Time shift resistivity method and time shift induced polarization method four-dimensional joint inversion method

Through the four-dimensional joint inversion method of time shift resistivity method and time shift excitation polarization method, the cross gradient term and time shift term are used to solve the problem that polarization result depends on a single resistivity in traditional inversion, and the underground medium is improved The similarity of the model and the credibility and resolution of the inversion results.

CN119045072AInactive Publication Date: 2024-11-29OIL & GAS SURVEY CGS +1

Patent Information

Application Number
CN202411423175.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-12
Publication Date
2024-11-29
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In the inversion of traditional resistivity and excitation polarization, the polarization result is overly dependent on the single resistivity result, lacking constraint information, which reduces the credibility of the inversion result, and the resistivity and polarization distribution of underground dielectrics are complex. , resulting in difficulty in interpreting in spatial distribution and structural features.

Method used

The four-dimensional inversion method is adopted for the time shift resistivity method and the time shift excitation polarization method. The cross gradient term is calculated by the cross gradient function, combined with the time shift term and the cross gradient term, the weight factor is set, and the alternating iteration is performed to improve the resistivity and pole. The structural similarity of the rate model ensures the temporal continuity of the inversion result and anti-interference ability.

Benefits of technology

The structural similarity and resolution of the resistivity and polarization model are improved, the credibility and anti-interference ability of the inversion results are enhanced, and the continuity of the inversion results are ensured at different moments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119045072A_ABST
    Figure CN119045072A_ABST
Patent Text Reader

Abstract

The invention provides a time shift resistivity method and time shift induced polarization method four-dimensional joint inversion method. The method comprises the following steps: preprocessing observation data; reading a file required by inversion; performing time shift inversion through a resistivity method and an induced polarization method, and calculating to obtain a time shift resistivity model and a time shift polarizability model with abnormal forms; calculating a cross gradient term between the time-shifted resistivity model and the time-shifted polarizability model; calculating to obtain a resistivity and polarizability inversion result with cross gradient constraint; calculating cross gradient terms for multiple times, performing alternate inversion iteration, and updating the resistivity model and the polarizability model until an iteration termination condition is reached; and calculating a cross gradient mean value of inversion results of the time-shifted resistivity and the time-shifted polarizability, a difference value between the inversion results and a real model and a model difference value of adjacent moments so as to evaluate the inversion results. The method is reasonable in conception, and has higher resolution and extremely strong anti-interference capability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of geophysical exploration, and in particular to a four-dimensional joint inversion method of a time-shift resistivity method and a time-shift induced polarization method. Background Art

[0002] Resistivity and induced polarization are two branch methods of electrical prospecting. The two methods are often combined to solve problems such as mineral exploration, hydrogeology and engineering geology. When solving problems such as environmental protection, ecological restoration, seawater intrusion, geological disasters, etc., long-term monitoring is often required, that is, multiple observations of the same underground area. In addition to the three-dimensional spatial dimension, the monitoring data also contains the time dimension. In recent years, domestic and foreign scholars have realized the time-shift resistivity inversion and time-shift induced polarization inversion of monitoring data based on traditional algorithms. Time-shift inversion can ensure the temporal continuity of the inversion results at different times. It is superior to traditional methods in processing monitoring data, especially in eliminating background noise interference. It has significant advantages.

[0003] The objective functions of the traditional resistivity method and the induced polarization method inversion are:

[0004]

[0005] It can be seen that the objective function contains the data term Φ d and the model term Φ m Two parts, where the data item Φ d Used to calculate the difference between the observed data and the inversion result response data, the model term Φ m Used to calculate the model smoothness, and It is the weight factor of the model term, which is used to adjust the proportion of the data term and the model term in the objective function.

[0006] In the previous inversion methods, the resistivity was inverted first and then the polarizability was inverted, that is, formula (01) was minimized first and then formula (02), which resulted in the polarizability result being overly dependent on the single resistivity result. In addition, the traditional inversion lacked other constraint information, which reduced the credibility of the inversion results and limited its application areas and scope.

[0007] After years of development, the separate inversion and time-shift inversion of the resistivity method and induced polarization method have become relatively mature in principle and application, but the following problems still exist: the resistivity and polarizability of underground media are complexly distributed in space, which leads to large differences in the spatial distribution, structural characteristics and other aspects of the inversion results of the resistivity method and the induced polarization method, causing certain difficulties in the interpretation of the results; in the traditional algorithm, the inversion of the induced polarization method is carried out on the basis of the inversion results of the resistivity method. Regardless of whether it is a single method inversion or a time-shift method inversion, the resistivity result will directly affect the polarizability inversion result, while the polarizability result does not constrain the resistivity inversion result, reducing the credibility of the inversion of the induced polarization method.

[0008] To sum up, it is necessary to make further innovations on existing technologies. Summary of the invention

[0009] In view of the technical problems existing in the above-mentioned background technology, the present invention proposes a four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method. The method has a reasonable conception and combines the advantages of time-shift inversion and joint inversion. It can not only ensure the continuity of inversion results at different times over time, but also improve the structural similarity of resistivity and polarizability models. It has higher resolution and strong anti-interference ability.

[0010] In order to solve the above technical problems, the present invention provides a four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method, which comprises the following steps:

[0011] (1) Preprocess the observation data;

[0012] (2) Read the files required for inversion;

[0013] (3) Time-shift inversion of apparent resistivity data and apparent polarizability data is performed using the resistivity method and induced polarization method, respectively, to obtain a time-shift resistivity model and a time-shift polarizability model with abnormal morphology;

[0014] (4) The cross-gradient term between the time-shift resistivity model and the time-shift polarizability model with abnormal morphology is calculated through the cross-gradient function, and then the cross-gradient term is substituted into the objective function of the joint inversion of the time-shift resistivity method and the time-shift induced polarization method, and a weight factor is set for the cross-gradient term. The initial model is the model after the most recent iteration. Through alternating iterations, the joint inversion results of the time-shift resistivity method and the time-shift induced polarization method with cross-gradient constraints are calculated;

[0015] (5) Repeat the above step (4), calculate the cross-gradient term and perform inversion iteration alternately, and update the time-shift resistivity model and the time-shift polarizability model until the data fitting error of the inversion results of the time-shift resistivity method and the time-shift induced polarization method is less than the expected value or the number of iterations reaches the set value, and the iteration is terminated;

[0016] (6) Calculate the cross-gradient mean of the time-shift resistivity and time-shift polarizability inversion results, the model error between the model and the true model, and the model difference at adjacent times to evaluate the inversion results.

[0017] The four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method, wherein: the observation data contains the coordinates of the transmitting electrode and the receiving electrode corresponding to all the measuring points, the apparent resistivity, the apparent polarizability and the current information; the preprocessing process includes data conversion, elimination of distortion points and adjustment of the data format; the data conversion is to obtain the original data for inversion, namely the apparent resistivity and the apparent polarizability, through the observed electric field calculation; the elimination of distortion points is to eliminate the distortion points of the observed data, and then fill the missing data points by using the interpolation method after elimination; the adjustment of the data format includes adjusting the data to a fixed format.

[0018] The four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method, wherein: the files required for the inversion in the step (2) include preprocessed observation data, parameter data and an initial model; the parameter data are mainly weight factors of model terms and time-shift terms; the initial model includes an initial time-shift resistivity model and an initial time-shift polarizability model.

[0019] The four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method, wherein the specific process of step (3) is as follows:

[0020] First, the apparent resistivity data is inverted by the time-lapse resistivity method. After multiple iterations, the time-lapse resistivity model M with an abnormal shape is obtained. ρ It is expressed as:

[0021] M ρ =[m ρ,1 m ρ,2 ... m ρ,nt ] T (1);

[0022] In the above formula (1), m ρ,nt is the resistivity model of the underground at a certain moment, nt represents the number of moments;

[0023] Then, based on the resistivity time-shift inversion results, the apparent polarizability data is inverted by time-shift induced polarization method. After multiple iterations, the time-shift polarizability model M with an abnormal shape is obtained. η It is expressed as:

[0024] M η =[m η,1 m η,2 ... m η,nt ] T (2);

[0025] In the above formula (2), m η,nt is the polarizability model of the underground at a certain moment, and nt represents the number of moments.

[0026] The four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method, wherein the specific process of step (4) is:

[0027] First, according to the gradient of the time-lapse resistivity model and the gradient of the time-shifted polarizability model The cross gradient function t is obtained by multiplication:

[0028]

[0029] The cross gradient term Φ is calculated according to the cross gradient function t cg :

[0030]

[0031] The cross-gradient term is mainly used to measure the structural similarity between the time-shift resistivity model and the time-shift polarizability model. The larger the cross-gradient term, the greater the structural difference, and the smaller the cross-gradient term, the greater the structural difference. After the cross-gradient term is calculated, the calculated cross-gradient term is introduced into the objective function of the joint inversion of the time-shift resistivity method and the time-shift induced polarization method, which is and When the objective function is minimized, the cross-gradient term will play a constraint role and promote the structural similarity of the time-shift resistivity model and the time-shift polarizability model.

[0032] The joint inversion of the time-shift resistivity method and the time-shift induced polarization method adopts the strategy of constructing two objective functions. The objective function of the joint inversion of the time-shift resistivity method and the time-shift induced polarization method is shown in the following formula (5):

[0033]

[0034] In the above formula (5), Φ(M ρ ) is the objective function of the resistivity method in the joint inversion of the time-shift resistivity method and the time-shift induced polarization method; Φ(M η ) is the objective function of the induced polarization method in the joint inversion of the time-shift resistivity method and the time-shift induced polarization method; Φ d (M ρ ) and Φ d (M η ) are the data items of the resistivity method and the induced polarization method, respectively, which are used to ensure that the difference between the forward response of the inversion result and the original data is minimized; Φ m (M ρ ) and Φ m (M η) are the model terms of the resistivity method and the induced polarization method, respectively, which are used to ensure the smoothness of the inversion results and reduce the multiple solutions of the inversion; Φ t (M ρ ) and Φ t (M η ) are the time-shift terms of the resistivity method and the induced polarization method, which are used to ensure the temporal continuity of the inversion results at different times; Φ cg (M ρ ,M η ) and Φ cg (M η ,M ρ ) are the cross gradient terms of the resistivity method and the induced polarization method, which are used to ensure that the resistivity results and the polarizability results are more similar in morphology and structure; and are the weight factors of the model term, time shift term and cross gradient term in the objective function of the resistivity method; and are the weight factors of the model term, time shift term and cross gradient term in the objective function of the induced polarization method. By adjusting the size of the weight factor, the strength of the constraint effect of each term in the inversion can be controlled to obtain the optimal inversion result; the initial weight factor of the cross gradient term is 0, and it is set to a non-zero value after a certain number of iterations;

[0035] The data item is obtained by the following formula (6):

[0036]

[0037] In the above formula (6), D ρ,obs is the apparent resistivity data observed at all times; D η,obs is the apparent polarizability data observed at all times; F(M ρ ) is the forward response function of the resistivity method; F(M η ) is the forward response function of the induced polarization method, and the forward response of the model can be calculated through the forward function; and The data covariance matrices of the resistivity method and induced polarization are respectively, and the diagonal elements are D ρ,obs and D η,obs A diagonal matrix with 5% of the elements and the remaining elements being 0;

[0038] The model term is obtained by the following formula (7):

[0039]

[0040] In the above formula (7), C M,ρ and C M,η is the model covariance matrix; M ρ,0 and M η,0They are the initial time-shift resistivity model and the initial time-shift polarizability model respectively;

[0041] The time shift term is obtained by the following formula (8):

[0042]

[0043] In the above formula (8), m ρ,k and m η,k They are the resistivity and polarizability models of the underground at time k respectively.

[0044] The four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method, wherein, in each inversion iteration process of updating the model in the step (5), it is determined whether the data fitting error is less than the expected value or the number of iterations reaches the set value, and if so, the inversion iteration is terminated, and if not, the inversion iteration is continued; if one of the resistivity method and the induced polarization method reaches the inversion iteration termination condition, the inversion iteration is terminated by the method, and the other method continues to iterate until the iteration termination condition is met;

[0045] The termination condition of the inversion iteration is that the RMS of the data fitting error calculated at all times is less than the expected value or the number of iterations reaches the set maximum number of iterations:

[0046]

[0047] In the above formula (9), RMS ρ and RMS η They are the data fitting differences of the resistivity method and the induced polarization method.

[0048] The four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method, wherein the cross gradient mean in step (6) is expressed as It is obtained by the following formula (10):

[0049]

[0050] In the above formula (10), N is the total number of underground grids.

[0051] The four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method, wherein the model error in the step (6) is expressed as δ, which is the difference between the inversion result and the true model, and the model errors of the resistivity and polarizability are both obtained by the following formula (11):

[0052]

[0053] When the resistivity model error is calculated using the above formula (11), and are the inverted resistivity value and the true resistivity value of the ith underground unit respectively; when using the above formula (11) to calculate the model error of the polarizability, and are the inverted polarizability value and the true polarizability value of the ith underground unit, respectively.

[0054] The four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method, wherein the model difference at adjacent moments in the step (6) is expressed as It is the sum of the absolute values ​​of the model differences at adjacent moments and is obtained by the following formula (12):

[0055]

[0056] In the above formula (12), m k is the model at time k, m k+1 is the model for time k+1, and nt represents the number of time moments.

[0057] By adopting the above technical solution, the present invention has the following beneficial effects:

[0058] The four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method of the present invention is reasonably conceived. The present invention synchronously inverts the apparent resistivity and apparent polarizability data at multiple times, and simultaneously obtains the resistivity and polarizability inversion results at multiple times, which belongs to a four-dimensional inversion method. The present invention adopts a strategy of alternating iteration of the time-shift resistivity method and the time-shift induced polarization method, which avoids the problem that the polarizability inversion results are overly dependent on a single resistivity model. The present invention adds time-shift terms and cross-gradient terms to the objective function, combining the advantages of time-shift inversion and joint inversion, which can not only ensure the continuity of the inversion results at different times over time, but also improve the structural similarity of the resistivity and polarizability models, and has higher resolution and extremely strong anti-interference ability. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0060] Figure 1 It is a four-dimensional joint inversion flow chart of the time-shift resistivity method and the time-shift induced polarization method involved in the four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method of the present invention;

[0061] Figure 2Schematic diagram of the time-shift resistivity model and time-shift polarizability model constructed in the four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method of the present invention, from left to right are T 1 、T 2 、T 3 Model at three moments (A has a resistivity of 100Ω·m and a polarizability of 0.01; B has a resistivity of 100Ω·m and a polarizability of 0.01; 1 and B 2 The resistivity is 10Ω·m and the polarizability is 0.2);

[0062] Figure 3 A schematic diagram of an observation system designed on the surface involved in the four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method of the present invention (the red triangles in the observation system are the positions of the transmitting electrodes, the blue dots are the positions of the receiving electrodes, and the response data of the underground model can be obtained by controlling the transmitting and receiving electrodes);

[0063] Figure 4 The comparison of resistivity inversion results of the traditional method (left) and the method of the present invention (right) at a depth of 8m ( Figure 4 In the figure, the 1st, 2nd and 3rd rows are T 1 、T 2 、T 3 Profiles of inversion results at three moments);

[0064] Figure 5 Comparison of polarizability inversion results of the traditional method (left) and the method of the present invention (right) at a depth of 8m ( Figure 5 In the figure, the 1st, 2nd and 3rd rows are T 1 、T 2 、T 3 Cross-sections of inversion results at three moments). DETAILED DESCRIPTION

[0065] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] The present invention is further explained below in conjunction with specific implementation modes.

[0067] like Figure 1 As shown, this embodiment provides a four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method, comprising the following steps:

[0068] S100, preprocessing the observed data. The observed data includes information such as the coordinates of the transmitting electrode and the receiving electrode corresponding to all the measuring points, the apparent resistivity, the apparent polarizability and the current. The preprocessing process includes data conversion, elimination of distortion points, adjustment of data format, etc.; the data conversion is to obtain the original data for inversion, namely the apparent resistivity and the apparent polarizability, through the observed electric field calculation; the elimination of distortion points is to eliminate the distortion points of the observed data, and then fill the missing data points by interpolation after elimination; the adjustment of data format includes adjusting the data to a fixed format.

[0069] In this embodiment, the forward response of the model is used as the observation data. Figure 2 As shown, the model contains T 1 、T 2 、T 3 Data at three moments. The resistivity of surrounding rock A is 100Ω·m, the polarizability is 0.01, and the abnormal body B is 1 , B 2 The resistivity is 10Ω·m and the polarizability is 0.2. Figure 3 This is a schematic diagram of the observation system. Data is collected by arranging transmitting and receiving electrodes on the surface. The electrodes are arranged in a rectangular array. The red triangles are the locations of the transmitting electrodes, the blue dots are the locations of the receiving electrodes, and the black solid line box is the anomaly B. 1 , B 2 Projection on the surface. When each transmitting electrode is powered, all adjacent receiving electrodes in the X direction receive simultaneously to ensure that enough information in the three-dimensional space is obtained. After forward modeling, 5%, 5%, and 20% Gaussian random errors are added to the observation data at T1, T2, and T3 respectively to simulate the noise interference in the actual acquisition process, especially the addition of 20% high error at the third moment to simulate the situation with large noise. The data after adding Gaussian random errors are used as observation data.

[0070] S200, read the files required for inversion. The files include the processed observation data, parameter data, initial model, etc. The parameter data is mainly the weight factor of the model term and the time-shift term; the initial model includes the initial time-shift resistivity model and the initial time-shift polarizability model. In this embodiment, the resistivity value (100Ω·m) and polarizability value (0.01) of the surrounding rock A are selected as the initial model.

[0071] S300, firstly perform resistivity method and induced polarization method time-shift inversion, first perform resistivity time-shift inversion, after the resistivity time-shift inversion is performed by minimizing the objective function and performing several iterations to obtain a time-shift resistivity model with an abnormal morphology, then perform polarizability time-shift inversion based on the time-shift resistivity inversion result to obtain a time-shift polarizability model with an abnormal morphology; and the specific process of this step is:

[0072] First, the apparent resistivity data is inverted by the time-shift resistivity method. After multiple iterations (40 times in this embodiment), a time-shift resistivity model M with an abnormal shape is obtained. ρ It is expressed as:

[0073] M ρ =[m ρ,1 m ρ,2 ... m ρ,nt ] T (1);

[0074] In the above formula (1), m ρ,nt is the resistivity model of the underground at a certain moment, nt represents the number of moments;

[0075] Then, based on the resistivity time-shift inversion results, the apparent polarizability data is inverted by time-shift induced polarization method. After multiple iterations (40 times in this embodiment), a time-shift polarizability model M with an abnormal morphology is obtained. η It is expressed as:

[0076] M η =[m η,1 m η,2 ... m η,nt ] T (2);

[0077] In the above formula (3), m η,nt is the polarizability model of the underground at a certain moment.

[0078] S400, after the time-shift resistivity model and the time-shift polarizability model in step (3) have certain abnormal morphologies, the cross-gradient term between the latest time-shift resistivity model and the time-shift polarizability model is calculated through the cross-gradient function, and then the cross-gradient term is substituted into the time-shift joint inversion objective function, and a weight factor is set for the cross-gradient term. The initial model is the model after the most recent iteration, and the resistivity and polarizability inversion results with cross-gradient constraints are calculated through alternating iterations; and the specific process of this step is:

[0079] First, according to the gradient of the time-lapse resistivity model and the gradient of the time-shifted polarizability model The cross gradient function t is obtained by multiplication:

[0080]

[0081] The cross gradient term Φ is calculated according to the cross gradient function t cg :

[0082]

[0083] The cross-gradient term is mainly used to measure the structural similarity between the time-shift resistivity model and the time-shift polarizability model. The larger the cross-gradient term, the greater the structural difference, and the smaller the cross-gradient term, the greater the structural difference. After the cross-gradient term is calculated, the calculated cross-gradient term is introduced into the objective function of the joint inversion of the time-shift resistivity method and the time-shift induced polarization method (referred to as the time-shift joint inversion), which is and Assignment, when minimizing the objective function, the cross-gradient term will play a constraining role, promoting the structural similarity of the time-shift resistivity model and the time-shift polarizability model; Due to the constraint of the cross-gradient term, the resistivity and polarizability results obtained in this step have certain structural similarities;

[0084] The time-lapse joint inversion adopts the strategy of constructing two objective functions. The objective function of the time-lapse joint inversion is shown in the following formula (5):

[0085]

[0086] In the above formula (5), Φ(M ρ ) is the objective function of the resistivity method in the time-lapse joint inversion; Φ(M η ) is the objective function of the induced polarization method in the time-lapse joint inversion; Φ d (M ρ ) and Φ d (M η ) are the data items of the resistivity method and the induced polarization method, respectively, which are used to ensure that the difference between the forward response of the inversion result and the original data is minimized; Φ m (M ρ ) and Φ m (M η ) are the model terms of the resistivity method and the induced polarization method, respectively, which are used to ensure the smoothness of the inversion results and reduce the multiple solutions of the inversion; Φ t (M ρ ) and Φ t (M η ) are the time-shift terms of the resistivity method and the induced polarization method, which are used to ensure the temporal continuity of the inversion results at different times; Φ cg (M ρ ,M η ) and Φ cg (M η ,M ρ ) are the cross-gradient terms of the resistivity method and the induced polarization method, respectively, which are used to ensure that the resistivity results and the polarizability results are more similar in morphology and structure. and are the weight factors of the model term, time shift term and cross gradient term in the objective function of the resistivity method, and are the weight factors of the model term, time-shift term and cross-gradient term in the objective function of the induced polarization method. By adjusting the size of the weight factor, the strength of the constraint effect of each term in the inversion can be controlled to obtain the optimal inversion result. Among them, when the first 40 inversions of the time-shift resistivity method and the time-shift induced polarization method are performed separately, and When it is set to 0, the cross-gradient term is not calculated at this time, and the cross-gradient term does not constrain the inversion effect; after 40 inversion iterations, a time-shift resistivity model and a time-shift polarizability model with abnormal morphology are obtained. The initial weight factor of the cross-gradient term in this step S400 is 0, because when starting the inversion from a uniform half-space, a certain abnormal morphology must be present so that the cross-gradient term can play a better constraining role;

[0087] The calculation of the above data items is as follows:

[0088]

[0089] Among them, D in the above formula (6) ρ,obs is the apparent resistivity data observed at all times, D η,obs is the apparent polarizability data observed at all times; F(M ρ ) is the forward response function of the resistivity method; F(M η ) is the forward response function of the induced polarization method, and the forward response of the model can be calculated through the forward function; and is the data covariance matrix of the resistivity method and induced polarization, with diagonal elements D ρ,obs and D η,obs A diagonal matrix with 5% of , and the rest of the elements are 0;

[0090] The calculation of the above model terms is as follows:

[0091]

[0092] In the above formula (7), C M,ρ and C M,η is the model covariance matrix; M ρ,0 and M η,0 They are the initial time-shifted resistivity model and the initial time-shifted polarizability model, which are generally set as a uniform half-space with constant resistivity and polarizability values;

[0093] The calculation of the above time shift term is shown in formula (8):

[0094]

[0095] Among them, m in the above formula (8) ρ,k and m η,kis the resistivity and polarizability model of the underground at time k;

[0096] S500, repeating step (4), calculating the cross-gradient term multiple times and iteratively updating the time-shift resistivity model and the time-shift polarizability model alternately; in each inversion iterative model updating process, judging whether the data fitting error is less than the expected value or the number of iterations reaches the set value, if satisfied, terminating the inversion iteration, if not satisfied, continuing the inversion iteration; if one of the resistivity method and the induced polarization method meets the inversion iteration termination condition, the method terminates the inversion iteration, and the other method continues to iterate until the iteration termination condition is met;

[0097] The termination condition of the inversion iteration is that the RMS of the data fitting error calculated at all times is less than the expected value or the number of iterations reaches the set maximum number of iterations;

[0098]

[0099] Among them, RMS in the above formula (9) ρ and RMS η The data fitting of the resistivity method and the induced polarization method are poor.

[0100] S600, calculate the cross gradient mean of resistivity and polarizability inversion results The model error δ between the model and the true model, the model difference at adjacent moments The inversion results are evaluated; the cross gradient mean calculation formula is shown in formula (10):

[0101]

[0102] Wherein, N in the above formula (10) is the total number of underground grids.

[0103] The model error δ is the difference between the inversion result and the true model. The model errors of resistivity and polarizability are obtained using the following formula (11):

[0104]

[0105] When the resistivity model error is calculated using the above formula (11), and are the inverted resistivity value and the true resistivity value of the ith underground unit respectively; when using the above formula (11) to calculate the model error of the polarizability, and are the inverted polarizability value and the true polarizability value of the ith underground unit, respectively.

[0106] The calculation formula of the time-shift model difference is as follows (12), which is the sum of the absolute values ​​of the model differences at adjacent moments:

[0107]

[0108] In this example, the cross gradient mean of the inversion result of the traditional method is 6.29×10 -5 After the inversion of the present invention, the cross gradient mean is 6.70×10 -6 , the cross gradient mean is significantly reduced, indicating that the resistivity and polarizability inversion results of the present invention are more similar in structure. After the traditional method inversion, the model similarity of the resistivity result is 5.83, and the model similarity of the polarizability result is 0.55; after the present invention inversion, the model similarity of the resistivity result is 5.77, and the model similarity of the polarizability result is 0.36, indicating that the present invention improves the accuracy of the inversion result. After the traditional method inversion, the model difference of the resistivity result is 2.4×10 5 The model difference of the polarizability result is 79.7; after the inversion of the present invention, the model difference of the resistivity result is 1.14×10 5 , the model difference of the polarizability result is 24.87, indicating that the present invention improves the continuity of the inversion results at different times over time.

[0109] observe Figure 4 , Figure 5 It can be found that the resistivity and polarizability results of the present invention are more consistent in structure and have a higher degree of fit with the real model, and there are fewer false anomalies in the background of the resistivity results, which confirms the effectiveness and superiority of the four-dimensional joint inversion method of the time-shift resistivity method and the time-shift induced polarization method developed by the present invention.

[0110] The present invention has a reasonable conception and combines the advantages of time-shift inversion and joint inversion, which can not only ensure the continuity of inversion results at different times over time, but also improve the structural similarity of resistivity and polarizability models, and has higher resolution and extremely strong anti-interference ability.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method, characterized in that: The following steps are involved: (1) Preprocess the observation data; (2) Read the files required for inversion; (3) Time-shift inversion of apparent resistivity data and apparent polarizability data is performed by resistivity method and induced polarization method respectively, and the time-shift resistivity model and time-shift polarizability model with abnormal morphology are obtained; (4) The cross-gradient term between the time-shift resistivity model and the time-shift polarizability model with abnormal morphology is calculated through the cross-gradient function, and then the cross-gradient term is substituted into the objective function of the four-dimensional joint inversion of the time-shift resistivity method and the time-shift induced polarization method, and a weight factor is set for the cross-gradient term. The initial model is the model after the most recent iteration. Through alternating iterations, the joint inversion results of the time-shift resistivity method and the time-shift induced polarization method with cross-gradient constraints are calculated; (5) Repeat the above step (4), calculate the cross-gradient term and perform inversion iteration alternately, and update the time-shift resistivity model and the time-shift polarizability model until the data fitting error of the inversion results of the time-shift resistivity method and the time-shift induced polarization method is less than the expected value or the number of iterations reaches the set value, and the iteration is terminated; (6) Calculate the cross-gradient mean of the time-shift resistivity and time-shift polarizability inversion results, the model error between the model and the true model, and the model difference at adjacent times to evaluate the inversion results.

2. The four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method according to claim 1 is characterized in that: The observation data includes information on the coordinates of the transmitting electrode and the receiving electrode corresponding to all the measuring points, the apparent resistivity, the apparent polarizability and the current; The preprocessing process includes data conversion, distortion point elimination and data format adjustment; the data conversion is to obtain the original data for inversion, namely apparent resistivity and apparent polarizability, through the observed electric field calculation; the distortion point elimination is to eliminate the distortion points of the observed data, and then use the interpolation method to fill the missing data points after elimination; the data format adjustment includes adjusting the data to a fixed format.

3. The four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method according to claim 1 is characterized in that: The files required for the inversion in step (2) include preprocessed observation data, parameter data and initial model; the parameter data are mainly weight factors of model terms and time-shift terms; the initial model includes an initial time-shift resistivity model and an initial time-shift polarizability model.

4. The four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method according to claim 1 is characterized in that: The specific process of step (3) is as follows: First, the apparent resistivity data is inverted by the time-lapse resistivity method. After multiple iterations, the time-lapse resistivity model M with an abnormal shape is obtained. ρ It is expressed as: M ρ =[m ρ,1 m ρ,2 ... m ρ,nt ] T (1); In the above formula (1), m ρ,nt is the resistivity model of the underground at a certain moment, nt represents the number of moments; Then, based on the resistivity time-shift inversion results, the apparent polarizability data is inverted by time-shift induced polarization method. After multiple iterations, the time-shift polarizability model M with an abnormal shape is obtained. η It is expressed as: M η =[m η,1 m η,2 ... m η,nt ] T (2); In the above formula (2), m η,nt is the polarizability model of the underground at a certain moment, and nt represents the number of moments.

5. The four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method according to claim 4 is characterized in that: The specific process of step (4) is as follows: First, according to the gradient of the time-lapse resistivity model and the gradient of the time-shifted polarizability model The cross gradient function t is obtained by multiplication: The cross gradient term Φ is calculated according to the cross gradient function t cg : The cross-gradient term is mainly used to measure the structural similarity between the time-shift resistivity model and the time-shift polarizability model. The larger the cross-gradient term, the greater the structural difference, and the smaller the cross-gradient term, the greater the structural difference. After the cross-gradient term is calculated, the calculated cross-gradient term is introduced into the objective function of the joint inversion of the time-shift resistivity method and the time-shift induced polarization method, which is and When the objective function is minimized, the cross-gradient term will play a constraint role and promote the structural similarity of the time-shift resistivity model and the time-shift polarizability model. The joint inversion of the time-shift resistivity method and the time-shift induced polarization method adopts the strategy of constructing two objective functions. The objective function of the joint inversion of the time-shift resistivity method and the time-shift induced polarization method is shown in the following formula (5): In the above formula (5), Φ(M ρ ) is the objective function of the resistivity method in the joint inversion of the time-shift resistivity method and the time-shift induced polarization method; Φ(M η ) is the objective function of the induced polarization method in the joint inversion of the time-shift resistivity method and the time-shift induced polarization method; Φ d (M ρ ) and Φ d (M η ) are the data items of the resistivity method and the induced polarization method, respectively, which are used to ensure that the difference between the forward response of the inversion result and the original data is minimized; Φ m (M ρ ) and Φ m (M η ) are the model terms of the resistivity method and the induced polarization method, respectively, which are used to ensure the smoothness of the inversion results and reduce the multiple solutions of the inversion; Φ t (M ρ ) and Φ t (M η ) are the time-shift terms of the resistivity method and the induced polarization method, which are used to ensure the temporal continuity of the inversion results at different times; Φ cg (M ρ ,M η ) and Φ cg (M η ,M ρ ) are the cross gradient terms of the resistivity method and the induced polarization method, which are used to ensure that the resistivity results and the polarizability results are more similar in morphology and structure; and are the weight factors of the model term, time shift term and cross gradient term in the objective function of the resistivity method; and are the weight factors of the model term, time shift term and cross gradient term in the objective function of the induced polarization method. By adjusting the size of the weight factor, the strength of the constraint effect of each term in the inversion can be controlled to obtain the optimal inversion result; the initial weight factor of the cross gradient term is 0, and it is set to a non-zero value after a certain number of iterations; The data item is obtained by the following formula (6): In the above formula (6), D ρ,obs is the apparent resistivity data observed at all times; D η,obs is the apparent polarizability data observed at all times; F(M ρ ) is the forward response function of the resistivity method; F(M η ) is the forward response function of the induced polarization method, and the forward response of the model can be calculated through the forward function; and The data covariance matrices of the resistivity method and induced polarization are respectively, and the diagonal elements are D ρ,obs and D η,obs A diagonal matrix with 5% of the elements and the remaining elements being 0; The model term is obtained by the following formula (7): In the above formula (7), C M,ρ and C M,η is the model covariance matrix; M ρ,0 and M η,0 They are the initial time-shift resistivity model and the initial time-shift polarizability model respectively; The time shift term is obtained by the following formula (8): In the above formula (8), m ρ,k and m η,k They are the resistivity and polarizability models of the underground at time k respectively.

6. The four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method according to claim 1, characterized in that: In each inversion iteration process of the model update in step (5), it is determined whether the data fitting error is less than the expected value or the number of iterations reaches the set value. If so, the inversion iteration is terminated; if not, the inversion iteration is continued; If one of the resistivity method and the induced polarization method reaches the inversion iteration termination condition, then this method terminates the inversion iteration, and the other method continues to iterate until the iteration termination condition is met; The termination condition of the inversion iteration is that the RMS of the data fitting error calculated at all times is less than the expected value or the number of iterations reaches the set maximum number of iterations: In the above formula (9), RMS ρ and RMS η They are the data fitting differences of the resistivity method and the induced polarization method.

7. The four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method according to claim 1 is characterized in that: The cross gradient mean in step (6) is represented by t, which is obtained by the following formula (10): In the above formula (10), N is the total number of underground grids.

8. The four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method according to claim 1, characterized in that: The model error in step (6) is expressed as δ, which is the difference between the inversion result and the true model. The model errors of resistivity and polarizability are obtained using the following formula (11): When the resistivity model error is calculated using the above formula (11), and are the inverted resistivity value and the true resistivity value of the ith underground unit respectively; when using the above formula (11) to calculate the model error of the polarizability, and are the inverted polarizability value and the true polarizability value of the ith underground unit, respectively.

9. The four-dimensional joint inversion method of time-shift resistivity method and time-shift induced polarization method according to claim 1, characterized in that: The model difference between adjacent moments in step (6) is expressed as It is the sum of the absolute values ​​of the model differences at adjacent moments and is obtained by the following formula (12): In the above formula (12), m k is the model at time k, m k+1 is the model for time k+1, and nt represents the number of time moments.

Citation Information

Patent Citations

  • Tunnel resistivity polarizability joint inversion gradient optimization method based on deep learning

    CN115577612A

  • Time shift resistivity method normalized data four-dimensional inversion method

    CN118033750A

  • Adaptive time-lapse sub-surface electrical resistivity monitoring

    US20150006081A1

Cited By

  • Time-shifting resistivity inversion imaging method based on time sequence gradient structure constraint

    CN119414471A

  • Time-lapse resistivity inversion imaging method based on time series gradient structure constraint

    CN119414471B

  • 4D direct current method inversion imaging method based on time sequence gradient structure constraint

    CN119758465A

  • Time shifting inversion method based on time shifting item constraint

    CN119960066A

  • A time-lapse inversion method based on time-lapse term constraints

    CN119960066B