Geophysical data fusion method based on logic function

Abstract

The invention relates to a data fusion method for geophysical data based on a logic function. According to the method, typical features of single geophysical data can be highlighted on the basis of a logic function, deeper, richer and finer information is obtained, the function of the data is played to the maximum extent, and the utilization rate of the data is increased so that the data cannot be wasted; and secondly, through fusion of various geophysical data, information which is richer and deeper than unit data is mined, the anomaly delineation range is narrowed, the exploration difficulty is reduced, and the exploration precision is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of mathematical three-dimensional models in geophysical exploration, and relates to a geophysical data fusion method based on logical functions. Background Art

[0002] From the current development of geophysical exploration, the information volume of single data is limited. The seismic exploration method has a high resolution for the stratigraphic structure, but has a low resolution for some special structures; the gravity and magnetic exploration method has good lateral resolution ability and plays an important role in dividing the distribution of geological structures, but has poor vertical resolution ability and shallow exploration depth; the electromagnetic exploration method has special advantages in studying the deep structure of the earth, but has poor lateral resolution.

[0003] Only by comprehensively using various geophysical data from different sources can the limitations of a single method be overcome and the non-uniqueness of inversion be reduced. Traditional geophysical data fusion is to comprehensively interpret the original data collected by different geophysical methods after processing, based on the representative characteristics of different geophysical methods according to certain rules. Although the current geophysical data fusion method overcomes the influence of non-uniqueness to a certain extent and improves the credibility of comprehensive interpretation, there are still some problems, such as the limited use of data. Mainly, the processing and interpretation of single geophysical data are easily restricted by the complex properties of geological bodies, and characteristics are easily lost in the fusion interpretation process, and more abundant information cannot be displayed, thus affecting the processing and interpretation effect. Summary of the Invention

[0004] In view of the problems existing in the prior art, the present invention proposes a geophysical data fusion method based on logical functions, including the following steps:

[0005] Step 1: Preparation of geophysical data based on the work area

[0006] Comprehensively collect the geological background information of the work area, and based on the specific situation of the work area, clarify the specific exploration objectives and prepare the existing measured geophysical data of the work area;

[0007] Step 2: Conduct three-dimensional individual inversion based on the measured geophysical data;

[0008] Step 3: Convert the inversion results of geophysical data based on logical functions to obtain three-dimensional logical anomaly data;

[0009] Step 4: Fusion of the individual inversion results of three-dimensional geophysical data.

[0010] On the basis of the above solution, the inversion iteration formula in Step 2 is:

[0011]

[0012] where: T is the matrix transpose, G T 、 is the transpose matrix of G, W d 、W m .

[0013] Based on the above scheme, during inversion, the conjugate gradient method is used to solve for the updated amount m of the model, and then the model is updated separately for the next inversion iteration until the optimal solution that meets the preset data fitting error accuracy is obtained.

[0014] Based on the above scheme, the specific method for step 3 is: converting different geophysical inversion data using the following logical function;

[0015] Logical function:

[0016] where a represents a natural number in [0, 10]; k represents the extreme value of the logical function;

[0017] Taking m obtained from step 2 as the independent variable x and inputting it into the logical function for conversion to obtain the corresponding three-dimensional logical anomaly data.

[0018] Based on the above scheme, the value-taking rule of the above a is as follows: if the physical property classification statistically based on known petrophysical and logging data is 1 type, then take a = 1; if it is 2 types, then take a = 2; if it is 3 types, then take a = 3; and so on. If the physical property classification is 10 types, then take a = 10. Usually, a = 10 can already meet the underground physical property classification situation.

[0019] Based on the above scheme, in step 1, the geological background information includes stratigraphic distribution, geological structure characteristics, rock types and their physical properties.

[0020] Based on the above scheme, in step 1, the measured geophysical data are gravity data and magnetic data; after conversion in step 3, the obtained data are three-dimensional gravity (density) logical anomalies and three-dimensional magnetic (magnetic susceptibility) logical anomaly data; during data fusion in step 4, it is carried out by calculating the product of the three-dimensional gravity (density) logical anomalies and the three-dimensional magnetic (magnetic susceptibility) logical anomalies.

[0021] Based on the above scheme, in step 1, the measured geophysical data also include seismic data and / or electrical method data. Correspondingly, after conversion in step 3, the obtained data also include three-dimensional seismic (velocity) logical anomalies and / or three-dimensional electrical method (resistivity) logical anomaly data. During data fusion in step 4, it is carried out by calculating the product of the three-dimensional logical anomalies obtained from the selected measured geophysical data.

[0022] Advantages of the present invention:

[0023] The method of the present invention can highlight the typical features of a single geophysical data based on a logical function, obtain deeper, richer, and more refined information, maximize the role of the data itself, improve the utilization rate of the data so that it will not be wasted; secondly, through the fusion of multiple geophysical data, more abundant and profound information than the unit data can be mined, the scope of anomaly delineation can be narrowed, the exploration difficulty can be reduced, and the exploration accuracy can be improved. Brief Description of the Drawings

[0024] Figure 1 It is the logical function diagram with different a values in Embodiment 1 of the present invention;

[0025] Figure 2 It is the gravity and magnetic separate inversion slice diagram obtained by using the method of the present invention in Embodiment 2;

[0026] Figure 3 It is the logical function diagram corresponding to a = 8 in Embodiment 2;

[0027] Figure 4 It is the density logic anomaly slice diagram converted based on the method of the present invention in Embodiment 2;

[0028] Figure 5 It is the magnetic susceptibility logic slice diagram converted based on the method of the present invention in Embodiment 2;

[0029] Figure 6 It is the gravity and magnetic data fusion slice diagram converted based on the method of the present invention in Embodiment 2. Detailed Embodiments

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. Usually, the components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0031] Embodiment 1

[0032] The present invention provides a new method for fusing and interpreting geophysical data. Specifically, the method of the present invention includes the following steps:

[0033] Step 1: Preparation of geophysical data based on the work area

[0034] Comprehensively collect the geological background information of the work area, including stratigraphic distribution, geological structure characteristics (such as faults, folds, etc.), rock types and their physical properties (such as velocity, density, magnetism, conductivity, etc.). The background information is used to determine which geophysical data in the work area are used for fusion interpretation by the method of the present invention.

[0035] Based on the specific situation of the work area, prepare the measured geophysical data such as seismic data, gravity data, magnetic data, and electrical method data existing in this work area.

[0036] Step 2: Conduct 3D separate inversion based on the measured geophysical data

[0037] The 3D inversion is carried out based on the Haotuo geophysical software V2024.

[0038] The separate inversion of geophysical data adopts the regularized smooth constraint inversion, and its objective function includes a data fitting term and a model smoothing constraint term.

[0039]

[0040] In the formula: m is the model vector (in the present invention, m represents velocity, density, magnetic susceptibility, or resistivity); d is the observed data vector; φ d (m) is the data fitting objective function; φ m (m) is the model objective function; G is the Jacobian matrix; W d is the data diagonal weighting matrix; W m is the model weighting matrix; m 0 is the reference model vector; β represents the regularization factor.

[0041] The inversion calculation is the process of solving the minimum value of the objective function. Therefore, taking the partial derivative of the objective function and setting it equal to 0, we can obtain:

[0042]

[0043] Based on the solution of the above formula, the inversion iteration formula can be obtained as:

[0044]

[0045] In the formula, T is the matrix transpose, G T 、 are the transpose matrices of G, W d 、W m respectively.

[0046] Then, use the conjugate gradient method to solve for the updated amount m of the model (m represents velocity, density, magnetic susceptibility, or resistivity), and then update the model separately for the next inversion iteration until the optimal solution that meets the preset data fitting error accuracy is obtained.

[0047] Step 3: Convert the inversion results of geophysical data based on the logical function

[0048] In the method of the present invention, the following logical function is used to convert different geophysical inversion data. The logical function can retain more useful information in a single geophysical data, improve the utilization rate of the data, and can also highlight the typical characteristics of a single geophysical data by constructing different logical functions, obtaining more profound, richer and more refined information than the separate processing and interpretation of single geophysical data, realizing the advantages and disadvantages complementarity and mutual complement of multiple geophysical data.

[0049] Construct a reasonable logical function:

[0050] In the formula, a represents a natural number in the range of [0, 10], reflecting the simplicity of the logical function curve shape, which is mainly related to the number of physical property classifications statistically obtained from known rock physics and logging data. In the ideal case where the geophysical data is completely homologous, if the physical property classification statistically obtained from known rock physics and logging data is 1 type, then a = 1; if it is 2 types, then a = 2; if it is 3 types, then a = 3; and so on. If the physical property classification is 10 types, then a = 10. Usually, a = 10 can already meet the underground physical property classification situation. k represents the extreme value of the logical function, reflecting that the maximum value after the conversion of various geophysical data through the logical function is 2.

[0051] Take the m (m represents velocity, density, magnetic susceptibility or resistivity) obtained from step 2 as the independent variable x and input it into the logical function for conversion to obtain the corresponding y, that is, the 3D seismic (velocity) logical anomaly, 3D gravity (density) logical anomaly, 3D magnetic (magnetic susceptibility) logical anomaly, and 3D electrical method (resistivity) logical anomaly.

[0052] Step 4: Fusion and analysis of the separate inversion results of 3D geophysical data

[0053] Perform data fusion on the y obtained from step 3 by calculating the product of the 3D seismic (velocity) logical anomaly, 3D gravity (density) logical anomaly, 3D magnetic (magnetic susceptibility) logical anomaly, and 3D electrical method (resistivity) logical anomaly.

[0054] The core idea of the method of the present invention is that after the original seismic, gravity, magnetic, electrical method and other geophysical data are subjected to inversion processing, the obtained inversion results are transformed based on a logical function. On the basis of retaining the original seismic, gravity, magnetic, electrical method geophysical anomaly characteristics, more useful inversion information is further highlighted, and the parts related to the inversion results in the geological body are combined, thereby increasing the reliability of the interpretation; in addition, in the process of inversion and fusion interpretation of geophysical data, the geophysical data inversion anomalies transformed by the logical function can be directly subjected to data fusion, which can simplify the process of comprehensive interpretation of geophysical data, greatly reduce the time, and improve the interpretation efficiency.

[0055] Example 2

[0056] Based on the method in Example 1, this application performs data fusion using the actual geophysical data of a certain work area. The specific steps are as follows:

[0057] Step 1: Preparation of geophysical data based on the work area

[0058] Comprehensively collect the geophysical data of this work area, and determine that the existing geophysical data in this work area are measured gravity data and measured magnetic data, as shown in Table 1 and Table 2. Among them, the gravity data format is (x, y, g), and the magnetic data format is (x, y, m). Where x and y are coordinate values, g is the measured gravity data value, and T is the measured magnetic data value.

[0059] Table 1 Measured gravity data

[0060] x y g … … … 0 0 -0.439 500 0 -1.122 1000 0 -1.658 1500 0 -2.026 2000 0 -2.392 2500 0 -2.920 3000 0 -3.490 3500 0 -3.931 4000 0 -4.190 4500 0 -4.193 5000 0 -3.940 5500 0 -3.477 6000 0 -2.719 6500 0 -1.656 7000 0 -0.488 7500 0 0.462 8000 0 1.139 8500 0 1.835 9000 0 2.633 … … …

[0061] Table 2 Measured magnetic data

[0062] x y T … … … 0 0 -107.146 500 0 -113.702 1000 0 -116.807 1500 0 -126.000 2000 0 -145.211 2500 0 -147.437 3000 0 -151.621 3500 0 -162.492 4000 0 -192.072 4500 0 -194.270 5000 0 -165.849 5500 0 -187.941 6000 0 -227.937 6500 0 -318.211 7000 0 -264.799 7500 0 -213.338 8000 0 -57.059 8500 0 -26.706 9000 0 -176.666 … … …

[0063] Step 2: Perform 3D separate inversion based on the measured geophysical data

[0064] In this example, the 3D inversion is carried out based on the Haotuo geophysical software V2024.

[0065] The separate inversion of the measured gravity data and measured magnetic data in this work area uses regularized smooth constraint inversion, and its objective function includes a data fitting term and a model smoothing constraint term.

[0066]

[0067] In the formula: m is the model vector (m represents density, magnetic susceptibility); d is the observed data vector; φ d (m) is the data fitting objective function; φ m (m) is the model objective function; G is the Jacobian matrix; W d is the data diagonal weighting matrix; W m is the model weighting matrix; m 0 is the reference model vector; β represents the regularization factor. The inversion calculation is the process of solving the minimum value of the objective function. The inversion calculation is the process of solving the minimum value of the objective function. Therefore, taking the partial derivative of the objective function and setting it equal to 0, we can get:

[0068]

[0069] Based on the solution of the above formula, the inversion iteration formula can be obtained as:

[0070]

[0071] Then, the conjugate gradient method is used to solve the update amount m of the model (m represents density and magnetic susceptibility), and then the model is updated separately for the next inversion iteration until the optimal solution that meets the preset error accuracy is obtained. (In this embodiment, the preset error accuracy is 1.00E-06)

[0072] After separately performing 3D inversion on the measured gravity data and measured magnetic data of this work area, the results shown in Figure 2 , Table 3 and Table 4 are obtained.

[0073] Figure 2 shows the density and magnetic susceptibility slice maps at a depth of 5000m of the results of separately performing 3D gravity and magnetic inversion based on the existing measured gravity data and measured magnetic data of this work area. It can be seen from the inversion results that there is more useful information shown in the density slice. In the southernmost part, there is a density distribution with a stepped shape horizontally. In the eastern and southeastern parts, there are spike-shaped density distributions. In the northeastern part, there is a density distribution with alternating high and low values. In the western part, there is an irregular-shaped density distribution; there is less information shown in the magnetic susceptibility slice. In the northeastern part, there is a magnetic susceptibility distribution with alternating high and low values. In the western and southern parts, there are scattered magnetic susceptibility distributions.

[0074] Table 3 Data of the density results of gravity-only inversion

[0075] x y z Density … … … … 48473 0 1500 2.486 49338 0 1500 2.514 37220 850 1500 2.514 46741 1700 1500 2.486 51069 1700 1500 2.486 46741 2550 1500 2.514 55397 2550 1500 2.486 46741 4250 1500 2.486 48473 4250 1500 2.486 51069 4250 1500 2.514 34623 5100 1500 2.514 50204 5100 1500 2.514 34623 5950 1500 2.486 52801 5950 1500 2.486 51935 0 2100 2.514 29429 850 2100 2.514 43279 850 2100 2.514 47607 850 2100 2.514 … … … …

[0076] Table 4 Data of the magnetic susceptibility results of magnetic-only inversion

[0077] x y z Magnetic susceptibility … … … … 31938 425 1500 405.649 32812 425 1500 406.580 33688 425 1500 406.168 34562 425 1500 407.914 35438 425 1500 408.326 36312 425 1500 409.141 37188 425 1500 407.807 38062 425 1500 410.567 38938 425 1500 404.289 39812 425 1500 403.062 40688 425 1500 405.114 41562 425 1500 409.167 42438 425 1500 408.474 43312 425 1500 407.354 44188 425 1500 408.540 45062 425 1500 406.662 45938 425 1500 406.687 46812 425 1500 402.782 … … … …

[0078] Step 3: Convert the inversion results of geophysical data based on logical functions

[0079] The results of inverting the measured gravity and magnetic data of this work area are converted based on different logical functions to obtain corresponding logical anomaly maps.

[0080] Since there are only 2 types of existing measured geophysical data in this work area, namely gravity and magnetism, the physical property corresponding to the gravity data is density, and the physical property corresponding to the magnetic data is magnetic susceptibility. Through the physical property measurement of the geological outcrop rocks and the statistical analysis of well logging data in the work area, the physical property classification of density and magnetic susceptibility in the work area is 8 types, as shown in Table 5. Therefore, when constructing the logical function, the value of a is 8.

[0081] Table 5 Statistical table of physical properties of density and magnetic susceptibility in the work area

[0082]

[0083]

[0084] Figure 1 and 3 In it, the abscissa and ordinate represent the domain x read by the logic function, where x ∈ [-24, 24], and the value range y read by the logic function, where y ∈ [0, 2).

[0085] Construct a reasonable logic function:

[0086] Take m (representing density and magnetic susceptibility) obtained from step 2 as the independent variable x and input it into the logic function for conversion to obtain the corresponding y, that is, the three-dimensional gravity (density) logic anomaly and the three-dimensional magnetic (magnetic susceptibility) logic anomaly. The results are shown in Tables 6 and 7 Figure 4 and Figure 5 as shown

[0087] Table 6 Density logic anomaly data obtained by converting the logic anomaly function constructed with a = 8

[0088] x y z Density logic anomaly … … … … 48473 0 1500 0.862 49338 0 1500 1.138 37220 850 1500 1.138 46741 1700 1500 0.862 51069 1700 1500 0.862 46741 2550 1500 1.138 55397 2550 1500 0.862 46741 4250 1500 0.862 48473 4250 1500 0.862 51069 4250 1500 1.138 34623 5100 1500 1.138 50204 5100 1500 1.138 34623 5950 1500 0.862 52801 5950 1500 0.862 51935 0 2100 1.138 29429 850 2100 1.138 43279 850 2100 1.138 47607 850 2100 1.138 … … … …

[0089] Table 7 Magnetic susceptibility logic anomaly data obtained by converting the logic anomaly function constructed with a = 8

[0090] x y z Magnetic susceptibility logic anomaly … … … … 31938 425 1500 0.614 32812 425 1500 0.617 33688 425 1500 0.616 34562 425 1500 0.622 35438 425 1500 0.623 36312 425 1500 0.626 37188 425 1500 0.622 38062 425 1500 0.631 38938 425 1500 0.609 39812 425 1500 0.605 40688 425 1500 0.612 41562 425 1500 0.626 42438 425 1500 0.624 43312 425 1500 0.620 44188 425 1500 0.624 45062 425 1500 0.618 45938 425 1500 0.618 46812 425 1500 0.604 … … … …

[0091] Figure 4 The three-dimensional gravity single inversion results in this work area are shown. Based on this method, the density logic anomaly slice maps at a depth of 5000 m are obtained by selecting different logic functions. It can be seen from the figure that the converted gravity inversion results show more abundant information while retaining many original gravity inversion characteristics.

[0092] Figure 5 The three-dimensional magnetic single inversion results in this work area are shown. Based on this method, the magnetic susceptibility logic anomaly slice maps at a depth of 5000 m are obtained by selecting different logic functions. It can be seen from the figure that the converted magnetic inversion results also show more abundant useful information.

[0093] Step 4: Fusion and analysis of three-dimensional geophysical data single inversion results

[0094] Fuse the y obtained from step 3 by calculating the product of the three-dimensional gravity (density) logic anomaly and the three-dimensional magnetic (magnetic susceptibility) logic anomaly.

[0095] Multiply the inversion logic anomaly results based on the logic function conversion to obtain the gravity and magnetic logic anomaly fusion maps and data based on different logic function conversions. Table 8 shows the data after fusing the gravity and magnetic results obtained by converting the logic anomaly function constructed with a = 8.

[0096] Data after fusing gravity and magnetic results using the logical anomaly function constructed with a = 8 in Table 8

[0097] x y z Fusion result … … … … 31938 1275 1500 0.632 32812 1275 1500 0.635 33688 1275 1500 0.633 34562 1275 1500 0.630 35438 1275 1500 0.631 36312 1275 1500 0.633 37188 1275 1500 0.637 38062 1275 1500 0.628 38938 1275 1500 0.570 39812 1275 1500 0.634 40688 1275 1500 0.636 41562 1275 1500 0.632 42438 1275 1500 0.646 43312 1275 1500 0.645 44188 1275 1500 0.633 45062 1275 1500 0.586 45938 1275 1500 0.625 46812 1275 1500 0.622 … … … …

[0098] Figure 6 It shows the gravity and magnetic logical anomaly fusion slice map at a depth of 5000m obtained by data fusion based on logical function conversion in this work area (the map is generated using Voxler 4.0 software).

[0099] It can be seen from the fusion results that it combines the useful information for lithology identification and structure division in the density slice and magnetic susceptibility slice, so as to obtain more accurate interpretation results and provide more precise auxiliary information for geological interpretation.

[0100] The method proposed by the present invention can not only retain the useful information in the 3D gravity and magnetic individual inversion results, improve the data utilization rate, but also perform geophysical data conversion by constructing a reasonable logical function to highlight the typical characteristics of the 3D gravity and magnetic individual inversion results, and obtain deeper, richer and more refined information than the separate processing and interpretation of 3D gravity and magnetic individual inversions, realizing the complementary advantages of gravity and magnetic data. Secondly, the obtained data can be fully and reasonably utilized to maximize the role of the data itself and prevent it from being wasted; secondly, through logical function conversion and data fusion, the typical characteristics of the data can be highlighted, and more abundant and profound information than the unit data can be mined, narrowing the anomaly delineation range, reducing the exploration difficulty, and improving the exploration accuracy.

[0101] It should be noted that the features in the embodiments of this application can be combined with each other without conflict.

[0102] The above are only the preferred embodiments of this application and are not used to limit this application. For those skilled in the art, this application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this application shall be included in the protection scope of this application.

Claims

1. A geophysical data fusion method based on logical function, characterized in that: The steps include: Step 1: Geophysical data preparation based on the work area Comprehensively collect geological background information of the work area, and based on the specific conditions of the work area, clarify specific exploration targets and prepare the existing measured geophysical data of the work area; Step 2: Perform three-dimensional separate inversion based on measured geophysical data; Step 3: Convert the geophysical data inversion results based on the logic function to obtain three-dimensional logical anomaly data; Step 4: Fusion of individual inversion results of 3D geophysical data.

2. The geophysical data fusion method based on logical function according to claim 1 is characterized in that: The inversion iteration formula in step 2 is: Where: T is the matrix transpose, G T , G, W d , W m The transposed matrix of .

3. The geophysical data fusion method based on logical function according to claim 2 is characterized in that: During inversion, the conjugate gradient method is used to solve the update amount m of the model, and then the model is updated separately for the next inversion iteration until the optimal solution that meets the preset data fitting error accuracy is obtained.

4. The geophysical data fusion method based on logical function according to claim 3 is characterized in that: The specific method of step 3 is: using the following logic function to transform different geophysical inversion data; Logical functions: In the formula, a represents a natural number in [0,10]; k represents the extreme value of the logic function; The m obtained in step 2 is used as the independent variable x to input the logic function for transformation to obtain the corresponding three-dimensional logic abnormality data.

5. The geophysical data fusion method based on logical function according to claim 4 is characterized in that: In step 3, the value of a is as follows: if the physical property classification based on the known rock physics and logging data is 1, then a=1; if the physical property classification is 2, then a=2; if the physical property classification is 3, then a=3; and so on, if the physical property classification is 10, then a=10.

6. The geophysical data fusion method based on logical function according to claim 1, characterized in that: In step 1, the geological background data include stratum distribution, geological structure characteristics, rock types and their physical properties.

7. The geophysical data fusion method based on logical function according to claim 1 is characterized in that: In step 1, the measured geophysical data are gravity data and magnetic data; after conversion in step 3, three-dimensional gravity (density) logical anomaly and three-dimensional magnetic (magnetic susceptibility) logical anomaly data are obtained; when fusion is performed in step 4, data fusion is performed by calculating the product of three-dimensional gravity (density) logical anomaly and three-dimensional magnetic (magnetic susceptibility) logical anomaly.

8. The geophysical data fusion method based on logical function according to claim 7 is characterized in that: In step 1, the measured geophysical data also includes seismic data and / or electrical data, and the corresponding data obtained after conversion in step 3 also includes three-dimensional seismic (velocity) logical anomaly and / or three-dimensional electrical (resistivity) logical anomaly data. In step 4, data fusion is performed by taking the product of the three-dimensional logical anomaly obtained from the selected measured geophysical data.

Images