Dynamic monitoring method of mine disaster water source based on nuclear magnetic resonance and resistivity

Abstract

The present invention provides a method for dynamically monitoring water sources of mine disasters based on nuclear magnetic resonance and resistivity, and obtains two types of geophysical detection data obtained by using nuclear magnetic resonance method and resistivity method in the target detection area respectively; according to the data detected by the resistivity method, the electrical information thereof is extracted by resistivity four-dimensional inversion imaging technology, the electrical model is determined, the electrical model is gridded, and the gridded model is used as the initial model; an objective function considering the initial model, spatial smoothness constraint and temporal smoothness constraint is constructed, and based on the data detected by the nuclear magnetic resonance method, the nuclear magnetic resonance four-dimensional inversion is performed, and the objective function is optimized by a nonlinear fitting method, and it is continuously iterated until the set requirements are met, and a joint inversion result is obtained, and the water content and water volume information of the formation are obtained. The present invention can ensure that the inversion result is more accurate.

Description

Technical Field

[0001] The invention belongs to the technical field of data processing and inversion imaging in the field of geophysics, and specifically relates to a method for dynamic monitoring of mine disaster water sources based on the combination of nuclear magnetic resonance and resistivity. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Nuclear magnetic resonance detection is a geophysical detection method that has been developed in recent years. This method uses the paramagnetic properties of hydrogen protons in water to reflect the water content, porosity and distribution of aquifers by applying and shutting off the nuclear magnetic resonance signal generated by the excitation magnetic field. Since the greater the water content, the stronger the information that this technology can display, and the water volume can be directly quantitatively detected, it has been widely used in the fields of groundwater resource detection and exploration, and disaster warning that may be caused by groundwater. It is recognized as a technology that can directly measure the amount of groundwater. However, this method also has certain limitations. First, this method ignores the distribution of resistivity of the formation and the water-containing body. The water content in the geological body will affect the conductivity of the formation and ultimately weaken the real nuclear magnetic signal intensity. Therefore, this has a great impact on the imaging results. Secondly, traditional nuclear magnetic resonance mostly uses one-dimensional inversion based on uniform half-space or layered models. When the underground environment is complex, the error will be large, and it cannot quickly reflect the time-related water volume changes in real time.

[0004] The resistivity four-dimensional inversion imaging technology can reflect the time-related conductivity changes of the strata and underground water-bearing bodies, and indirectly obtain the groundwater parameters based on the negative correlation between conductivity and water content. Therefore, the present invention considers combining nuclear magnetic resonance with resistivity four-dimensional inversion to efficiently and accurately monitor the dynamics of mine disaster water sources.

[0005] At present, nuclear magnetic resonance detection technology and resistivity four-dimensional inversion technology can be used to detect groundwater information and determine whether there is a flood risk. However, each has its own shortcomings in terms of detection accuracy:

[0006] First, NMR is sensitive to differences in the conductivity of underground media. Therefore, the underground resistivity distribution is one of the key factors affecting the results of NMR detection. It is particularly important to determine the appropriate electrical model, but NMR technology ignores this sensitive factor.

[0007] Secondly, traditional nuclear magnetic resonance mostly uses one-dimensional forward and inversion theory. This method can indeed give better imaging results for groundwater with approximately horizontal layered distribution. However, in reality, the underground environment is mostly complex. At this time, since the layered distribution is not satisfied, forcibly treating the complex three-dimensional structure as a one-dimensional structure will result in large errors, and it is even more impossible to obtain changes in hydrological parameters related to the four-dimensional time.

[0008] Third, the four-dimensional resistivity inversion obtains the distribution information of underground resistivity, and cannot directly obtain hydrological parameters such as water content, and does not have the ability to quantitatively analyze flood risks. Summary of the invention

[0009] In order to solve the above problems, the present invention proposes a dynamic monitoring method for mine disaster water sources based on the combination of nuclear magnetic resonance and resistivity. The present invention cleverly combines nuclear magnetic resonance four-dimensional inversion and resistivity four-dimensional inversion, which can not only quantitatively measure the water content and water volume, but also reduce the error caused by the conductivity of the underground medium to the nuclear magnetic inversion, forming a real joint constraint on the inversion process, and the inversion result is more accurate.

[0010] According to some embodiments, the present invention adopts the following technical solutions:

[0011] A method for dynamic monitoring of mine disaster water sources based on nuclear magnetic resonance and resistivity, comprising the following steps:

[0012] (1) Obtain two types of geophysical exploration data in the target exploration area, using nuclear magnetic resonance and resistivity methods respectively;

[0013] (2) Based on the data detected by the resistivity method, the electrical information is extracted through resistivity four-dimensional inversion imaging technology, the electrical model is determined, the electrical model is gridded, and the gridded model is used as the initial model;

[0014] (3) Construct an objective function that takes into account the initial model, spatial smoothness constraints, and temporal smoothness constraints. Perform nuclear magnetic resonance four-dimensional inversion based on the data detected by the nuclear magnetic resonance method. Use a nonlinear fitting method to optimize the objective function and iterate continuously until the set requirements are met. Obtain a joint inversion result and obtain information on the water content and water volume of the formation.

[0015] As an optional implementation, the following step (4) is also included: using the Bayesian inversion method to quantitatively describe the reliability of the joint inversion result and reduce the inversion error, performing imaging processing on the inversion result to obtain an optimized inversion result.

[0016] As an optional implementation, in step (1), monitoring is performed using nuclear magnetic resonance and resistivity methods respectively, and data collection is performed continuously. The nuclear magnetic resonance method collects information by applying and shutting down the nuclear magnetic resonance signal generated by the excitation magnetic field, and the resistivity method obtains resistance data by measuring the electrodes.

[0017] As an optional implementation, in step (2), the electrical information is extracted by resistivity four-dimensional inversion imaging technology, and the specific process of determining the electrical model includes:

[0018] A resistivity four-dimensional inversion objective function is constructed, wherein the resistivity four-dimensional inversion objective function includes a data fitting term, a spatial smoothness constraint term and a temporal smoothness constraint term, and then a four-dimensional resistivity inversion equation is determined. The four-dimensional resistivity inversion equation is solved, and an inversion result, i.e., a corresponding electrical model, is obtained by continuous iteration.

[0019] As an optional implementation, in step (2), the specific process of meshing the electrical model includes: using the finite element method to mesh according to the stratum structure and boundary information reflected by the electrical model.

[0020] As an optional implementation, in step (3), the specific process of performing nuclear magnetic resonance four-dimensional inversion based on the data detected by the nuclear magnetic resonance method includes: calculating the distribution of the underground excitation magnetic field, solving it using the finite element method, calculating the kernel function and the initial amplitude, and then calculating the inverted water content value.

[0021] As an optional implementation, in step (3), the objective function is optimized by a nonlinear fitting method, and the optimization is performed continuously until the set requirements are met. The specific process of obtaining the joint inversion result includes: optimizing by a nonlinear fitting method based on least squares, and defining the objective function as:

[0022] Where C is the kernel function matrix, n is the water content model parameter, α and β are the weights of the influence of spatial and temporal smoothness constraints on the objective function;

[0023] Based on the initial NMR inversion model, the objective function is minimized and the joint inversion result is obtained through continuous iteration.

[0024] As an optional implementation, the specific process of step (4) includes: determining the prior probability distribution of the current model based on the previously output joint inversion results, and establishing an initial model of Bayesian inversion;

[0025] Perform Bayesian inversion calculations to obtain all acceptable models and their likelihood probabilities, and calculate their posterior probability distributions;

[0026] Calculate the mean and variance of the model at this time. The mean model represents the most likely inversion result, and the variance model represents the error range. If the error range is less than the set value, the inversion result at this time is the optimized inversion result. Otherwise, continue to iterate the Bayesian inversion until the requirements are met.

[0027] A mine disaster water source dynamic monitoring system based on nuclear magnetic resonance and resistivity, comprising:

[0028] A data acquisition module is configured to acquire two types of geophysical exploration data obtained in the target exploration area by using a nuclear magnetic resonance method and a resistivity method respectively;

[0029] The resistivity four-dimensional inversion module is configured to extract electrical information of the data detected by the resistivity method through the resistivity four-dimensional inversion imaging technology, determine the electrical model, grid the electrical model, and use the gridded model as the initial model;

[0030] The joint inversion module is configured to perform nuclear magnetic resonance four-dimensional inversion based on the data detected by the nuclear magnetic resonance method, construct an objective function that takes into account the initial model, spatial smoothness constraints and temporal smoothness constraints, and optimize the objective function using a nonlinear fitting method. It is continuously iterated until the set requirements are met, and the joint inversion results are obtained to obtain the water content and water volume information of the formation.

[0031] As an optional implementation, it also includes a joint inversion optimization module, which is configured to use a Bayesian inversion method to quantitatively describe the reliability of the joint inversion result, reduce the inversion error, perform imaging processing on the inversion result, and obtain an optimized inversion result.

[0032] As an optional implementation, it also includes a data acquisition module, including a nuclear magnetic resonance acquisition system and a resistivity acquisition system.

[0033] An electronic device comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps in the above method are completed.

[0034] A computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the steps in the above method are completed.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] The present invention uses a four-dimensional resistivity inversion method, through which information such as the boundary and resistivity distribution of the underground aquifer is explored, providing a reference basis for nuclear magnetic resonance grid segmentation, and adding time smoothness constraints to obtain time-related resistivity changes and make real-time adjustments.

[0037] The present invention proposes a four-dimensional nuclear magnetic resonance inversion strategy, which breaks the spatial and temporal limitations of traditional one-dimensional inversion, adds spatial smoothness constraints and temporal smoothness constraints, and greatly reduces the error of nuclear magnetic resonance detection results caused by forcibly assuming a layered distribution in complex underground environments. At the same time, four-dimensional inversion can also capture time-related changes in water content and water volume in real time.

[0038] The present invention proposes a method for dynamic monitoring of mine disaster water sources based on the combination of nuclear magnetic resonance and resistivity four-dimensional inversion, which uses an electrical model to limit the stratum structure, so as to achieve the effect of resistivity inversion guiding nuclear magnetic inversion. Nuclear magnetic inversion can directly and quantitatively measure the water content and water volume, and resistivity inversion can be used to reduce the error caused by the conductivity of the underground medium to the nuclear magnetic inversion. Combining the advantages of the two, a real joint constraint is formed on the inversion process, and more accurate and efficient imaging is achieved.

[0039] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0041] Figure 1 It is a schematic diagram of the dynamic monitoring process of mine disaster water sources of the present invention;

[0042] Figure 2 It is the Bayesian inversion optimization flow chart of the present invention. DETAILED DESCRIPTION

[0043] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0044] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0045] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0046] A dynamic monitoring method for mine disaster water sources based on nuclear magnetic resonance and resistivity four-dimensional inversion. Figure 1 As shown, the following steps are included:

[0047] (1) Select the detection area, use the nuclear magnetic resonance method and resistivity method to collect relevant data, and import the working parameters.

[0048] (2) Define the resistivity measured data set DR = (DR 1 ,DR 2 ,…,DR Nk ) T .

[0049] (3) Four-dimensional resistivity inversion: construct a four-dimensional resistivity inversion equation, solve the model parameter increment vector ΔM, and obtain the electrical model through continuous iteration.

[0050] (4) Based on the stratum structure and boundary information reflected by the electrical model, the finite element method is used to divide the grid, and the gridded model is used as the initial model;

[0051] (5) Using the initial model, based on the nuclear magnetic resonance detection data, perform nuclear magnetic resonance four-dimensional inversion, construct a nuclear magnetic resonance inversion objective function, use a nonlinear fitting method to iterate continuously until the objective function value reaches the accuracy requirement, and output the joint inversion result.

[0052] In the step (1), the same detection area is selected, and the nuclear magnetic resonance and resistivity methods are used for monitoring respectively, and data collection is performed continuously, wherein the nuclear magnetic resonance collects information by applying and shutting down the nuclear magnetic resonance signal generated by the excitation magnetic field, and the resistivity method uses a certain device, such as a dipole-dipole device, a Wenner device, a Schlumberger device, etc., or a combination of certain devices, to supply power to points A and B, measure the potential difference between MN, and then divide it by the power supply current at this time to obtain resistance data.

[0053] In step (3), the objective function is as follows:

[0054]

[0055] The first term on the right side of the equation is the data fitting term, the second term is the spatial smooth constraint term, and the third term is the temporal smooth constraint term. are the corresponding functions in space and time respectively, and λ and μ are the weights of the influence of spatial and temporal smoothness constraints on the objective function.

[0056] The first data error term W dis the data weight matrix, e is the difference vector between the measured data and the theoretical data, e = Dr-G(M)-JΔM, where Dr is the measured data, G represents the forward modeling, and M n is a set of model parameters, corresponding to the measured data one by one, J n It is an extension of a single model sensitivity matrix and corresponds one-to-one to the model parameters.

[0057] The obtained four-dimensional resistivity inversion equation is as follows:

[0058] J T W d T W d J+λK T K+μK T T K T )ΔM=J T W d T (W d Dr-G(M))-μK T T K T M

[0059] Where K is the spatial smoothness matrix, K T is the temporal smoothness matrix.

[0060] By solving the inversion equation to obtain ΔM, the inversion result, i.e. the corresponding electrical model, is obtained by continuous iteration.

[0061] In step (5), during the inversion process, the distribution of the underground excited magnetic field is calculated. Under the premise of only an external electrical source, it is assumed that the harmonic change factor is Then the electromagnetic field satisfies the following Maxwell equations:

[0062]

[0063]

[0064] Where E is the electric field strength, H is the magnetic field strength, μ is the magnetic permeability, is the electrical conductivity. D = εE is the electric flux intensity or electric displacement, where ε is the dielectric constant. J s is the current density of the external charge source, is the circular frequency of the electrical source.

[0065] When using the finite element method to solve, the electric field or vector potential can be calculated first, and then the magnetic field value can be calculated by derivative. T =μH, B T is the magnetic induction intensity.

[0066] Calculate the kernel function and initial amplitude. When isotropic, the initial amplitude of NMR can be expressed as: Where N is the number of small grids, C j is the kernel function of the jth unit, n j is the moisture content of the jth unit.

[0067]

[0068] Among them, f L is the frequency in the transmitting coil (Larmor frequency, f L =0.04258×|B 0 |, B 0 is the actual geomagnetic field strength at the measuring point), M 0 is the nuclear magnetization intensity of hydrogen protons, η is the magnetic gyrometry ratio of hydrogen protons, B ⊥ is the exciting magnetic field B T Perpendicular to the Earth's magnetic field B 0 The amount.

[0069] It can be expressed in matrix form as C·n=E 0 , by solving the equation group, the inverted water content value can be obtained.

[0070] In step (5), a four-dimensional nuclear magnetic resonance inversion is performed to obtain information such as water content and water volume. The optimization is performed using a nonlinear fitting method based on least squares, and the objective function is defined as:

[0071] Where C is the kernel function matrix, n is the water content model parameter, and α and β are the weights of the influence of spatial and temporal smoothness constraints on the objective function.

[0072] Given the initial NMR inversion model, the objective function is minimized and the joint inversion result is obtained through continuous iteration.

[0073] The above inversion results are relatively reliable. However, in order to improve the monitoring effect, the present invention also provides a method for optimizing the above joint inversion results. Figure 2 As shown, including:

[0074] (6) Bayesian inversion: Based on the previously output joint inversion results, the prior probability distribution of the current model is determined and the initial model of the Bayesian inversion is established.

[0075] (7) Bayesian inversion calculation, save all acceptable models and their likelihood probabilities, and calculate their posterior probability distribution.

[0076] (8) The mean and variance of the model are calculated. The mean model represents the most likely inversion result, and the variance model represents the error range. If the error range is less than the set value, the inversion result is the optimized inversion result. Otherwise, the Bayesian inversion is iterated continuously until the requirements are met.

[0077] (9) Perform imaging processing on the inversion results to provide flood and water volume warnings for the monitored area.

[0078] Specifically, in step (7), the prior probability distribution of the model is determined based on the joint inversion result previously obtained by least squares inversion, and the initial model is established using this as a constraint condition. Where g(m) is the forward modeling result, C -1 is the inverse of the covariance matrix, which includes the effect of the observation error of each model on the data.

[0079] In step (8), Bayesian inversion is performed, using the Metropolis criterion as the standard, to save all acceptable models and their likelihood probabilities. Statistics are then taken of all inverted model samples to calculate their posterior probability distributions. Where P(m) and P(d) are the prior probability distributions of model parameters and observed data, respectively; P(dm) is the likelihood probability of specific observed data under given model parameters; and σ(md) is the posterior probability of model parameters obtained by combining prior information and likelihood probability.

[0080] In step (9), the mean and variance of the model are calculated, the mean model is the most likely inversion result, and the variance model is the error range.

[0081] As a typical real-time method, (1) the same detection area is continuously collected using nuclear magnetic resonance and resistivity methods. Nuclear magnetic resonance collects information by applying and shutting down the nuclear magnetic resonance signal generated by the excitation magnetic field. The resistivity method uses a Wenner-Schlumberger device to supply power to points A and B, collect the potential of the remaining electrode points, and divide it by the power supply current at this time to extract the resistance data.

[0082] (2) Define the resistivity measured data set DR = (DR 1 ,DR 2 ,…,DR Nk ) T Nk represents the number of data sets used for inversion.

[0083] (3) Perform four-dimensional resistivity inversion. The objective function is: The first term on the right side of the equation is the data fitting term, the second term is the spatial smooth constraint term, and the third term is the temporal smooth constraint term. are the corresponding functions in space and time respectively, and λ and μ are the weights of the influence of spatial and temporal smoothness constraints on the objective function. W d is the data weight matrix, e is the difference vector between the measured data and the theoretical data, e = Dr-G(M)-JΔM, where Dr is the measured data, G represents the forward modeling, and M n is a set of model parameters, corresponding to the measured data one by one, J n It is an extension of a single model sensitivity matrix and corresponds one-to-one with the model parameters. The four-dimensional resistivity inversion equation is:

[0084] J T W d T W d J+λK T K+μK T T K T )ΔM=J T W d T (W d Dr-G(M))-μK T T K T M, where K is the spatial smoothness matrix, K T is the time smoothness matrix. By solving the inversion equation, ΔM is obtained, and the inversion result, i.e. the corresponding electrical model, is obtained by continuous iteration.

[0085] (4) Based on the obtained electrical model, the finite element method is used to perform meshing.

[0086] (5) Calculate the distribution of the underground excited magnetic field. First calculate the electric field or vector potential, and then take the derivative to calculate the magnetic field value, B T =μH, B T is the magnetic induction intensity. Under the premise of only external electrical source, it is assumed that the harmonic change factor is Then the electromagnetic field satisfies the following Maxwell equations:

[0087]

[0088]

[0089] Where E is the electric field strength, H is the magnetic field strength, μ is the magnetic permeability, is the electrical conductivity. D = εE is the electric flux intensity or electric displacement, where ε is the dielectric constant. J s is the current density of the external charge source, is the circular frequency of the electrical source.

[0090] Calculate the kernel function, where f L is the frequency in the transmitting coil (Larmor frequency, f L =0.04258×|B 0 |, B 0 is the actual geomagnetic field strength at the measuring point), M 0 is the nuclear magnetization intensity of hydrogen protons, η is the magnetic gyrometry ratio of hydrogen protons, B ⊥ is the exciting magnetic field B T Perpendicular to the Earth's magnetic field B 0 The amount.

[0091] Calculate the initial amplitude. Assuming isotropy, the initial amplitude of the NMR is expressed as: Where N is the number of small grids, C j is the kernel function of the jth unit, n j is the moisture content of the jth unit.

[0092] Solve C·n=E 0 The inverted water content value is obtained by using the equation group.

[0093] (6) NMR four-dimensional inversion, the objective function is defined as: Where C is the kernel function matrix, n is the water content model parameter, α and β are the weights of the spatial and temporal smoothness constraints on the objective function. The optimization is performed using the least squares-based nonlinear fitting method.

[0094] The objective function is minimized and the joint inversion result is obtained through continuous iteration.

[0095] (7) Based on the joint inversion results obtained previously through least squares inversion, the prior probability distribution of the model is determined, and this is used as a constraint to establish the initial model. Where g(m) is the forward modeling result, C -1 is the inverse of the covariance matrix.

[0096] (8) Bayesian inversion calculation, using the Metropolis criterion as the standard, saves all acceptable models and their likelihood probabilities. And counts all inverted model samples and calculates their posterior probability distribution. Where P(m) and P(d) are the prior probability distributions of model parameters and observation data, respectively; P(d|m) is the likelihood probability of specific observation data given the model parameters; and σ(m|d) is the posterior probability of the model parameters obtained by combining the prior information and the likelihood probability.

[0097] (9) The mean and variance of the model are calculated. The mean model is the most likely inversion result, and the variance model is the error range.

[0098] (10) The obtained model is imaged and processed to provide early warning of floods and water volume in the monitored area.

[0099] In order to verify that the joint inversion of NMR and resistivity is better than the inversion of NMR alone, a simulation example can be performed using the following electrical model.

[0100] A. The NMR data were collected under the following conditions: the NMR transmitting coil and receiving coil were both 1 turn, the size was 60m×60m, the surrounding resistivity was 300Ω·m, the geomagnetic field strength was 50000nT, the geomagnetic inclination was 45°, the geomagnetic declination was 0°, and the Larmor frequency was f L =0.04258×|B 0 |, here is 2129Hz. The water-bearing body model is 20m long, 20m wide, 8m thick, and centered at (0, 0, 10). The resistivity of the water-bearing body is 10Ω·m, and the water content is 50%. 36 measuring points are arranged on the ground, and the distance between the measuring points in the x and y directions is 40m. The range of X is -80~80m and Y is -80~80m is observed.

[0101] B. Collect resistivity data using the Wenner-Schlumberger apparatus using the high-density resistivity method.

[0102] C. Resistivity four-dimensional inversion imaging to obtain the corresponding electrical model.

[0103] D. Discretize the detection area in the x, y, and z directions with a grid spacing of 10m×10m×5m.

[0104] E. Use the 300Ω·m uniform half-space electrical model to perform NMR inversion alone, output the inversion results, and perform Bayesian inversion on this basis to obtain its mean model.

[0105] F. Under the same discrete grid, the electrical model obtained by resistivity four-dimensional inversion imaging is used for nuclear magnetic resonance inversion, and the inversion results are output. Bayesian inversion is also performed on this basis to obtain its mean model.

[0106] G. Comparing the inversion results obtained by E and F, it is found that although both can reflect the true contour of the water body, false anomalies will appear in E, and the local water content has a large error compared with the true water content, while F has higher accuracy.

[0107] This example shows that the electrical model will have a certain impact on the inversion results of NMR. Therefore, it is particularly important to determine the appropriate electrical model before conducting NMR. This also shows that the joint inversion of NMR and resistivity can better reflect the distribution of underground water content than the traditional NMR inversion alone.

[0108] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0109] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0110] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0111] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0112] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0113] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A method for dynamic monitoring of mine disaster water sources based on nuclear magnetic resonance and resistivity, characterized by: The following steps are involved: (1) Obtain two types of geophysical exploration data in the target exploration area, using nuclear magnetic resonance and resistivity methods respectively; The nuclear magnetic resonance and resistivity methods are used for monitoring and data collection is performed continuously. The nuclear magnetic resonance method collects information by applying and shutting down the nuclear magnetic resonance signal generated by the excitation magnetic field, and the resistivity method obtains resistance data through electrode measurement; (2) Based on the data detected by the resistivity method, the electrical information is extracted through resistivity four-dimensional inversion imaging technology, the electrical model is determined, the electrical model is gridded, and the gridded model is used as the initial model; The specific process of extracting electrical information through resistivity four-dimensional inversion imaging technology and determining the electrical model includes: Construct a resistivity four-dimensional inversion objective function, which includes a data fitting term, a spatial smoothness constraint term, and a temporal smoothness constraint term, and then determine a four-dimensional resistivity inversion equation, solve the four-dimensional resistivity inversion equation, and iterate continuously to obtain an inversion result, that is, the corresponding electrical model Perform four-dimensional resistivity inversion, and the objective function is: The first term on the right side of the equation is the data fitting term, the second term is the spatial smooth constraint term, and the third term is the temporal smooth constraint term; in are the corresponding functions in space and time respectively, λ and μ are the weights of the influence of spatial and temporal smoothness constraints on the objective function; in the data error term W d is the data weight matrix, e is the difference vector between the measured data and the theoretical data, e = Dr-G(M)-JΔM, where Dr is the measured data, G represents the forward modeling, and M n is a set of model parameters, corresponding to the measured data one by one, J n It is an extension of the single model sensitivity matrix, which corresponds one-to-one to the model parameters. The four-dimensional resistivity inversion equation is: (J T W d T W d J+λK T K+μK T T K T )ΔM=J T W d T (W d Dr-G(M))-μK T T K T M, Where K is the spatial smoothness matrix, K T is the time smoothness matrix; ΔM is obtained by solving the inversion equation, and the inversion result, i.e. the corresponding electrical model, is obtained by continuous iteration; (3) Construct an objective function that takes into account the initial model, spatial smoothness constraints, and temporal smoothness constraints. Based on the data detected by the nuclear magnetic resonance method, perform nuclear magnetic resonance four-dimensional inversion. Use a nonlinear fitting method to optimize the objective function, and iterate continuously until the set requirements are met. Obtain a joint inversion result and obtain the water content and water volume information of the formation. The objective function is optimized by nonlinear fitting method and iterated continuously until the set requirements are met. The specific process of obtaining the joint inversion result includes: optimizing by nonlinear fitting method based on least squares and defining the objective function as: Where C is the kernel function matrix, n is the water content model parameter, α and β are the weights of the influence of spatial and temporal smoothness constraints on the objective function; Based on the initial NMR inversion model, the objective function is minimized and the joint inversion result is obtained through continuous iteration. (4) Using the Bayesian inversion method, quantitatively describe the reliability of the joint inversion results, reduce the inversion error, perform imaging processing on the inversion results, and obtain optimized inversion results; Save all acceptable models and their likelihood probabilities; and perform statistics on all inverted model samples to calculate their posterior probability distribution. Where P(m) and P(d) are the prior probability distributions of model parameters and observed data, respectively; P(dm) is the likelihood probability of specific observed data under given model parameters; and σ(md) is the posterior probability of model parameters obtained by combining prior information and likelihood probability.

2. A method for dynamic monitoring of mine disaster water sources based on nuclear magnetic resonance and resistivity as claimed in claim 1, characterized in that: In the step (2), the specific process of meshing the electrical model includes: using the finite element method to mesh according to the stratum structure and boundary information reflected by the electrical model.

3. The method for dynamic monitoring of mine disaster water sources based on nuclear magnetic resonance and resistivity as claimed in claim 1, characterized in that: In the step (3), the specific process of performing nuclear magnetic resonance four-dimensional inversion based on the data detected by the nuclear magnetic resonance method includes: calculating the distribution of the underground excitation magnetic field, solving it using the finite element method, calculating the kernel function and the initial amplitude, and then calculating the inverted water content value.

4. The method for dynamic monitoring of mine disaster water sources based on nuclear magnetic resonance and resistivity as claimed in claim 1, characterized in that: The specific process of step (4) includes: determining the prior probability distribution of the current model based on the previously output joint inversion results, and establishing an initial model of Bayesian inversion; Perform Bayesian inversion calculations to obtain all acceptable models and their likelihood probabilities, and calculate their posterior probability distributions; Calculate the mean and variance of the model at this time. The mean model represents the most likely inversion result, and the variance model represents the error range. If the error range is less than the set value, the inversion result at this time is the optimized inversion result. Otherwise, continue to iterate the Bayesian inversion until the requirements are met.

5. A dynamic monitoring system for mine disaster water sources based on nuclear magnetic resonance and resistivity, characterized by: include: The data acquisition module is configured to acquire two types of geophysical exploration data obtained in the target exploration area by using a nuclear magnetic resonance method and a resistivity method respectively; The nuclear magnetic resonance and resistivity methods are used for monitoring and data collection is performed continuously. The nuclear magnetic resonance method collects information by applying and shutting down the nuclear magnetic resonance signal generated by the excitation magnetic field, and the resistivity method obtains resistance data through electrode measurement; The resistivity four-dimensional inversion module is configured to extract electrical information of the data detected by the resistivity method through the resistivity four-dimensional inversion imaging technology, determine the electrical model, grid the electrical model, and use the gridded model as the initial model; The specific process of extracting electrical information through resistivity four-dimensional inversion imaging technology and determining the electrical model includes: Construct a resistivity four-dimensional inversion objective function, which includes a data fitting term, a spatial smoothness constraint term, and a temporal smoothness constraint term, and then determine a four-dimensional resistivity inversion equation, solve the four-dimensional resistivity inversion equation, and iterate continuously to obtain an inversion result, that is, the corresponding electrical model Perform four-dimensional resistivity inversion, and the objective function is: The first term on the right side of the equation is the data fitting term, the second term is the spatial smooth constraint term, and the third term is the temporal smooth constraint term; in are the corresponding functions in space and time respectively, λ and μ are the weights of the influence of spatial and temporal smoothness constraints on the objective function; in the data error term W d is the data weight matrix, e is the difference vector between the measured data and the theoretical data, e = Dr-G(M)-JΔM, where Dr is the measured data, G represents the forward modeling, and M n is a set of model parameters, corresponding to the measured data one by one, J n It is an extension of the single model sensitivity matrix, which corresponds one-to-one to the model parameters. The four-dimensional resistivity inversion equation is: The joint inversion module is configured to perform NMR four-dimensional inversion based on the data detected by the NMR method, construct an objective function that takes into account the initial model, spatial smoothness constraints and temporal smoothness constraints, and optimize the objective function using a nonlinear fitting method, iterating continuously until the set requirements are met, obtaining the joint inversion results and obtaining the water content and water volume information of the formation; The objective function is optimized by nonlinear fitting method and iterated continuously until the set requirements are met. The specific process of obtaining the joint inversion result includes: optimizing by nonlinear fitting method based on least squares and defining the objective function as: Where C is the kernel function matrix, n is the water content model parameter, α and β are the weights of the influence of spatial and temporal smoothness constraints on the objective function; Based on the initial NMR inversion model, the objective function is minimized and the joint inversion result is obtained through continuous iteration. The Bayesian inversion method is used to quantitatively describe the reliability of the joint inversion results, reduce the inversion error, and perform imaging processing on the inversion results to obtain optimized inversion results. Save all acceptable models and their likelihood probabilities; and perform statistics on all inverted model samples to calculate their posterior probability distribution. Where P(m) and P(d) are the prior probability distributions of model parameters and observed data, respectively; P(dm) is the likelihood probability of specific observed data under given model parameters; and σ(md) is the posterior probability of model parameters obtained by combining prior information and likelihood probability.

6. A mine disaster water source dynamic monitoring system based on nuclear magnetic resonance and resistivity as claimed in claim 5, characterized in that: It also includes a joint inversion optimization module, which is configured to use a Bayesian inversion method to quantitatively describe the reliability of the joint inversion result, reduce the inversion error, perform imaging processing on the inversion result, and obtain an optimized inversion result; Or, it also includes a data acquisition module, including a nuclear magnetic resonance acquisition system and a resistivity acquisition system.

Images