ^M are the first unit, The current density on the Mth cell; Among them, the magnetic anomaly vector of any i-th magnetic observation point Distance matrix between the i-th magnetic observation point and the k-th cell The current density Jk on the ^kth cell is as follows: In the formula, magnetic anomaly vector The components in the x, y, and z directions in the spatial Cartesian coordinate system xyz, μ is the magnetic permeability of free space, ν is the unit volume, r _ik is the space between the i-th magnetic observation point and the k-th unit distance, are the spatial distance between the i-th magnetic observation point and the k-th unit, respectively The components in the x, y, and z directions in the space Cartesian coordinate system xyz; are the components of the current density J ^k on the kth unit in the x, y, and z directions in the spatial Cartesian coordinate system xyz, respectively. 3. The method according to claim 2, characterized in that: in step S3, the characteristic point method, tangent method, two-dimensional inversion method or linear three-dimensional inversion of the magnetic anomaly curve is adopted based on the relationship between the magnetic anomaly field and the current density field. The method deduces the current density distribution. 4. method according to claim 3 is characterized in that: the process that adopts linear three-dimensional inversion method to deduce current density distribution is as follows: First, establish the objective function Φ(J) according to the principle of the least multiplication algorithm and obtain the solution of the linear inverse problem of the objective function; Among them, the objective function is as follows: Φ(J)=[W _d (d _obs -KJ)] ^T [W _d (d _obs -KJ)]+λ[W _m (JJ ₀ )] ^T [W _m (JJ ₀ )] The formula for the solution is as follows: In the formula, λ is the regularization factor, J ₀ is the prior model of the current density, d _obs is the observation data, W _d and W _m are weight matrices, and the weight matrix W _d is the error covariance matrix of the observation data d _obs . The diagonal matrix composed of reciprocals, the weight matrix W _m is the smoothness matrix; Then, the magnetic anomaly field matrix and the position matrix on the dam are substituted into the formula of the solution to calculate the current density field matrix to obtain the current density distribution on the dam; Among them, the magnetic anomaly field matrix B _s is the observation data d _obs . 5 . The method according to claim 1 , wherein the magnetic induction intensity vector is collected by using a magnetic probe at the observation point. 6 . 6. A device based on the method according to any one of claims 1-5, characterized in that: comprising: transmitting electrodes A and B, a signal transmitter, a receiver, and an analysis processor; Wherein, the transmitting electrodes A and B are respectively arranged on both sides of the dam, the transmitting electrode A is located in the reservoir and is in contact with the water, and the signal transmitter is connected between the transmitting electrodes A and B; on the dam The magnetic observation points are arranged according to the preset line distance and point distance, and the receiver is arranged at the magnetic observation point to collect the magnetic induction intensity vector; The analysis processor is connected with the receiver, and is used for acquiring the background magnetic induction intensity vector at each magnetic observation point collected by the receiver when the signal transmitter is not powered and each magnetic observation collected when the signal transmitter is powered on The total magnetic induction intensity vector at the point, and identify the location of the leakage channel on the dam. 7. The device according to claim 6, wherein the receiver is a magnetic probe or a magnetometer. 8. The device according to claim 6, wherein the transmitting electrode B is located on land. 9 . The device according to claim 6 , wherein the transmission signal of the signal transmitter is a DC signal or an AC signal with a frequency of 0.1-380 Hz. 10 .