Dike hidden danger detection and early warning method and system based on electric field and electromagnetic field coupling

Abstract

The present invention discloses a method and system for detecting and warning embankment hidden dangers based on the coupling of electric and electromagnetic fields. The method includes the following steps: obtaining and processing the resistivity and dielectric constant of the embankment body; constructing a joint inversion objective function based on the processed results; establishing a joint inversion model and solving the problem by minimizing the joint inversion objective function; obtaining multi-physics field joint inversion results to delineate abnormal areas; deploying a time-lapse resistivity monitoring network in the area for detection; obtaining seepage velocity; defining soil mechanical parameters; constructing a numerical model for embankment hidden dangers using seepage velocity as a boundary condition input; simulating the stability safety factor of the embankment body under different seepage velocities; setting an early warning threshold; and issuing an alarm when the stability safety factor of the embankment body reaches the early warning threshold. The present invention effectively eliminates data multi-solutions through coupled analysis of geoelectric fields, seepage velocity, and other factors, significantly improving defect resolution and greatly reducing the risk of misjudgment.

Description

Technical Field

[0001] The present invention belongs to the technical field of water conservancy project safety monitoring, and specifically relates to a method and system for detecting and warning hidden dangers in embankments based on the coupling of electric and electromagnetic fields. The method is suitable for the precise detection and dynamic warning of hidden defects such as leakage, cracks, and cavities in water conservancy facilities such as embankments and earth-rock dams. Background Art

[0002] The detection and early warning of levee hazards in a low-detectability physical environment presents a pressing technical challenge in the safety management of levee projects. Traditional levee hazard detection methods primarily rely on single-physics field techniques (such as high-density resistivity and geological radar). These methods suffer from the following drawbacks: First, they suffer from high multi-solution variability. Existing technologies lack a multi-physics field collaborative inversion mechanism, and a single physical field cannot distinguish between hazard targets of different origins (such as seepage channels and clay layers), resulting in insufficient hazard location accuracy. Second, there is a conflict between resolution and detection depth. Geological radar has high resolution in shallow layers but insufficient penetration depth, while resistivity has deep penetration but slightly lower resolution. For example, high-density resistivity is susceptible to shallow interference when detecting deep seepage, while geological radar is sensitive to shallow cracks but cannot effectively detect deep, porous areas. Third, dynamic monitoring capabilities are weak. Existing technologies lack dynamic, time-shifted, and continuous observation capabilities, making it impossible to track the temporal and spatial evolution of hazards in real time, assess the risk of hazard catastrophic changes, and implement dynamic early warnings. Therefore, it is necessary to develop a detection technology and system based on the coupling of geoelectric and electromagnetic fields to enhance the ability to accurately detect and provide dynamic early warnings for levee hazards. Summary of the Invention

[0003] In order to solve the problems in the related art, the present application provides a dike hidden danger detection and early warning method and system based on the coupling of electric field and electromagnetic field, which solves the problems mentioned in the background technology.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a method for detecting and warning embankment hidden dangers based on the coupling of electric and electromagnetic fields, comprising the following steps:

[0005] Step S1: Obtain the resistivity and dielectric constant of the embankment body;

[0006] Step S2: Process the resistivity and dielectric constant, and construct a joint inversion objective function based on the processed results;

[0007] Step S3: Establishing a joint inversion model, solving the joint inversion model by minimizing the joint inversion objective function, and obtaining a multi-physics joint inversion result; delineating the abnormal area based on the multi-physics joint inversion result;

[0008] Step S4: using a time-lapse resistivity monitoring network to perform periodic resistivity detection on the abnormal area to obtain time series resistivity data;

[0009] Step S5: converting the time series resistivity data into a spatiotemporal continuous three-dimensional resistivity field using the Kriging interpolation method, and calculating the seepage velocity based on the spatiotemporal continuous three-dimensional resistivity field;

[0010] Step S6: Based on the results of the multi-physics field joint inversion, a numerical model of the embankment hidden danger is established using the COMSOL stress-seepage coupling module; soil mechanical parameters are defined, and the seepage velocity is input into the numerical model of the embankment hidden danger as a boundary condition to simulate the stability safety factor of the embankment under different seepage velocities;

[0011] Step S7: Set an early warning threshold, and issue an alarm when the embankment dam stability safety factor reaches the early warning threshold.

[0012] Furthermore, in step S2, a joint inversion objective function is constructed, and the specific process is as follows:

[0013] First, the acquired resistivity and dielectric constant are processed, and a joint inversion objective function is constructed based on the processed results. The joint inversion objective function consists of a fitting error term, a model regularization term, and a cross-gradient constraint term.

[0014] Minimize the joint inversion objective function, expressed as:

[0015] (1);

[0016] Where, represents resistivity; represents the dielectric constant; represents the minimization of the joint inversion objective function; and are the fitted differences of resistivity and dielectric constant respectively; is the model regularization term; is the cross gradient constraint; and are the weight coefficients of the resistivity and dielectric constant fitting difference terms respectively; is the weight coefficient of the model regularization term; is the weight coefficient of the cross gradient constraint;

[0017] Among them, F ERT and F GPR Calculated using formula (2) and formula (3) respectively, it is expressed as:

[0018] (2);

[0019] (3);

[0020] Where, and are the observed resistivity and dielectric constant of the i-th measuring point respectively; and are the calculated resistivity and the calculated dielectric constant of the i-th measuring point respectively; represents the error weight of the resistivity of the i-th measuring point; represents the error weight of the dielectric constant of the i-th measuring point; n represents the total number of measuring points;

[0021] The model regularization term is calculated using formula (4), which is expressed as:

[0022] (4);

[0023] Where, are the joint inversion model parameters including resistivity and dielectric constant;

[0024] The cross gradient constraint term is calculated using formula (5), which is expressed as:

[0025] (5);

[0026] Where, 、 、 are the gradients of resistivity at x, y, and z coordinates, respectively; 、 、 are the gradients of the dielectric constant at the x, y, and z coordinates, respectively.

[0027] Furthermore, in step S3, the multi-physics field joint inversion result is obtained. The specific process is as follows:

[0028] The joint inversion model is constructed based on the Bayesian framework. The parameters of the joint inversion model are solved by minimizing the joint inversion objective function. The gradient descent method is used in the optimization process to iteratively update the parameters of the joint inversion model, which is expressed as:

[0029] (6);

[0030] Where, are the joint inversion model parameters of the kth iteration; represents the joint inversion model parameters of the k+1th iteration; η is the learning rate;

[0031] When the change rate of the joint inversion model parameters iteratively updated by minimizing the joint inversion objective function is less than 1% or the number of iterations reaches the maximum value, the iteration is terminated and the multi-physics field joint inversion result is obtained.

[0032] Furthermore, in step S4, the time series resistivity data is obtained. The specific process is as follows:

[0033] A time-lapse resistivity monitoring network is arranged to conduct periodic resistivity detection in abnormal areas to obtain time series resistivity data ρ(x, y, z, t); where x, y, z are the x, y, z coordinates of the embankment soil space; and t is the detection time.

[0034] Furthermore, in step S5, the seepage velocity is calculated based on the spatiotemporal continuous three-dimensional resistivity field. The specific process is as follows:

[0035] The time series resistivity data ρ(x,y,z,t) are converted into a spatiotemporal continuous three-dimensional resistivity field using Kriging interpolation method.

[0036] The seepage velocity is calculated based on the time-space continuous three-dimensional resistivity field, which is expressed as:

[0037] (7);

[0038] Where v represents the seepage velocity, Δx is the displacement of the seepage front within Δt; Indicates a time interval; represents the resistivity at time t; Indicates the change in resistivity; represents the rate of change of resistivity with time t; Resistivity changes with coordinates The rate of change.

[0039] Furthermore, in step S6, the stability safety factor of the embankment dam under different seepage velocities is simulated. The specific process is as follows:

[0040] According to the results of the multi-physics field joint inversion in step S3, the COMSOL stress-seepage coupling module is used to establish a numerical model of embankment hidden dangers and define the soil mechanical parameters; the soil mechanical parameters include elastic modulus E, Poisson's ratio , permeability coefficient K, effective cohesion c′ and effective internal friction angle ϕ′ of soil; and the seepage velocity is input into the numerical model of embankment hidden danger as the boundary condition; COMSOL stress-seepage coupling module is used to simulate the stability safety factor of embankment dam under different seepage velocities based on soil mechanical parameters and the numerical model of embankment hidden danger.

[0041] Furthermore, the specific process of step S7 is as follows:

[0042] When 1.5>F s When the value is ≥1.0, an alarm is issued, and the early warning information is sent through the data processing PC and transmitted to the host computer; F s It represents the safety factor of embankment dam stability.

[0043] Furthermore, the embankment dam stability safety factor is calculated using the simplified Bishop method in the limit equilibrium method through the following formula (8):

[0044] (8);

[0045] Where W is the weight of the soil strip; θ is the inclination angle of the sliding surface; l is the length of the sliding surface strip.

[0046] A dike hidden danger detection and early warning system based on electric field and electromagnetic field coupling, applied to the dike hidden danger detection and early warning method based on electric field and electromagnetic field coupling, comprising:

[0047] Multi-physics field synchronous acquisition module: used to obtain the resistivity and dielectric constant of the embankment;

[0048] The first data processing and analysis unit is used to process the resistivity and dielectric constant and construct a joint inversion objective function based on the processed results;

[0049] The second data processing and analysis unit is used to establish a joint inversion model, solve the joint inversion model by minimizing the joint inversion objective function, obtain multi-physics joint inversion results; and delineate abnormal areas based on the multi-physics joint inversion results;

[0050] The third data processing and analysis unit is used to perform periodic resistivity detection on the abnormal area using the time-lapse resistivity monitoring network to obtain time series resistivity data;

[0051] The fourth data processing and analysis unit is used to convert the time series resistivity data into a spatiotemporal continuous three-dimensional resistivity field using the Kriging interpolation method, and calculate the seepage velocity based on the spatiotemporal continuous three-dimensional resistivity field;

[0052] The fifth data processing and analysis unit is used to establish a numerical model of embankment hidden dangers based on the results of multi-physics field joint inversion using the COMSOL stress-seepage coupling module. It defines soil mechanical parameters and inputs the seepage velocity as a boundary condition into the numerical model of embankment hidden dangers to simulate the stability safety factor of the embankment under different seepage velocities.

[0053] The sixth data processing and analysis unit is used to set a warning threshold and issue an alarm when the stability safety factor of the embankment dam reaches the warning threshold.

[0054] Compared with the existing technology, the present invention has the following beneficial effects:

[0055] (1) The present invention breaks through the limitations of traditional single physical field detection and effectively eliminates data multi-solutions through coupled analysis of geoelectric field, seepage velocity, etc., significantly improves defect resolution, and greatly reduces the risk of misjudgment.

[0056] (2) The present invention adopts the Bayesian framework and time-lapse resistivity monitoring network technology, combined with the COMSOL stress-permeability coupling module for real-time modeling, to achieve a leap from static detection to dynamic tracking, and can monitor the defect evolution trend (such as leakage expansion rate) in real time, shortening the early warning response time. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 Flow chart of the method of the present invention.

[0058] Figure 2 It is a schematic diagram of the embankment dam structure of the present invention.

[0059] Figure numerals: 1. embankment dam body; 2. tracked multi-frequency geological radar vehicle; 3. high-density electrical survey line; 4. radar data transmission cable; 5. electrical data transmission cable; 6. geological radar host; 7. high-density electrical host; 8. data processing PC. DETAILED DESCRIPTION

[0060] like Figure 1 As shown, the present invention provides a technical solution: a dike hidden danger detection and early warning method based on the coupling of electric field and electromagnetic field, comprising the following steps:

[0061] Step S1: Obtain the resistivity and dielectric constant of the embankment body;

[0062] Step S2: Process the resistivity and dielectric constant, and construct a joint inversion objective function based on the processed results;

[0063] Step S3: Establishing a joint inversion model, solving the joint inversion model by minimizing the joint inversion objective function, and obtaining a multi-physics joint inversion result; delineating the abnormal area based on the multi-physics joint inversion result;

[0064] Step S4: using a time-lapse resistivity monitoring network to perform periodic resistivity detection on the abnormal area to obtain time series resistivity data;

[0065] Step S5: converting the time series resistivity data into a spatiotemporal continuous three-dimensional resistivity field using the Kriging interpolation method, and calculating the seepage velocity based on the spatiotemporal continuous three-dimensional resistivity field;

[0066] Step S6: Based on the results of the multi-physics field joint inversion, a numerical model of the embankment hidden danger is established using the COMSOL stress-seepage coupling module; soil mechanical parameters are defined, and the seepage velocity is input into the numerical model of the embankment hidden danger as a boundary condition to simulate the stability safety factor of the embankment under different seepage velocities;

[0067] Step S7: Set an early warning threshold, and issue an alarm when the embankment dam stability safety factor reaches the early warning threshold.

[0068] The specific process of obtaining the resistivity and dielectric constant of the embankment dam is as follows:

[0069] Figure 2 Figure 1 is the embankment dam, 2 is the tracked multi-frequency geological radar vehicle, 3 is the high-density electrical survey line, 4 is the radar data transmission cable, 5 is the electrical data transmission cable, 6 is the geological radar host, 7 is the high-density electrical host, and 8 is the data processing PC.

[0070] Specifically, a high-density electrical survey line 3 is laid out along the axis parallel to the embankment dam body 1, and the geoelectric field data (resistivity) of the embankment dam body 1 is acquired using the high-density electrical survey line 3. The geoelectric field data (resistivity) is then transmitted to a high-density electrical host computer 7 via an electrical data transmission cable 5. Simultaneously, a tracked multi-frequency geological radar vehicle 2 is used to collect electromagnetic field data (dielectric constant) of the embankment dam body 1, and the electromagnetic field data (dielectric constant) is then transmitted to a geological radar host computer 6 via a radar data transmission cable 4. After pre-processing, the resistivity and dielectric constant are uniformly transmitted to a data processing PC 8 via 5G / wireless.

[0071] Specifically, the detection of levee hidden dangers is divided into two stages: general survey and detailed investigation. The general survey stage focuses on identifying suspected hidden danger locations, while the detailed investigation stage further confirms the spatial distribution of hidden dangers and tracks their spatiotemporal changes based on the suspected hidden danger locations identified in the general survey stage.

[0072] During the survey phase, tracked multi-frequency geological radar vehicles were used, with a 100MHz antenna for continuous scanning of the embankment and dam body, with a detection depth of 8-10m. High-density electrical survey lines (high-density resistivity method) were used simultaneously to collect data along the embankment and dam body, with an electrode spacing of 5m and a detection depth of 20m. During the detailed investigation phase, intensified detection was carried out on abnormal areas discovered during the survey phase. Based on the previous electrode spacing of 5m and a detection depth of 20m, the spacing between resistivity measurement points was increased to 2m, and the tracked multi-frequency geological radar vehicle used a 400MHz antenna to improve the resolution. A time-lapse resistivity monitoring network was arranged in abnormal areas (suspected leakage areas) to continuously collect geoelectric field data and analyze the dynamic changes in seepage.

[0073] Among them, the joint inversion objective function is constructed in step S2, and the specific process is:

[0074] First, the acquired resistivity and dielectric constant are processed, and a joint inversion objective function is constructed based on the processed results. The joint inversion objective function consists of a fitting error term, a model regularization term, and a cross-gradient constraint term.

[0075] Minimize the joint inversion objective function, expressed as:

[0076] (1);

[0077] Where, represents resistivity; represents the dielectric constant; represents the minimization of the joint inversion objective function; and are the fitted differences of resistivity and dielectric constant respectively; is the model regularization term; is the cross gradient constraint; and are the weight coefficients of the resistivity and dielectric constant fitting difference terms respectively; is the weight coefficient of the model regularization term; is the weight coefficient of the cross gradient constraint;

[0078] Among them, F ERT and F GPR Calculated using formula (2) and formula (3) respectively, it is expressed as:

[0079] (2);

[0080] (3);

[0081] Where, and are the observed resistivity and dielectric constant of the i-th measuring point respectively; and are the calculated resistivity and the calculated dielectric constant of the i-th measuring point respectively; represents the error weight of the resistivity of the i-th measuring point; represents the error weight of the dielectric constant of the i-th measuring point; n represents the total number of measuring points;

[0082] The model regularization term uses sparse constraints (L1 norm) and is calculated through formula (4), which is expressed as:

[0083] (4);

[0084] Where, are the parameters of the joint inversion model including resistivity and permittivity;

[0085] Among them, the cross gradient constraint term is calculated using formula (5), which is expressed as:

[0086] (5);

[0087] Where, 、 、 are the resistivity gradients in the x, y, and z directions, respectively; 、 、 are the gradients of the dielectric constant in the x, y, and z directions, respectively.

[0088] Among them, the multi-physics field joint inversion result is obtained in step S3, and the specific process is:

[0089] A joint inversion model is constructed based on a Bayesian framework. The joint inversion model parameters are solved by minimizing the joint inversion objective function. The gradient descent method is used in the optimization process to iteratively update the joint inversion model parameters. When the rate of change of the iterative update of the joint inversion model parameters by minimizing the joint inversion objective function is less than 1% or the number of iterations reaches the maximum, the iteration is terminated and the multi-physics field joint inversion result is obtained.

[0090] Iteratively update the joint inversion model parameters, expressed as:

[0091] (6);

[0092] Where, are the joint inversion model parameters of the kth iteration; represents the joint inversion model parameters of the k+1th iteration; η is the learning rate, which is generally 0.01;

[0093] The above-mentioned multi-physical field joint inversion method can eliminate the multi-solution characteristics of a single physical field, improve the inversion resolution, and accurately locate hidden dangers.

[0094] In step S4, the time series resistivity data is obtained, and the specific process is as follows:

[0095] A time-lapse resistivity monitoring network is arranged to conduct periodic resistivity detection in abnormal areas to obtain time series resistivity data ρ(x, y, z, t); where x, y, z are the x, y, z coordinates of the embankment soil space; and t is the detection time.

[0096] In step S5, the seepage velocity is calculated based on the spatiotemporal continuous three-dimensional resistivity field. The specific process is as follows:

[0097] The time series resistivity data ρ(x,y,z,t) are converted into a spatiotemporal continuous three-dimensional resistivity field using Kriging interpolation method.

[0098] The seepage velocity is calculated based on the time-space continuous three-dimensional resistivity field, which is expressed as:

[0099] (7);

[0100] Where v represents the seepage velocity, Δx is the displacement of the seepage front within Δt; Indicates a time interval; represents the resistivity at time t; Indicates the change in resistivity; represents the rate of change of resistivity with time t; Resistivity changes with coordinates The rate of change.

[0101] Among them, the stability safety factor of the embankment and dam body under different seepage velocities is simulated in step S6. The specific process is:

[0102] According to the results of the multi-physics field joint inversion in step S3, the COMSOL stress-seepage coupling module is used to establish a numerical model of embankment hidden dangers and define the soil mechanical parameters; the soil mechanical parameters include elastic modulus E, Poisson's ratio , permeability coefficient K, effective cohesion c′ and effective internal friction angle ϕ′ of soil; and the seepage velocity is input into the numerical model of embankment hidden danger as the boundary condition; COMSOL stress-seepage coupling module is used to simulate the stability safety factor of embankment dam under different seepage velocities based on soil mechanical parameters and the numerical model of embankment hidden danger.

[0103] The specific process of step S7 is as follows:

[0104] Set the three-level warning threshold: when F s ≥1.5 is the yellow warning threshold, when 1.5>F s ≥1.0 is the orange warning threshold, when F s <1.0 is the red warning threshold; F s It represents the stability safety factor of the embankment and dam body;

[0105] When 1.5>F s When the value is ≥1.0, an alarm is issued, and a warning message is sent through the data processing PC and transmitted to the host computer to remind the staff.

[0106] The embankment dam stability safety factor is calculated using the simplified Bishop method in the limit equilibrium method using the following formula (8):

[0107] (8);

[0108] Where W is the weight of the soil strip; θ is the inclination angle of the sliding surface; l is the length of the sliding surface strip.

[0109] A dike hidden danger detection and early warning system based on electric field and electromagnetic field coupling, applied to the dike hidden danger detection and early warning method based on electric field and electromagnetic field coupling, comprising:

[0110] Multi-physics field synchronous acquisition module: used to obtain the resistivity and dielectric constant of the embankment;

[0111] The first data processing and analysis unit is used to process the resistivity and dielectric constant and construct a joint inversion objective function based on the processed results;

[0112] The second data processing and analysis unit is used to establish a joint inversion model, solve the joint inversion model by minimizing the joint inversion objective function, obtain multi-physics joint inversion results; and delineate abnormal areas based on the multi-physics joint inversion results;

[0113] The third data processing and analysis unit is used to perform periodic resistivity detection on the abnormal area using the time-lapse resistivity monitoring network to obtain time series resistivity data;

[0114] The fourth data processing and analysis unit is used to convert the time series resistivity data into a spatiotemporal continuous three-dimensional resistivity field using the Kriging interpolation method, and calculate the seepage velocity based on the spatiotemporal continuous three-dimensional resistivity field;

[0115] The fifth data processing and analysis unit is used to establish a numerical model of embankment hidden dangers based on the results of multi-physics field joint inversion using the COMSOL stress-seepage coupling module. It defines soil mechanical parameters and inputs the seepage velocity as a boundary condition into the numerical model of embankment hidden dangers to simulate the stability safety factor of the embankment under different seepage velocities.

[0116] The sixth data processing and analysis unit is used to set a warning threshold and issue an alarm when the stability safety factor of the embankment dam reaches the warning threshold.

[0117] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for detecting and warning embankment hidden dangers based on the coupling of electric field and electromagnetic field, characterized in that: The following steps are involved: Step S1: Obtain the resistivity and dielectric constant of the embankment body; Step S2: Process the resistivity and dielectric constant, and construct a joint inversion objective function based on the processed results; Step S3: Establishing a joint inversion model, solving the joint inversion model by minimizing the joint inversion objective function, and obtaining a multi-physics joint inversion result; delineating the abnormal area based on the multi-physics joint inversion result; Step S4: using a time-lapse resistivity monitoring network to perform periodic resistivity detection on the abnormal area to obtain time series resistivity data; Step S5: converting the time series resistivity data into a spatiotemporal continuous three-dimensional resistivity field using the Kriging interpolation method, and calculating the seepage velocity based on the spatiotemporal continuous three-dimensional resistivity field; Step S6: Based on the results of the multi-physics field joint inversion, a numerical model of the embankment hidden danger is established using the COMSOL stress-seepage coupling module; soil mechanical parameters are defined, and the seepage velocity is input into the numerical model of the embankment hidden danger as a boundary condition to simulate the stability safety factor of the embankment under different seepage velocities; Step S7: Set an early warning threshold, and issue an alarm when the embankment dam stability safety factor reaches the early warning threshold.

2. The method for detecting and warning embankment hidden dangers based on coupling of electric and electromagnetic fields according to claim 1, characterized in that: In step S2, the joint inversion objective function is constructed. The specific process is as follows: First, the acquired resistivity and dielectric constant are processed, and a joint inversion objective function is constructed based on the processed results. The joint inversion objective function consists of a fitting error term, a model regularization term, and a cross-gradient constraint term. Minimize the joint inversion objective function, expressed as: （1）； Where, represents resistivity; represents the dielectric constant; represents the minimization of the joint inversion objective function; and are the fitted differences of resistivity and dielectric constant respectively; is the model regularization term; is the cross gradient constraint; and are the weight coefficients of the resistivity and dielectric constant fitting difference terms respectively; is the weight coefficient of the model regularization term; is the weight coefficient of the cross gradient constraint; Among them, F ERT and F GPR Calculated using formula (2) and formula (3) respectively, it is expressed as: （2）； （3）； Where, and are the observed resistivity and dielectric constant of the i-th measuring point respectively; and are the calculated resistivity and the calculated dielectric constant of the i-th measuring point respectively; represents the error weight of the resistivity of the i-th measuring point; represents the error weight of the dielectric constant of the i-th measuring point; n represents the total number of measuring points; The model regularization term is calculated using formula (4), which is expressed as: （4）； Where, are the joint inversion model parameters including resistivity and dielectric constant; The cross gradient constraint term is calculated using formula (5), which is expressed as: （5）； Where, 、 、 are the gradients of resistivity at x, y, and z coordinates, respectively; 、 、 are the gradients of the dielectric constant at the x, y, and z coordinates, respectively.

3. The method for detecting and warning embankment hidden dangers based on coupling of electric and electromagnetic fields according to claim 2, characterized in that: In step S3, the multi-physics field joint inversion results are obtained. The specific process is as follows: The joint inversion model is constructed based on the Bayesian framework. The parameters of the joint inversion model are solved by minimizing the joint inversion objective function. The gradient descent method is used in the optimization process to iteratively update the parameters of the joint inversion model, which is expressed as: （6）； Where, are the joint inversion model parameters of the kth iteration; represents the joint inversion model parameters of the k+1th iteration; η is the learning rate; When the change rate of the joint inversion model parameters iteratively updated by minimizing the joint inversion objective function is less than 1% or the number of iterations reaches the maximum value, the iteration is terminated and the multi-physics field joint inversion result is obtained.

4. The method for detecting and warning embankment hidden dangers based on coupling of electric and electromagnetic fields according to claim 3, characterized in that: In step S4, the time series resistivity data is obtained. The specific process is as follows: A time-lapse resistivity monitoring network is arranged to conduct periodic resistivity detection in abnormal areas to obtain time series resistivity data ρ(x, y, z, t); where x, y, z are the x, y, z coordinates of the embankment soil space; and t is the detection time.

5. The method for detecting and warning embankment hidden dangers based on electric field and electromagnetic field coupling according to claim 4 is characterized in that: In step S5, the seepage velocity is calculated based on the spatiotemporal continuous three-dimensional resistivity field. The specific process is as follows: The time series resistivity data ρ(x,y,z,t) are converted into a spatiotemporal continuous three-dimensional resistivity field using Kriging interpolation method. The seepage velocity is calculated based on the time-space continuous three-dimensional resistivity field, which is expressed as: （7）； Where v represents the seepage velocity, Δx is the displacement of the seepage front within Δt; Indicates a time interval; represents the resistivity at time t; Indicates the change in resistivity; represents the rate of change of resistivity with time t; Resistivity changes with coordinates The rate of change.

6. The method for detecting and warning embankment hidden dangers based on electric field and electromagnetic field coupling according to claim 5, characterized in that: In step S6, the stability safety factor of the embankment dam under different seepage velocities is simulated. The specific process is as follows: According to the multi-physics field joint inversion results in step S3, the COMSOL stress-seepage coupling module is used to establish a numerical model of embankment hidden dangers and define soil mechanical parameters; Soil mechanical parameters include elastic modulus E, Poisson's ratio , permeability coefficient K, effective soil cohesion c′ and effective soil internal friction angle ϕ′; The seepage velocity is input into the numerical model of embankment hidden danger as the boundary condition; the COMSOL stress-seepage coupling module is used to simulate the stability safety factor of the embankment under different seepage velocities based on the soil mechanical parameters and the numerical model of embankment hidden danger.

7. The method for detecting and warning embankment hidden dangers based on electric field and electromagnetic field coupling according to claim 6, characterized in that: The specific process of step S7 is: When 1.5>F s When the value is ≥1.0, an alarm is issued, and the early warning information is sent through the data processing PC and transmitted to the host computer; F s It represents the safety factor of embankment dam stability.

8. The method for detecting and warning embankment hidden dangers based on electric field and electromagnetic field coupling according to claim 7, characterized in that: The safety factor of embankment dam stability is calculated using the simplified Bishop method in the limit equilibrium method through the following formula (8): （8）； Where W is the weight of the soil strip; θ is the inclination angle of the sliding surface; l is the length of the sliding surface strip.

9. A dike hidden danger detection and early warning system based on electric field and electromagnetic field coupling, applied to a dike hidden danger detection and early warning method based on electric field and electromagnetic field coupling according to any one of claims 1 to 8, characterized in that: include: Multi-physics field synchronous acquisition module: used to obtain the resistivity and dielectric constant of the embankment; The first data processing and analysis unit is used to process the resistivity and dielectric constant and construct a joint inversion objective function based on the processed results; The second data processing and analysis unit is used to establish a joint inversion model, solve the joint inversion model by minimizing the joint inversion objective function, obtain multi-physics joint inversion results; and delineate abnormal areas based on the multi-physics joint inversion results; The third data processing and analysis unit is used to perform periodic resistivity detection on the abnormal area using the time-lapse resistivity monitoring network to obtain time series resistivity data; The fourth data processing and analysis unit is used to convert the time series resistivity data into a spatiotemporal continuous three-dimensional resistivity field using the Kriging interpolation method, and calculate the seepage velocity based on the spatiotemporal continuous three-dimensional resistivity field; The fifth data processing and analysis unit is used to establish a numerical model of embankment hidden dangers based on the results of multi-physics field joint inversion using the COMSOL stress-seepage coupling module. It defines soil mechanical parameters and inputs the seepage velocity as a boundary condition into the numerical model of embankment hidden dangers to simulate the stability safety factor of the embankment under different seepage velocities. The sixth data processing and analysis unit is used to set a warning threshold and issue an alarm when the stability safety factor of the embankment dam reaches the warning threshold.

Images