Ground nuclear magnetic resonance underground water inversion method, device and equipment and storage medium

Abstract

The invention provides a ground nuclear magnetic resonance underground water inversion method and device, equipment and a storage medium. The method comprises the following steps: acquiring a ground detection nuclear magnetic resonance observation data envelope, a ground nuclear magnetic resonance forward modeling data envelope signal and an initial inversion model space parameter; mapping the space parameters of the initial inversion model in a preset range to obtain the updating parameters of the current iteration model; wherein the preset range is a numerical range meeting hydrogeological significance; constructing a ground nuclear magnetic resonance inversion objective function according to ground detection nuclear magnetic resonance observation data envelope, ground nuclear magnetic resonance forward modeling data envelope signals and current iteration model updating parameters; the ground nuclear magnetic resonance inversion objective function is solved through a quasi-Newton method, and updating parameters of a next-round iteration model are obtained; and if the difference value between the current-round iteration model updating parameter and the next-round model updating parameter is smaller than a preset error, taking the next-round model updating parameter as an inversion result. The accuracy of an inversion result can be improved.

Description

Technical Field

[0001] This application relates to the technical field of groundwater detection, and particularly to a ground nuclear magnetic resonance groundwater inversion method, device, equipment and storage medium. Background Technique

[0002] The surface nuclear magnetic resonance (SNMR) detection technology is an important geophysical exploration means in the world at present. Its principle is mainly to non-invasively directly detect the relaxation signal generated by hydrogen protons in groundwater excited by a radio frequency radiation field through a ground coil, so as to quantitatively evaluate the hydrogeological information of the surveyed area. Compared with other geophysical exploration methods, the SNMR method has the advantages of high measurement efficiency, unique inversion interpretation, and detailed exploration information, and has been widely used in the fields of hydrogeological exploration research and water disaster monitoring and early warning.

[0003] However, the groundwater SNMR detection is limited by its own weak nano-volt level signal, and complex environmental noise is very likely to affect its acquisition, resulting in poor actual detection effect. The common solution is to perform inversion based on actual detection data and improve the accuracy of the inversion result to make up for the poor detection effect. However, the existing inversion methods have the problem of inaccurate inversion results.

[0004] Therefore, a method is needed to improve the accuracy of the obtained inversion result. Summary of the Invention

[0005] This application provides a ground nuclear magnetic resonance groundwater inversion method to solve the technical problem of improving the accuracy of the obtained inversion result.

[0006] In a first aspect, this application provides a ground nuclear magnetic resonance groundwater inversion method, including:

[0007] Obtain a ground detection nuclear magnetic resonance observation data envelope, a ground nuclear magnetic resonance forward data envelope signal, and an initial inversion model space parameter; wherein, the ground detection nuclear magnetic resonance observation data envelope represents the actually collected groundwater characteristic data, the ground nuclear magnetic resonance forward data envelope signal represents the theoretical value of the underground medium obtained by simulation calculation, and the initial inversion model space parameter represents the physical characteristics initially describing the aquifer and the surrounding geological structure;

[0008] Map the initial inversion model space parameter within a preset range to obtain the updated parameter of the current round of iterative model; wherein, the preset range is a numerical range that meets the hydrogeological significance;

[0009] Construct a ground nuclear magnetic resonance inversion objective function according to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the updated parameter of the current round of iterative model;

[0010] Solve the ground nuclear magnetic resonance inversion objective function by the quasi-Newton method to obtain the updated parameters of the next iteration model.

[0011] If the difference between the updated parameters of the current iteration model and the updated parameters of the next iteration model is less than the preset error, then use the updated parameters of the next iteration model as the inversion result.

[0012] Optionally, for the above method, map the initial inversion model space parameters within a preset range to obtain the updated parameters of the current iteration model, including:

[0013] Express the ground nuclear magnetic resonance forward modeling envelope signal in the form of a discrete matrix to obtain the first formula.

[0014] Convert the first formula into the form of a linear differential equation to obtain the second formula; where the second formula is used to calculate the derivative of the updated parameters of the current iteration model.

[0015] Calculate the second formula by the natural logarithm transformation method and the cotangent transformation method, so that the initial inversion model space parameters are mapped within a preset range to obtain the updated parameters of the current iteration model.

[0016] Optionally, for the above method, calculate the second formula by the natural logarithm transformation method and the cotangent transformation method, so that the initial inversion model space parameters are mapped within a preset range to obtain the updated parameters of the current iteration model, including:

[0017] Obtain the first Jacobian matrix according to the second formula.

[0018] Process the first Jacobian matrix by the natural logarithm transformation method to obtain the second Jacobian matrix.

[0019] Process the second Jacobian matrix by the cotangent transformation method to obtain the final Jacobian matrix.

[0020] Map the initial inversion model space parameters within a preset range according to the final Jacobian matrix.

[0021] Optionally, for the above method, obtain the ground detection nuclear magnetic resonance observation envelope and the ground nuclear magnetic resonance forward modeling envelope signal, including:

[0022] Collect groundwater measurement data through a ground nuclear magnetic resonance water exploration instrument.

[0023] Preprocess the groundwater measurement data to obtain the ground detection nuclear magnetic resonance observation envelope; where the preprocessing includes noise suppression, envelope extraction, frequency correction, phase correction, and data decimation.

[0024] Obtain the water content, relaxation time, and any spatial position of the underground medium to be analyzed.

[0025] According to the water content, relaxation time, and any position in each space, a forward modeling envelope signal of ground nuclear magnetic resonance is obtained.

[0026] Optionally, for the method as above, based on the observed envelope of ground detection nuclear magnetic resonance, the forward modeling envelope signal of ground nuclear magnetic resonance, and the model update parameters of this round of iteration, a ground nuclear magnetic resonance inversion objective function is constructed, including:

[0027] Based on the observed envelope of ground detection nuclear magnetic resonance and the forward modeling envelope signal of ground nuclear magnetic resonance, a data error function is obtained;

[0028] Based on the initial inversion model space parameters and the model smoothness matrix, a model constraint function is obtained;

[0029] Through the standard form of Tikhonov regular solution, the data error function, the model constraint function, and the regularization coefficient, a first objective function is obtained;

[0030] The first objective function is arranged into a standard iterative solution format to construct the ground nuclear magnetic resonance inversion objective function.

[0031] Optionally, for the method as above, by using the quasi - Newton method to solve the ground nuclear magnetic resonance inversion objective function, the model update parameters of the next round of iteration are obtained, including:

[0032] The ground nuclear magnetic resonance inversion objective function is partially differentiated to obtain the gradient expression of the objective function;

[0033] By using the quasi - Newton method to solve the gradient expression of the objective function according to the gradient expression of the objective function, the model update parameters of the next round of iteration are obtained.

[0034] Optionally, for the method as above, it further includes:

[0035] If the difference between the model update parameters of this round of iteration and the model update parameters of the next round is greater than the preset error, then execute the step of constructing the ground nuclear magnetic resonance inversion objective function based on the observed envelope of ground detection nuclear magnetic resonance, the forward modeling envelope signal of ground nuclear magnetic resonance, and the model update parameters of the next round of iteration.

[0036] In a second aspect, the present application provides a ground nuclear magnetic resonance groundwater inversion device, including:

[0037] An acquisition module, configured to acquire a ground detection nuclear magnetic resonance observation data envelope, a ground nuclear magnetic resonance forward modeling data envelope signal, and initial inversion model space parameters; wherein, the ground detection nuclear magnetic resonance observation data envelope represents the actually collected groundwater characteristic data, the ground nuclear magnetic resonance forward modeling data envelope signal represents the theoretical numerical values of the underground medium obtained through simulation calculation, and the initial inversion model space parameters represent the physical characteristics initially describing the aquifer and the surrounding geological structure;

[0038] A first obtaining module, configured to map the initial inversion model space parameters within a preset range to obtain the model update parameters for this round of iteration; wherein, the preset range is a numerical range that satisfies hydrogeological significance;

[0039] A construction module, configured to construct a ground nuclear magnetic resonance inversion objective function according to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward modeling data envelope signal, and the model update parameters for this round of iteration;

[0040] A second obtaining module, configured to solve the ground nuclear magnetic resonance inversion objective function through the quasi-Newton method to obtain the model update parameters for the next round of iteration;

[0041] An acting as module, configured to use the model update parameters for the next round as the inversion result if the difference between the model update parameters for this round of iteration and the model update parameters for the next round is less than a preset error.

[0042] In a third aspect, an embodiment of the present application provides an electronic device, including: a memory, a processor;

[0043] The memory stores computer-executable instructions;

[0044] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the above first aspect and / or various possible implementation manners of the first aspect.

[0045] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, they are used to implement the above first aspect and / or various possible implementation manners of the first aspect.

[0046] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the above first aspect and / or various possible implementation manners of the first aspect.

[0047] The ground nuclear magnetic resonance groundwater inversion method, device, equipment, and storage medium provided by this application obtain the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the initial inversion model space parameters. Among them, the ground detection nuclear magnetic resonance observation data envelope represents the actually collected groundwater characteristic data, the ground nuclear magnetic resonance forward data envelope signal represents the theoretical values of the underground medium obtained through simulation calculations, and the initial inversion model space parameters represent the physical characteristics initially describing the aquifer and the surrounding geological structure. Mapping the initial inversion model space parameters within a preset range to obtain the model update parameters for this round of iteration. The preset range is a numerical range that meets the hydrogeological significance. By mapping the initial inversion model space parameters within the preset range, the inversion error can be reduced, the accuracy of the inversion result can be improved, and at the same time, the variation range of the inversion model space parameters can be narrowed, making the inversion process more stable. According to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the model update parameters for this round of iteration, a ground nuclear magnetic resonance inversion objective function is constructed. The ground nuclear magnetic resonance inversion objective function is solved by the quasi-Newton method to obtain the model update parameters for the next round of iteration. The quasi-Newton method can improve the efficiency of the inversion process. If the difference between the model update parameters for this round of iteration and the model update parameters for the next round is less than the preset error, the model update parameters for the next round are used as the inversion result. The method of this application can improve the accuracy of the inversion result and the efficiency and speed of the inversion. Description of the Drawings

[0048] The drawings here are incorporated into the specification and form a part of this specification, showing embodiments that conform to this application, and are used together with the specification to explain the principles of this application.

[0049] Figure 1 It is a schematic flow chart of a ground nuclear magnetic resonance groundwater inversion method provided by an embodiment of this application;

[0050] Figure 2 It is a schematic flow chart of another ground nuclear magnetic resonance groundwater inversion method provided by an embodiment of this application;

[0051] Figure 3 It is a schematic flow chart of the solution by the quasi-Newton method provided by an embodiment of this application;

[0052] Figure 4a It is a two-dimensional distribution map of water content of a simulation experiment provided by an embodiment of this application;

[0053] Figure 4b It is a two-dimensional distribution map of relaxation time of a simulation experiment provided by an embodiment of this application;

[0054] Figure 5a It is a two-dimensional imaging result map of water content provided by an embodiment of this application;

[0055] Figure 5b It is a two-dimensional imaging result diagram of relaxation time provided by an embodiment of the present application;

[0056] Figure 6 It is a schematic structural diagram of a ground nuclear magnetic resonance groundwater inversion device provided by an embodiment of the present application;

[0057] Figure 7 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application.

[0058] Through the above-mentioned drawings, specific embodiments of the present application have been shown, and there will be more detailed descriptions hereinafter. These drawings and textual descriptions are not intended to limit the scope of the concept of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. Specific Embodiments

[0059] Here, exemplary embodiments will be described in detail, and examples thereof are shown in the drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. On the contrary, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0060] Currently, the prior art proposes a three-dimensional magnetotelluric anisotropic inversion method based on the unstructured finite element method, which is based on the limited-memory quasi-Newton method (L-BFGS) and can achieve fast three-dimensional inversion of magnetotelluric data. By (1) obtaining magnetotelluric sounding data and screening the data for inversion according to anisotropic characteristics; (2) constructing a conductivity tensor model using the unstructured finite element method; (3) constructing an inversion objective function under anisotropic conditions; (3) performing forward modeling on the conductivity tensor to obtain a predicted full impedance tensor, and calculating the sensitivity matrix at the same time; (4) calculating the gradient of the objective function using the data fitting error; (5) calculating the model update amount using the L-BFGS algorithm and updating the parameters; iteratively executing (4) to (6), judging the iteration termination condition, and realizing three-dimensional inversion to obtain the inversion result. However, the prior art aims to invert the true geoelectric structure distribution of the underground space through the electromagnetic response measured on the earth's surface, and use this to assist geological interpreters to obtain a more accurate geological structure judgment. There are errors when directly used for groundwater detection. At the same time, the existing inversion methods may produce numerical values that do not conform to physical meaning, which not only affects the inversion efficiency and stability, but also brings large inversion errors. On the other hand, the hydrological parameters obtained by the prior art through simulation experiments (i.e., the final relaxation time and water content are inaccurate).

[0061] The ground nuclear magnetic resonance groundwater inversion method, device, equipment, and storage medium provided by this application obtain the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the initial inversion model space parameters, and then constrain the initial inversion model space parameters within a numerical range that meets hydrogeological significance. On the one hand, by narrowing the numerical range, the efficiency and stability of the inversion are improved. On the other hand, the occurrence of errors can be reduced. Then, according to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the model update parameters of this round of iteration, a ground nuclear magnetic resonance inversion objective function is constructed, and the ground nuclear magnetic resonance inversion objective function is solved by the quasi-Newton method to improve the inversion efficiency by the quasi-Newton method.

[0062] The technical solution of this application and how the technical solution of this application solves the above technical problems will be described in detail below with specific embodiments. These several specific embodiments below can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of this application will be described below in conjunction with the accompanying drawings.

[0063] Figure 1 It is a schematic flowchart of a ground nuclear magnetic resonance groundwater inversion method provided by an embodiment of this application. As Figure 1 shown, the method includes:

[0064] S101. Obtain the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the initial inversion model space parameters; among them, the ground detection nuclear magnetic resonance observation data envelope represents the actually collected groundwater characteristic data, the ground nuclear magnetic resonance forward data envelope signal represents the theoretical numerical values of the underground medium obtained according to simulation calculations, and the initial inversion model space parameters represent the physical characteristics initially describing the aquifer and the surrounding geological structure;

[0065] Among them, the field measurement data is collected by a ground nuclear magnetic resonance (SNMR) water detector, and then the field measurement data is preprocessed through processes such as noise suppression, envelope extraction, frequency / phase correction, and data channel extraction to obtain the ground detection nuclear magnetic resonance observation data envelope .

[0066] The ground nuclear magnetic resonance forward data envelope signal can be calculated through a preset forward formula.

[0067] The preset forward formula is:

[0068]

[0069] Among them, is the excitation pulse moment; is the excitation pulse time; r represents any position in space; The representative kernel function mainly maps the input data from a low-dimensional space to a high-dimensional space through a certain mapping, so that it is easier to find the linear separation of the data in the high-dimensional space. The use of the kernel function enables us to avoid explicitly performing this mapping, thus greatly reducing the computational complexity; and respectively refer to the water content and the relaxation time.

[0070] S102. Map the initial inversion model space parameters within a preset range to obtain the updated parameters of the current round of iteration; wherein, the preset range is a numerical range that satisfies the hydrogeological significance;

[0071] Among them, the numerical value that satisfies the hydrogeological significance cannot be negative. Therefore, the preset range needs to be a positive value and satisfy a certain value, specifically it can be 0.05 - 0.95.

[0072] In this embodiment, mapping the initial inversion model space parameters within a preset range to obtain the updated parameters of the current round of iteration includes:

[0073] Express the ground nuclear magnetic forward modeling envelope signal in the form of a discrete matrix to obtain the first formula;

[0074] Convert the first formula into the form of a linear differential equation to obtain the second formula; wherein, the second formula is used to calculate the derivative of the updated parameters of the current round of iteration;

[0075] Calculate the second formula through the natural logarithm transformation method and the cotangent transformation method, so that the initial inversion model space parameters are mapped within the preset range to obtain the updated parameters of the current round of iteration.

[0076] Among them, the first formula obtained by expressing the ground nuclear magnetic forward modeling envelope signal in the form of a discrete matrix is:

[0077]

[0078] Among them, represents the Jacobian matrix; the superscripts and respectively represent the real part and the imaginary part; the subscripts and respectively correspond to the partial derivatives with respect to the water content and the relaxation time, is the Jacobian matrix with respect to the water content, is the Jacobian matrix with respect to the relaxation time.

[0079] The Jacobian matrices with respect to the water content and the relaxation time are respectively expressed as:

[0080]

[0081] Convert the first formula into the form of a linear differential equation, and the second formula is obtained as follows:

[0082]

[0083] Wherein, is an intermediate variable for differential calculation, and m is the inversion model space parameter.

[0084] Through the natural logarithm transformation method and the cotangent transformation method, the second formula can be calculated, so that the initial inversion model space parameter is mapped within a preset range. The preset range is a value that satisfies the hydrogeological significance. Specifically, the preset range can be a positive number and between 0.05 and 0.95.

[0085] The advantage of such a setting is that: by optimizing the mapping and updating process of the model parameters, the technical effects of improving the inversion accuracy while enhancing the calculation efficiency and stability are achieved.

[0086] In this embodiment, through the natural logarithm transformation method and the cotangent transformation method, the second formula is calculated, so that the initial inversion model space parameter is mapped within a preset range, and the updated parameters of the current round of iteration model are obtained, including:

[0087] According to the second formula, the first Jacobian matrix is obtained;

[0088] Process the first Jacobian matrix through the natural logarithm transformation method to obtain the second Jacobian matrix;

[0089] Process the second Jacobian matrix through the cotangent transformation method to obtain the final Jacobian matrix;

[0090] According to the final Jacobian matrix, map the initial inversion model space parameter within a preset range.

[0091] Among them, mapping the initial inversion model parameter within a preset range is to achieve the purpose of imposing range constraints on it. Two requirements need to be met here: on the one hand, the initial inversion model parameter must be limited to a positive value, and on the other hand, its upper and lower boundaries need to be limited. That is to say, an intermediate function must be found to map the initial inversion model parameter from to , with and as the upper boundary and the lower boundary respectively. Specifically, lb can be 0.05 and ub can be 0.95.

[0092] By processing the second formula, the Jacobian matrix is obtained:

[0093]

[0094] To ensure that the initial inversion model parameters are positive, the natural logarithm transformation method is used , and the first Jacobian transformation matrix is obtained which is expressed as:

[0095]

[0096] To impose upper and lower bounds on the model parameters, the following cotangent transformation method is used:

[0097]

[0098] where, to ensure the smoothness of the subsequent inversion process, is 0.95 and is 0.05.

[0099] At this time, after using the cotangent transformation method, the final Jacobian transformation matrix is expressed as:

[0100]

[0101] Therefore, in each iteration of the inversion calculation, by solving the following equation each time, the next-round model parameter matrix can be obtained , and through the final Jacobian transformation matrix, the initial inversion model space parameters are mapped within the preset range:

[0102]

[0103] The advantage of such a setting is that: through the natural logarithm transformation and cotangent transformation of the Jacobian matrix, this method realizes the efficient mapping and update of the initial inversion model space parameters, significantly improving the inversion accuracy, calculation stability and convergence speed.

[0104] S103. Construct a ground nuclear magnetic resonance inversion objective function according to the ground detected nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the model update parameters of this round of iteration.

[0105] Among them, the ground nuclear magnetic resonance inversion objective function is an important tool for extracting the physical properties of underground media from the ground detected nuclear magnetic resonance observation data envelope. It provides a quantitative way to measure the difference between the model prediction and the actual observation data (i.e., the ground detected nuclear magnetic resonance observation data envelope). By minimizing this difference, the inversion process can estimate the model parameters that best fit the observation data.

[0106] S104. Solve the ground nuclear magnetic resonance inversion objective function by the quasi-Newton method to obtain the model update parameters for the next round of iteration;

[0107] Among them, the quasi - Newton method is a class of iterative algorithms used to find local minima in unconstrained optimization problems. They are variants of the Newton method and are mainly used to optimize multivariable functions. The Newton method requires calculating the second - order derivative (i.e., the Hessian matrix), while the quasi - Newton method avoids directly calculating the second - order derivative by gradually constructing an approximate Hessian matrix, thereby reducing the computational complexity.

[0108] S105. If the difference between the model update parameters of the current iteration and the model update parameters of the next iteration is less than the preset error, then use the model update parameters of the next iteration as the inversion result.

[0109] Among them, the fact that the difference between the model update parameters of the current iteration and the model update parameters of the next iteration is less than the preset error represents that the convergence condition is reached. The change in the model parameters is already very small, indicating that the algorithm may have approached the optimal solution or a local optimal solution. Therefore, the model update parameters of the next iteration can be used as the inversion result.

[0110] In the above - mentioned embodiments of the present application, obtaining the ground - penetrating nuclear magnetic resonance observation data envelope and the ground nuclear magnetic resonance forward - modeling data envelope signal includes:

[0111] Collecting groundwater measurement data through a ground nuclear magnetic resonance water - exploration instrument;

[0112] Pre - processing the groundwater measurement data to obtain the ground - penetrating nuclear magnetic resonance observation data envelope; where the pre - processing includes noise suppression, envelope extraction, frequency correction, phase correction, and data decimation;

[0113] Obtaining the water content, relaxation time, and any spatial positions of the underground medium to be analyzed;

[0114] According to the water content, relaxation time, and any spatial positions, obtaining the ground nuclear magnetic resonance forward - modeling data envelope signal.

[0115] Among them, noise suppression is an important task in signal processing, aiming to remove or reduce unwanted noise components from the signal while retaining as much useful signal information as possible.

[0116] Envelope extraction is a technique in signal processing used to obtain the envelope curve of a signal, that is, the contour of the signal amplitude changing with time. This is very useful in fields such as analyzing amplitude - modulated signals, seismic data processing, and biomedical signal processing (such as electrocardiograms and electromyograms).

[0117] Frequency correction refers to the process of adjusting the frequency components of a signal to correct or compensate for frequency deviations.

[0118] Phase correction refers to the process of adjusting the phase components of a signal to correct or compensate for phase deviations.

[0119] Data extraction refers to the method of extracting specific data from a set of data, belonging to the field of signal processing and used to exclude irrelevant or interfering data.

[0120] Collect groundwater measurement data through a ground nuclear magnetic resonance instrument, preprocess the groundwater measurement data, and obtain a ground exploration nuclear magnetic resonance observation data envelope.

[0121] According to the preset forward formula, calculate the ground nuclear magnetic resonance forward data envelope signal d based on the water content, relaxation time, and any spatial position of the underground medium to be analyzed.

[0122] The preset forward formula is:

[0123]

[0124] Among them, is the excitation pulse moment; is the excitation pulse time; r represents any spatial position; represents the kernel function. The main role of the kernel function is to map the data to a high-dimensional space, making the data that is non-linearly separable in the original space become linearly separable in the high-dimensional space. The kernel function can avoid explicitly calculating the coordinates in the high-dimensional space, thereby reducing the computational complexity; and respectively refer to the water content and relaxation time.

[0125] The advantage of such a setting is that through precise data processing and analysis, reliable groundwater information is provided, which is beneficial to improving the accuracy of subsequent inversion results.

[0126] In the above embodiments of the present application, according to the ground exploration nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the current round of iterative model update parameters, construct a ground nuclear magnetic resonance inversion objective function, including:

[0127] According to the ground exploration nuclear magnetic resonance observation data envelope and the ground nuclear magnetic resonance forward data envelope signal, obtain a data error function;

[0128] According to the initial inversion model space parameters and the model smoothness matrix, obtain a model constraint function;

[0129] Through the Tikhonov regularization solution standard form, the data error function, the model constraint function, and the regularization coefficient, obtain a first objective function;

[0130] Organize the first objective function into a standard iterative solution format to construct a ground nuclear magnetic resonance inversion objective function.

[0131] Among them, the data error function Characterize the error degree between the observed data envelope of surface nuclear magnetic resonance sounding and the forward modeling data envelope signal of surface nuclear magnetic resonance sounding.

[0132] Model constraint function Characterize that the value of the initial inversion model needs to be constrained within a preset range.

[0133] Data error function is:

[0134]

[0135] where represents the noise weighting coefficient matrix of the observed data, which is calculated from the covariance matrix of the observed data envelope of surface nuclear magnetic resonance sounding and the forward modeling data envelope signal d

[0136] Model constraint function is:

[0137]

[0138] where is the model smoothness matrix, is the spatial parameter of the initial inversion model.

[0139] According to the Tikhonov regular solution standard form of the ill-posed inverse problem (Tikhonov regular solution standard form), the first objective function is composed of the data error function and the model constraint function and is balanced by the regularization coefficient to obtain the first objective function as:

[0140]

[0141] To make the solved function value as close as possible to the minimum value of the objective function, the first objective function is arranged into the standard iterative solution format as:

[0142]

[0143] where 、 and respectively represent the model parameter value of the k-th iteration (i.e., the spatial parameter of the inversion model in the k-th round), the search step size, and the search direction.

[0144] The advantage of such a setting is that: the constructed inversion objective function of surface nuclear magnetic resonance sounding can effectively combine the observed data and the model constraints to achieve accurate inversion of the underground structure. The technical effect of this method is to improve the stability and reliability of the inversion results, and at the same time, it can handle noise and ill-posed problems.

[0145] In the above embodiments of the present application, the quasi-Newton method is used to solve the ground nuclear magnetic resonance inversion objective function to obtain the updated parameters of the next-round iteration model, including:

[0146] Perform partial derivative processing on the ground nuclear magnetic resonance inversion objective function to obtain the gradient expression of the objective function;

[0147] Use the quasi-Newton method to solve the gradient expression of the objective function according to the gradient expression of the objective function to obtain the updated parameters of the next-round iteration model.

[0148] Among them, the quasi-Newton method is a class of iterative algorithms used to find local minima in unconstrained optimization problems. They are variants of the Newton method and are mainly used to optimize multivariable functions. The Newton method requires the calculation of second-order derivatives (i.e., the Hessian matrix), while the quasi-Newton method avoids directly calculating second-order derivatives by gradually constructing an approximate Hessian matrix, thereby reducing the computational complexity.

[0149] Perform partial derivative processing on the ground nuclear magnetic resonance inversion objective function to obtain the gradient expression of the objective function as:

[0150]

[0151] Among them, T represents the transpose operation of the matrix.

[0152] Select the number of vector pairs to be 20, and set the initial model parameters .

[0153] Calculate the initial gradient .

[0154] Calculate the initial Hessian surrogate matrix :

[0155]

[0156] Among them, , , is the identity matrix.

[0157] Use the Two-loop recursion algorithm according to the initial gradient , that is, the search direction in the first round is ( ), and subsequent iterations are performed here to determine the search direction of the k-th iteration: ;

[0158]

[0159] Among them, , 。

[0160] Use a line search method with uniform gradient conditions to obtain the optimal step size , by solving the ground nuclear magnetic resonance inversion objective function, which satisfies:

[0161]

[0162] Among them, is the parameter for updating the model in the next iteration.

[0163] Among them, the uniform gradient conditions are composed of the following two equations:

[0164]

[0165]

[0166] is the uniform scaling factor, generally taken as 0.9. The uniform gradient conditions strictly limit the descent direction and gradient relationship between two adjacent iterations, thus ensuring that the selected step size approaches = 0 faster, improving the convergence speed and reducing the calculation time.

[0167] Judge the size relationship between k and n. If , then update the vector pair , and save the vector pairs of the last n rounds.

[0168] The advantages of such settings are as follows: improving the convergence speed: The quasi-Newton method achieves fast convergence by approximating the Hessian matrix and is usually more efficient than the simple gradient descent method. Reducing the computational complexity: There is no need to explicitly calculate the Hessian matrix, reducing the computational complexity, especially suitable for large-scale problems.

[0169] In the above embodiments of the present application, it further includes:

[0170] If the difference between the model update parameter in this iteration and the model update parameter in the next iteration is greater than the preset error, then execute the step of constructing the ground nuclear magnetic resonance inversion objective function according to the ground exploration nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward modeling data envelope signal, and the model update parameter in the next iteration.

[0171] Among them, if the difference between the model update parameter in this iteration and the model update parameter in the next iteration is greater than the preset error, then use the model update parameter in the next iteration as the model update parameter in this iteration, and execute the step of constructing the ground nuclear magnetic resonance inversion objective function according to the ground exploration nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward modeling data envelope signal, and the model update parameter in this iteration, until the difference between the model update parameter in this iteration and the model update parameter in the next iteration is less than the preset error.

[0172] The advantage of such a setting is that the ground nuclear magnetic resonance inversion process can more effectively analyze the underground structure and provide more reliable geological information.

[0173] The ground nuclear magnetic resonance groundwater inversion method, device, equipment, and storage medium provided in this application obtain the ground exploration nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward modeling data envelope signal, and the initial inversion model space parameters, and then constrain the initial inversion model space parameters within a numerical range that meets hydrogeological significance. On the one hand, by narrowing the numerical range, the efficiency and stability of the inversion are improved. On the other hand, the occurrence of errors can be reduced. Then, according to the ground exploration nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward modeling data envelope signal, and the updated parameters of the current round of iterative model, a ground nuclear magnetic resonance inversion objective function is constructed, and the quasi-Newton method is used to solve the ground nuclear magnetic resonance inversion objective function to improve the inversion efficiency through the quasi-Newton method.

[0174] Figure 2 It is a schematic flowchart of another ground nuclear magnetic resonance groundwater inversion method provided in the embodiments of this application. As shown in Figure 5, the method includes:

[0175] Step 1: Perform preprocessing processes such as noise suppression, envelope extraction, frequency / phase correction, and data trace extraction on the field measurement data collected by a ground nuclear magnetic resonance (SNMR) water detector to obtain the corresponding observation data envelope ;

[0176] Step 2: Calculate the SNMR forward modeling data (i.e., the forward modeling data envelope signal );

[0177] Among them, the envelope signal is calculated by the following formula:

[0178] (1)

[0179] Among them, is the excitation pulse moment; is the excitation pulse time; r represents any position in space; represents the kernel function. The main role of the kernel function is to map the data into a high-dimensional space, making the data that is non-linearly separable in the original space become linearly separable in the high-dimensional space. The kernel function can avoid explicitly calculating the coordinates in the high-dimensional space, thereby reducing the computational complexity; and correspond to the water content and relaxation time respectively.

[0180] Step 3: Perform variable substitution on the inverted model space parameters and constrain them within a reasonable range;

[0181] Among them, for the inversion model space parameters The method of performing variable substitution and constraining them within a reasonable range is as follows:

[0182] Express the SNMR envelope signal in Equation (1) as a discrete matrix form:

[0183] (2)

[0184] Among them, represents the Jacobian matrix; the superscripts and represent the real part and the imaginary part respectively; the subscripts and represent the partial derivatives of taking the water content and the relaxation time with respect to respectively, is the Jacobian matrix with respect to the water content, is the Jacobian matrix with respect to the relaxation time.

[0185] The Jacobian matrices with respect to the water content and the relaxation time are respectively expressed as:

[0186] (3)

[0187] Convert Equation (2) into the form of a linear differential equation, that is, only consider the derivative of the calculation data with respect to the model:

[0188] (4)

[0189] Among them, is an intermediate variable used for differential calculation, and m is the inversion model space parameter.

[0190] Perform variable substitution on the model parameters to achieve the purpose of imposing range constraints on the model. This transformation needs to meet two requirements: on the one hand, the model parameters need to be limited to positive values, and on the other hand, their upper and lower boundaries need to be limited. That is to say, an intermediate function must be found that can map to , and use and as the upper and lower boundaries respectively. The Jacobian matrix containing the parameter transformation is expressed as:

[0191] (5)

[0192] To ensure that the initial inversion model parameters are positive values, use the natural logarithm transformation method , and obtain the first Jacobian transformation matrix which is expressed as:

[0193] (6)

[0194] To impose upper and lower bounds on the model parameters, the following cotangent transformation method is used:

[0195] (7)

[0196] At this time, after using the cotangent transformation method, the final Jacobian transformation matrix is expressed as:

[0197] (8)

[0198] Therefore, in each iteration of the inversion calculation, solving the following equation each time can obtain the model parameter matrix of the next round , and by the final Jacobian transformation matrix, the initial inversion model space parameters can be mapped within the preset range:

[0199] (9)

[0200] Step 4: Construct the SNMR objective function, combine the observed data envelope and the forward data envelope signal in Steps 1 and 2, as well as the updated parameters after variable substitution in Step 3, to construct the objective function of SNMR inversion (i.e., the SNMR objective function), calculate its corresponding gradient , and express it in the standard iterative solution form;

[0201] Among them, the acquisition of the objective function gradient depends on the construction of the objective function. According to the standard form of the Tikhonov regular solution of the ill-posed inverse problem, the SNMR objective function is composed of the data error function and the model constraint function , and the regularization coefficient balances the weights of the two. Substituting the logarithmically transformed model parameters into it, the updated expression of the objective function is:

[0202] (10)

[0203] Among them, is the initial inversion model space parameter, is the model smoothness matrix, represents the noise weighting coefficient matrix of the observed data, which is calculated from the covariance matrix of the observed data envelope and the envelope signal d of the SNMR forward data; is the model smoothness matrix.

[0204] Take the partial derivative of (10) to obtain the gradient expression of the objective function as follows:

[0205] (11)

[0206] where T represents the transpose operation of the matrix.

[0207] To make the solved function value as close as possible to the minimum value of the objective function, the first objective function is arranged into the standard iterative solution format as follows:

[0208] (12)

[0209] where 、 and represent the model parameter values, search step sizes, and search directions at the k-th iteration, respectively

[0210] Step 5: Solve the iterative direction of parameter update based on the quasi-Newton method, and use line search with uniform gradient conditions to find the optimal step size. Specifically, use the quasi-Newton method to perform iterative optimization on the objective function until the error between and is less than the set error, and obtain the inversion results of water content and relaxation time.

[0211] Figure 3 is the flow chart of the solution by the quasi-Newton method provided in the embodiment of the present application. As Figure 3 shown, the method includes:

[0212] Step a: Initialize n and . Specifically, select the number of vector pairs to be 20, and set the initial model parameter ;

[0213] Step b: Calculate , specifically, calculate the gradient corresponding to the initial model parameter according to the gradient expression of the objective function ;

[0214] Step c: Calculate the initial substitution matrix ;

[0215] (13)

[0216] where , , is the identity matrix;

[0217] Step d: Use the Two-loop recursion algorithm according to the initial gradient , that is, the search direction in the first round is ( ), and then iterate on this to determine the search direction for the k-th iteration, thereby determining the search direction for the k-th iteration ;

[0218] (14)

[0219] where, , .

[0220] Step e: Use a line search method with uniform gradient conditions to obtain the optimal step size , which satisfies: ;

[0221] Step f: Judge the magnitude relationship between k and n. If , then update the vector pair , and save the vector pairs of the most recent n rounds. If k < n, then directly calculate , and save the vector pairs of the most recent n times.

[0222] Among them, the uniform gradient conditions in step e are composed of the following two equations:

[0223] (15)

[0224] (16)

[0225] is the uniform scaling factor, generally taken as 0.9. The uniform gradient conditions strictly limit the descent direction and gradient relationship between two adjacent iterations, thereby ensuring that the selected step size approaches = 0 faster, improving the convergence speed and reducing the calculation time

[0226] where, is the uniform scaling factor, generally taken as 0.9. The uniform gradient conditions strictly limit the descent direction and gradient relationship between two adjacent iterations, thereby ensuring that the selected step size approaches = 0 faster, improving the convergence speed and reducing the calculation time.

[0227] Step 6: Calculate the inversion result according to the optimal step size obtained in step 5. If the iteration stop condition is satisfied, perform SNMR two-dimensional high-efficiency fine imaging. If the iteration condition is not satisfied, construct the SNMR objective function for the next round. The iteration stop condition being satisfied means that the difference between the model space parameters of this round of inversion and the model space parameters of the next round is less than the preset error.

[0228] Another ground nuclear magnetic resonance groundwater inversion method provided by the embodiments of the present application can combine the natural logarithm transformation method and the tangent transformation method to constrain the model space parameters of SNMR inversion, making it conform to the actual hydrogeological significance; construct the objective function of SNMR inversion, and give the corresponding SNMR gradient expression for the calculation of the Hessian substitution matrix in the subsequent quasi-Newton method; propose to use a line search method with uniform gradient conditions to obtain the optimal step size, which significantly speeds up the optimization process of the quasi-Newton method for the objective function. In addition, the successful application of the method of the present invention provides a new idea for high-dimensional precise and rapid imaging of SNMR, and also provides strong support for expanding the application range of the SNMR method.

[0229] Figure 4a It is the two-dimensional water content distribution map of the simulation experiment provided by the embodiments of the present application. Figure 4b It is the two-dimensional relaxation time distribution map of the simulation experiment provided by the embodiments of the present application, which includes the first aquifer, the second aquifer, the third aquifer and the clay layer. Combining Figure 4a and Figure 4b it can be seen that the depth of the first aquifer is 2 - 7m, the water content is 0.2m 3 / m 3 , the relaxation time is 0.15s, simulating a coarse sand layer. The depth of the second aquifer is 7m - 16m, the water content is 0.4 m 3 / m 3 , the relaxation time is 0.3s, simulating a gravel layer. The first aquifer and the second aquifer are the main detection areas, and the accuracy of their morphology, water content and relaxation time is used to evaluate the quality of the inversion results. The depth of the third aquifer is 21m - 28m, the water content is 0.2m 3 / m 3 (that is, in every cubic meter of soil or medium, there is 0.2 cubic meters of water), the relaxation time is 0.2s, simulating an interlayer of fine sand and medium sand, and its setting purpose is to test the detection depth of the coil. The part between the second aquifer and the third aquifer is the clay layer. Although the water volume is rich (0.45m 3 / m 3 ), the relaxation time corresponding to the water signal is only 0.01s, and it cannot be detected by the SNMR method at present.

[0230] To verify that the ground nuclear magnetic resonance groundwater inversion method provided by the present application has a better imaging effect on the water-bearing structure in the real underground environment than the simulation experiment of the prior art, Figure 5a It is the two-dimensional imaging result map of water content provided by the embodiments of the present application. Figure 5b It is the two-dimensional imaging result map of relaxation time provided by the embodiments of the present application. Combining Figure 5a and Figure 5b Combining, Figure 4a and Figure 4bCombining, it can be seen that accurate inversion results can be obtained in all three aquifers by the method of this application, and the boundaries between the aquifers can be significantly distinguished. Through multiple experiments, it is found that the proposed method not only improves the inversion accuracy of deep aquifers, but also significantly improves the iterative solution efficiency of SNMR inversion, saving nearly two-thirds of the time on average.

[0231] Figure 6 FIG. is a structural example diagram of the ground nuclear magnetic resonance groundwater inversion device provided by an embodiment of this application. As Figure 6 shown, the ground nuclear magnetic resonance groundwater inversion device 60 includes: an acquisition module 601, a first obtaining module 602, a construction module 603, a second obtaining module 604, and a module 605. Among them:

[0232] The acquisition module 601 is used to acquire the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the initial inversion model space parameters; wherein, the ground detection nuclear magnetic resonance observation data envelope represents the actually collected groundwater characteristic data, the ground nuclear magnetic resonance forward data envelope signal represents the theoretical numerical value of the underground medium obtained according to simulation calculation, and the initial inversion model space parameters represent the physical characteristics initially describing the aquifer and the surrounding geological structure;

[0233] The first obtaining module 602 is used to map the initial inversion model space parameters within a preset range to obtain the updated model parameters for this round of iteration; wherein, the preset range is a numerical range that satisfies the hydrogeological significance;

[0234] The construction module 603 is used to construct the ground nuclear magnetic resonance inversion objective function according to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the updated model parameters for this round of iteration;

[0235] The second obtaining module 604 is used to solve the ground nuclear magnetic resonance inversion objective function by the quasi-Newton method to obtain the updated model parameters for the next round of iteration;

[0236] The module 605 is used to, if the difference between the updated model parameters for this round of iteration and the updated model parameters for the next round is less than the preset error, use the updated model parameters for the next round as the inversion result.

[0237] In a possible implementation manner, the acquisition module 601 is used for:

[0238] Collect groundwater measurement data through a ground nuclear magnetic resonance water exploration instrument;

[0239] Preprocess the groundwater measurement data to obtain the ground detection nuclear magnetic resonance observation data envelope; wherein, the preprocessing includes noise suppression, envelope extraction, frequency correction, phase correction, and data decimation;

[0240] Obtain the water content, relaxation time, and any arbitrary spatial position of the subsurface medium to be analyzed;

[0241] Based on the water content, relaxation time, and any arbitrary spatial position, obtain the forward modeling envelope signal of ground nuclear magnetic resonance.

[0242] In a possible implementation, the obtaining module 601 is further configured to:

[0243] Based on the observed envelope of ground probing nuclear magnetic resonance and the forward modeling envelope signal of ground nuclear magnetic resonance, obtain a data error function;

[0244] Based on the initial inversion model space parameters and the model smoothness matrix, obtain a model constraint function;

[0245] Through the standard form of the Tikhonov regular solution, the data error function, the model constraint function, and the regularization coefficient, obtain a first objective function;

[0246] Organize the first objective function into a standard iterative solution format to construct the inversion objective function of ground nuclear magnetic resonance.

[0247] In a possible implementation, the first obtaining module 602 is configured to:

[0248] Express the forward modeling envelope signal of ground nuclear magnetic resonance in a discrete matrix form to obtain a first formula;

[0249] Convert the first formula into a linear differential equation form to obtain a second formula; wherein, the second formula is used to calculate the derivative of the model update parameter for this round of iteration;

[0250] Through the natural logarithm transformation method and the cotangent transformation method, calculate the second formula to map the initial inversion model space parameters within a preset range to obtain the model update parameter for this round of iteration

[0251] In a possible implementation, the first obtaining module 602 is further configured to:

[0252] Based on the second formula, obtain a first Jacobian matrix;

[0253] Through the natural logarithm transformation method, process the first Jacobian matrix to obtain a second Jacobian matrix;

[0254] Through the cotangent transformation method, process the second Jacobian matrix to obtain a final Jacobian matrix;

[0255] Based on the final Jacobian matrix, map the initial inversion model space parameters within a preset range.

[0256] In a possible implementation, the second obtaining module 604 is configured to:

[0257] Take the partial derivative of the ground nuclear magnetic resonance inversion objective function to obtain the gradient expression of the objective function;

[0258] Use the quasi-Newton method to solve the gradient expression of the objective function according to the gradient expression of the objective function to obtain the update parameters of the next iteration model.

[0259] In a possible implementation, the module 605 is used for:

[0260] If the difference between the update parameters of the current iteration model and the update parameters of the next iteration model is greater than the preset error, then execute the step of constructing the ground nuclear magnetic resonance inversion objective function according to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward data envelope signal, and the update parameters of the next iteration model.

[0261] The device provided by the embodiments of the present application can execute the technical solutions shown in the above method embodiments, and its implementation principle and beneficial effects are similar, and will not be described in detail here.

[0262] It should be noted that it should be understood that the division of each module of the above device is only a logical function division. In actual implementation, it can be fully or partially integrated into a physical entity, or physically separated. And these modules can all be implemented in the form of software called by processing elements; they can also all be implemented in hardware; or some modules can be implemented in the form of software called by processing elements, and some modules can be implemented in hardware. For example, the processing module can be a separately established processing element, or can be integrated in a certain chip of the above device. In addition, it can also be stored in the memory of the above device in the form of program code, and called and executed by a certain processing element of the above device to perform the functions of the above processing module. The implementation of other modules is similar. In addition, all or part of these modules can be integrated together or can be independently implemented. Here, the processing element can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed by the integrated logic circuit in the processor element or the instruction in the form of software.

[0263] For example, the above modules can be one or more integrated circuits configured to implement the above methods, such as: one or more Application Specific Integrated Circuits (ASICs), or, one or more Digital Signal Processors (DSPs), or, one or more Field Programmable Gate Arrays (FPGAs), etc. Again, when a certain above module is implemented in the form of a processing element scheduling program code, the processing element can be a general-purpose processor, such as a Central Processing Unit (CPU) or other processors that can call program code. Again, these modules can be integrated together and implemented in the form of a System-On-a-Chip (SOC).

[0264] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer instructions are loaded and executed on a computer, the processes or functions according to the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as coaxial cable, optical fiber, Digital Subscriber Line (DSL)) or wireless (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that includes one or more integrated available media. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a Digital Video Disc (DVD)), or a semiconductor medium (for example, a solid state disk (SSD)), etc.

[0265] Figure 7 Schematic structural diagram of the electronic device provided by the embodiments of the present application. As Figure 7 shown, the electronic device 70 includes:

[0266] The electronic device 70 may include components such as a processor 701 with one or more processing cores, a memory 702 with one or more computer-readable storage media, and a communication component 703. Among them, the processor 701, the memory 702, and the communication component 703 are connected through a bus 704.

[0267] In a specific implementation process, at least one processor 701 executes the computer-executable instructions stored in the memory 702, so that at least one processor 701 executes a ground nuclear magnetic resonance groundwater inversion method as described above.

[0268] For the specific implementation process of the processor 701, reference can be made to the above method embodiments, and their implementation principles and technical effects are similar, so they will not be elaborated here in this embodiment.

[0269] In the above Figure 7 In the illustrated embodiment, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules in the processor.

[0270] The memory may include a high-speed random access memory (RAM), and may also include non-volatile memory (NVM), such as at least one disk memory.

[0271] The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the buses in the drawings of this application are not limited to only one bus or one type of bus.

[0272] In some embodiments, a computer program product is also proposed, including a computer program or instructions, and when the computer program or instructions are executed by a processor, the steps in any of the above ground nuclear magnetic resonance groundwater inversion methods are implemented.

[0273] For the specific implementation of each of the above operations, reference can be made to the previous embodiments and will not be elaborated herein.

[0274] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructions or by controlling related hardware through instructions. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0275] Therefore, an embodiment of the present application provides a computer-readable storage medium, which stores multiple instructions that can be loaded by a processor to execute the steps in any of the ground nuclear magnetic resonance groundwater inversion methods provided by the embodiments of the present application.

[0276] Among them, the storage medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk, etc.

[0277] According to one aspect of the present application, there is provided a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium.

[0278] Since the instructions stored in the storage medium can execute the steps in any of the ground nuclear magnetic resonance groundwater inversion methods provided by the embodiments of the present application, the beneficial effects achievable by any of the ground nuclear magnetic resonance groundwater inversion methods provided by the embodiments of the present application can be realized. For details, refer to the previous embodiments and will not be elaborated herein.

[0279] Those skilled in the art will readily think of other implementation manners of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any variations, uses, or adaptations of the present application, which follow the general principles of the present application and include common general knowledge or conventional technical means in the technical field not disclosed in the present application. The specification and the embodiments are only regarded as exemplary, and the true scope and spirit of the present application are pointed out by the following claims.

[0280] It should be understood that the present application is not limited to the precise structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope. The scope of the present application is only limited by the appended claims.

Claims

1. A ground nuclear magnetic resonance groundwater inversion method, characterized in that Including: Obtaining a ground detection nuclear magnetic resonance observation data envelope, a ground nuclear magnetic resonance forward modeling data envelope signal, and initial inversion model space parameters; wherein, the ground detection nuclear magnetic resonance observation data envelope represents the actually collected groundwater characteristic data, the ground nuclear magnetic resonance forward modeling data envelope signal represents the theoretical numerical values of the underground medium obtained according to simulation calculations, and the initial inversion model space parameters represent the physical characteristics initially describing the aquifer and the surrounding geological structure; Mapping the initial inversion model space parameters within a preset range to obtain the model update parameters for this round of iteration; wherein, the preset range is a numerical range that satisfies hydrogeological significance; Constructing a ground nuclear magnetic resonance inversion objective function according to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward modeling data envelope signal, and the model update parameters for this round of iteration; Solving the ground nuclear magnetic resonance inversion objective function by the quasi-Newton method to obtain the model update parameters for the next round of iteration; If the difference between the model update parameters for this round of iteration and the model update parameters for the next round is less than a preset error, then taking the model update parameters for the next round as the inversion result.

2. The method according to claim 1, wherein The mapping the initial inversion model space parameters within a preset range to obtain the model update parameters for this round of iteration includes: Expressing the ground nuclear magnetic resonance forward modeling data envelope signal in the form of a discrete matrix to obtain a first formula; Converting the first formula into the form of a linear differential equation to obtain a second formula; wherein, the second formula is used to calculate the derivative of the model update parameters for this round of iteration; Calculating the second formula by the natural logarithm transformation method and the cotangent transformation method to map the initial inversion model space parameters within the preset range to obtain the model update parameters for this round of iteration.

3. The method according to claim 2, wherein The calculating the second formula by the natural logarithm transformation method and the cotangent transformation method to map the initial inversion model space parameters within the preset range to obtain the model update parameters for this round of iteration includes: Obtaining a first Jacobian matrix according to the second formula; Processing the first Jacobian matrix by the natural logarithm transformation method to obtain a second Jacobian matrix; Processing the second Jacobian matrix by the cotangent transformation method to obtain a final Jacobian matrix; Mapping the initial inversion model space parameters within the preset range according to the final Jacobian matrix.

4. The method according to any one of claims 1 to 3, characterized in that The obtaining the ground detection nuclear magnetic resonance observation data envelope and the ground nuclear magnetic resonance forward modeling data envelope signal includes: Collecting groundwater measurement data by a ground nuclear magnetic resonance water exploration instrument; Preprocessing the groundwater measurement data to obtain the ground detection nuclear magnetic resonance observation data envelope; wherein, the preprocessing includes noise suppression, envelope extraction, frequency correction, phase correction, and data trace extraction; Obtaining the water content, relaxation time, and each spatial arbitrary position of the underground medium to be analyzed; Obtaining the ground nuclear magnetic resonance forward modeling data envelope signal according to the water content, the relaxation time, and each spatial arbitrary position.

5. The method according to any one of claims 1 to 3, characterized in that Constructing a ground nuclear magnetic resonance inversion objective function according to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward modeling data envelope signal, and the updated parameters of the current iteration model includes: Obtaining a data error function according to the ground detection nuclear magnetic resonance observation data envelope and the ground nuclear magnetic resonance forward modeling data envelope signal; Obtaining a model constraint function according to the initial inversion model space parameters and the model smoothness matrix; Obtaining a first objective function through the standard form of the Tikhonov regular solution, the data error function, the model constraint function, and the regularization coefficient; Rearranging the first objective function into a standard iterative solution format to construct the ground nuclear magnetic resonance inversion objective function.

6. The method according to any one of claims 1 to 3, characterized in that Solving the ground nuclear magnetic resonance inversion objective function by the quasi-Newton method to obtain the updated parameters of the next iteration model, including: Performing a partial derivative process on the ground nuclear magnetic resonance inversion objective function to obtain an expression of the objective function gradient; Solving the expression of the objective function gradient by the quasi-Newton method according to the expression of the objective function gradient to obtain the updated parameters of the next iteration model.

7. The method according to any one of claims 1 to 3, characterized in that Further including: If the difference between the updated parameters of the current iteration model and the updated parameters of the next iteration model is greater than a preset error, then execute the step of constructing a ground nuclear magnetic resonance inversion objective function according to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward modeling data envelope signal, and the updated parameters of the next iteration model.

8. A ground nuclear magnetic resonance groundwater inversion device, characterized in that, Including: An acquisition module, configured to acquire a ground detection nuclear magnetic resonance observation data envelope, a ground nuclear magnetic resonance forward modeling data envelope signal, and initial inversion model space parameters; wherein, the ground detection nuclear magnetic resonance observation data envelope represents the actually collected groundwater characteristic data, the ground nuclear magnetic resonance forward modeling data envelope signal represents the theoretical values of the underground medium obtained according to simulation calculations, and the initial inversion model space parameters represent the physical characteristics initially describing the aquifer and the surrounding geological structure; A first obtaining module, configured to map the initial inversion model space parameters within a preset range to obtain the updated parameters of the current iteration model; wherein, the preset range is a numerical range that satisfies hydrogeological significance; A construction module, configured to construct a ground nuclear magnetic resonance inversion objective function according to the ground detection nuclear magnetic resonance observation data envelope, the ground nuclear magnetic resonance forward modeling data envelope signal, and the updated parameters of the current iteration model; A second obtaining module, configured to solve the ground nuclear magnetic resonance inversion objective function by the quasi-Newton method to obtain the updated parameters of the next iteration model; A determination module, configured to, if the difference between the updated parameters of the current iteration model and the updated parameters of the next iteration model is less than a preset error, then use the updated parameters of the next iteration model as the inversion result.

9. An electronic device, characterized in that, Including: A memory and a processor; The memory stores computer execution instructions; The processor executes the computer execution instructions stored in the memory, so that the processor executes the method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions that, when executed by a processor, are used to implement the method according to any one of claims 1-7.

Images