Deep ground space abnormal structure identification method based on multi-source heterogeneous data fusion

Abstract

The invention discloses a deep ground space abnormal structure identification method based on multi-source heterogeneous data fusion, and the method comprises the steps: obtaining the detection data of the same deep ground space through geophysical prospecting and drilling, and obtaining the corresponding interpretation results; then, geological structure anomaly evaluation indexes are constructed, and the evaluation indexes are assigned according to different data interpretation results; then, on the basis of the D-S evidence theory, a basic probability matrix M of multi-source heterogeneous comprehensive investigation data response is constructed, a judgment matrix Y is constructed after anomaly recognition fusion indexes and normalization processing are carried out, and then a preliminary judgment result of the deep ground space anomaly geological structure is obtained according to the judgment matrix Y; and finally, entropy weight fusion is adopted to process the preliminary identification result in the step S3 so as to enlarge the confidence degree of abnormal point data. Therefore, a multi-source heterogeneous data DS-entropy weight identification vector is obtained, geological conditions of different positions of a deep ground space needing to be detected are obtained according to the vector, and finally accurate identification of the range and the position of geological structure anomaly is achieved.

Description

Technical Field

[0001] The present invention belongs to the field of deep earth space geological exploration, and specifically is a method for identifying deep earth space abnormal structures based on DS-entropy weight fusion of multi-source heterogeneous data. Background Art

[0002] Deep earth space involves hidden disaster-causing factors such as abnormal geological structures such as faults, sinkholes, and goafs. Common exploration methods include seismic exploration, high-density electrical exploration, and borehole transient electromagnetic method.

[0003] Seismic exploration technology analyzes the propagation characteristics of seismic waves in underground media to identify faults and the structural undulations of coal seam roofs and floors. Three-dimensional seismic exploration can identify faults deeper than five meters. High-density electrical exploration exploits electrical differences in underground rocks to detect geological structures and is suitable for detecting abnormal areas such as sinkholes and water-rich areas. Borehole transient electromagnetic (TEM) technology, by placing transmitters and receivers within the borehole, enables precise detection of water bodies around coal mine boreholes. Its pseudo-seismic full-waveform inversion technology improves the resolution of geological anomaly boundaries.

[0004] While these methods have played an important role in deep-earth geological structure exploration, they still have limitations. For geophysical exploration, seismic exploration has limited resolution for small-scale structural detection; high-density electrical exploration is susceptible to topography and interference; and for drilling, borehole transient electromagnetic methods present significant difficulties in interpreting complex geological conditions. Furthermore, different geophysical and drilling measurement methods can produce different interpretations of the same geological structure. Even the same technical approach can yield different results, making it impossible to accurately determine the extent and location of geological structures. Combining multiple detection methods, however, complicates processing of multi-source, heterogeneous geophysical data, requiring extensive computation and analysis, and presents the problem of non-uniqueness in data inversion. Multi-source, heterogeneous drilling data can only be detected using point and line methods, failing to fully reflect the geological conditions of the entire region and exhibiting a one-sided perspective. Addressing this contradiction is the primary challenge of data fusion, and improving data confidence is a pressing issue.

[0005] Therefore, in order to address the problem of inaccurate detection and unclear interpretation of deep-earth geological anomalies, how to provide a method that can separately obtain drilling and geophysical detection results as multi-source heterogeneous data, and output the final detection results after comprehensive processing of data fusion, weight allocation and increased confidence level of the data, so that the scope and position of the geological structure can be accurately located, is the research direction required by the present invention. Summary of the Invention

[0006] In response to the problems existing in the above-mentioned prior art, the present invention provides a method for identifying deep-earth spatial abnormal structures based on multi-source heterogeneous data fusion, which can respectively obtain drilling and geophysical detection results as multi-source heterogeneous data, and output the final detection results after comprehensive processing of data fusion, weight allocation and improvement of data confidence level, so that the scope and position of the geological structure can be accurately located.

[0007] To achieve the above objectives, the present invention adopts a technical solution: a method for identifying abnormal structures in deep space based on multi-source heterogeneous data fusion, comprising the following steps:

[0008] S1. Obtain multi-source heterogeneous comprehensive survey data:

[0009] The deep-earth geological structure signals are collected by geophysical prospecting and drilling methods, and the geophysical prospecting data interpretation results and drilling data interpretation results of deep-earth geological structure anomalies are obtained respectively; and the data from different sources are collected at the same sampling interval;

[0010] S2. Constructing evaluation indicators for multi-source heterogeneous comprehensive survey data:

[0011] Based on the required deep-earth space comprehensive exploration data, geological structural anomaly evaluation indicators are constructed as follows: Indicator 1: original geological exploration results; Indicator 2: high-density electrical method interpretation results; Indicator 3: seismic detection interpretation results; Indicator 4: borehole transient electromagnetic detection interpretation results. After the evaluation indicators are constructed, the data corresponding to each indicator is divided at the same scale and analyzed and processed, and then the indicators at different locations in the deep-earth space are assigned values.

[0012] S3. Fusion evaluation index based on improved DS evidence theory:

[0013] Based on the DS evidence theory, a basic probability matrix M of multi-source heterogeneous integrated exploration data responses is constructed. Anomaly identification fusion indicators are then established to process the individual data in matrix M. The processed data are then normalized to construct a judgment matrix Y for the multi-source heterogeneous integrated exploration data responses. Preliminary identification results of anomalous geological structures in deep space are then obtained based on the judgment matrix Y. DS evidence theory, as an extension of Bayesian reasoning, is a comprehensive theory that can handle ambiguity in data information. This theory uses a priori probability distribution functions to obtain posterior evidence intervals, thereby reducing the unknowns and uncertainties of fuzzy information. However, due to the poor accuracy of single-source test data, the sensitivity of the indicators for local anomaly diagnosis is enhanced by increasing the number of test sources. Due to the inconsistent dimensionality of the heterogeneous response data, the results of the multi-source heterogeneous integrated exploration data must be fused, dimensionlessized, and normalized.

[0014] S4. Entropy weight fusion is used to modify the evaluation index:

[0015] Entropy is a macroscopic quantity unique to information theory. Shannon proposed the concept of information entropy to characterize the complexity of system signals. The greater the degree of disorder of the signal, the greater the information entropy; the smaller the degree of disorder, the smaller the information entropy. Entropy weight fusion is an objective weight distribution method. This method obtains more objective indicator weights based on the degree of variation of the evaluation indicators. Entropy weight fusion is used to process the preliminary identification results of step S3 to expand the confidence level of the abnormal point data; the greater the entropy value of the data at a certain position in the evaluation result, the lower the credibility of the evaluation result; conversely, the smaller the entropy value, the greater the credibility of the evaluation result; thereby correcting the preliminary identification results;

[0016] S5. Identification and positioning of abnormal features of deep-earth space structures:

[0017] According to step S3 and step S4, the geological conditions at different locations in the deep earth space required for detection are obtained, thereby accurately obtaining the range and location of geological structural anomalies.

[0018] Furthermore, in step S1, data from different sources are collected at a sampling interval of 1 meter to ensure consistency in the collection of different data; the geophysical data interpretation results include high-density electrical interpretation results and seismic detection interpretation results, and the drilling data interpretation results are borehole transient electromagnetic detection interpretation results.

[0019] Further, the step S3 is specifically as follows:

[0020] Based on the DS evidence theory, the basic probability matrix M of the multi-source heterogeneous comprehensive survey data response is constructed:

[0021]

[0022] Where: They represent the original geological survey results, high-density electrical interpretation results, seismic detection interpretation results and borehole transient electromagnetic detection interpretation results at the sampling node n respectively; According to the DS evidence theory, when a certain The larger the value, the greater the possibility of anomaly at that location, and the greater the confidence level of that location;

[0023] Establish anomaly recognition fusion index, which is the basic probability function at sampling node n:

[0024]

[0025] Use the above function to process each data in the matrix M;

[0026] In order to eliminate the influence of different evaluation index dimensions, the matrix M is normalized to obtain the standardized matrix Y. The normalization formula is:

[0027] Thus, the judgment matrix Y of the multi-source heterogeneous comprehensive survey data response is constructed:

[0028]

[0029] According to the matrix Y, the fusion vector α=[y m1 y m2 ···y mn ], and the fusion vector α is used as the preliminary identification result of the abnormal geological structure in deep space.

[0030] Furthermore, in step S4, entropy weight fusion is used to process the preliminary identification result of step S3, specifically:

[0031] When the evaluation index contains multiple states, let the probability of occurrence of the i-th state be p i , i=1,2,…n, then the entropy of the indicator is expressed as:

[0032]

[0033] The frequency p of abnormal multi-source data in the i-th sampling node in the total information volume is i for:

[0034]

[0035] where p i The relationship satisfied is: x i The amount of self-information represented by I(x i ) is defined as: I(x i )=-log a P i ; where I(x i ) is a random variable, for all information I(x i ) to find the average value, that is, to calculate the mathematical expectation of all information self-information, which is the information entropy at the sampling node, denoted as P(x). The specific formula is as follows:

[0036]

[0037] Furthermore, in step S5, the geological conditions of different locations in the deep earth space required for detection are obtained according to steps S3 and S4, specifically:

[0038] The information entropy formula obtained in step S4 is used to process the data of the fusion vector α in step S3 to obtain the multi-source heterogeneous data DS-entropy weight identification vector as shown below:

[0039] P=[p 11 p 12···p ij ];

[0040] Based on the above vectors, the scope and location of geological structural anomalies can be accurately obtained.

[0041] Compared with the prior art, the present invention first uses geophysical exploration and drilling to respectively acquire detection data of the same deep earth space and obtains corresponding interpretation results respectively; then constructs geological structure anomaly evaluation indicators and assigns values to the evaluation indicators according to the interpretation results of different data; then, based on the DS evidence theory, constructs a basic probability matrix M of the multi-source heterogeneous comprehensive exploration data response, and after anomaly identification fusion indicators and normalization processing, constructs a judgment matrix Y, and then obtains preliminary identification results of abnormal geological structures in the deep earth space based on the judgment matrix Y; finally, entropy weight fusion is used to process the preliminary identification results of step S3 to increase the confidence level of the abnormal point data; thereby obtaining the multi-source heterogeneous data DS-entropy weight identification vector, based on which the geological conditions of different locations in the deep earth space to be detected are obtained, including possible hidden disaster-causing factors of abnormal geological structures such as faults, collapse columns, and goafs; compared with traditional single drilling, geophysical exploration or geological survey data, this method can effectively solve the shortcomings of single means of unclear exploration and narrow scope, thereby increasing the sensitivity of identifying geological structure anomalies in the deep earth space, thereby accurately obtaining the scope and location of geological structure anomalies. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic diagram of the overall process of the present invention;

[0043] Figure 2 It is a schematic diagram of obtaining different detection data for the same deep earth space in the present invention. DETAILED DESCRIPTION

[0044] The present invention will be further described below.

[0045] like Figure 1 As shown, the present invention includes the following steps:

[0046] S1. Obtain multi-source heterogeneous comprehensive survey data:

[0047] like Figure 2 As shown, deep-earth geological structure signals are collected by geophysical and drilling methods, and geophysical data interpretation results and drilling data interpretation results of deep-earth geological structure anomalies are obtained respectively; and data from different sources are collected at a sampling interval of 1 meter to ensure the consistency of different data collection; the geophysical data interpretation results include high-density electrical interpretation results and seismic detection interpretation results, and the drilling data interpretation results are borehole transient electromagnetic detection interpretation results.

[0048] S2. Constructing evaluation indicators for multi-source heterogeneous comprehensive survey data:

[0049] According to the deep space comprehensive exploration data to be detected, the geological structure anomaly evaluation index is constructed as follows: Index 1: original geological exploration results; Index 2: high-density electrical method interpretation results; Index 3: seismic detection interpretation results; Index 4: borehole transient electromagnetic detection interpretation results; after the evaluation index is constructed, the data corresponding to each indicator are divided at the same scale and analyzed and processed, and then the indicators at different positions in the deep space are assigned values. Specifically, the data corresponding to each indicator are first divided at the same scale to obtain the corresponding number of abnormal data points and normal data points, and the main peak value at the abnormal data point is obtained; the assignment is the quantification of the index, which is the ratio of the sum of the main peak values at all abnormal points to the sum of the values of all points. The specific formula is: Where: ε represents the number of outliers; j represents the number of data partitions; y ε Represents the main peak value at the abnormal point; use the above formula to assign values to the indicators at different positions in the deep space;

[0050] S3. Fusion evaluation index based on improved DS evidence theory:

[0051] Based on the DS evidence theory, a basic probability matrix M of the multi-source heterogeneous integrated exploration data response is constructed, and an anomaly identification fusion index is established to process each data in the matrix M. The processed data is then normalized to construct the judgment matrix Y of the multi-source heterogeneous integrated exploration data response. Preliminary identification results of abnormal geological structures in deep space are then obtained based on the judgment matrix Y. The DS evidence theory, as an extension of Bayesian reasoning, is a complete theory that can handle the ambiguity of data information. This theory uses a priori probability distribution function to obtain a posteriori evidence interval to reduce the unknown and uncertainty of fuzzy information. However, due to the poor accuracy of single-source test data, the sensitivity of the index to local anomaly diagnosis is increased by increasing the number of test source types. Due to the inconsistent dimensions of the heterogeneous response data, the results of the multi-source heterogeneous integrated exploration data must be fused and dimensionless and normalized. Specifically,

[0052] Based on the DS evidence theory, the basic probability matrix M of the multi-source heterogeneous comprehensive survey data response is constructed:

[0053]

[0054] Where: They represent the original geological survey results, high-density electrical interpretation results, seismic detection interpretation results and borehole transient electromagnetic detection interpretation results at the sampling node n respectively; According to the DS evidence theory, when a certain The larger the value, the greater the possibility of anomaly at that location, and the greater the confidence level of that location;

[0055] Establish anomaly recognition fusion index, which is the basic probability function at sampling node n:

[0056]

[0057] Use the above function to process each data in the matrix M;

[0058] In order to eliminate the influence of different evaluation index dimensions, the matrix M is normalized to obtain the standardized matrix Y. The normalization formula is:

[0059] Thus, the judgment matrix Y of the multi-source heterogeneous comprehensive survey data response is constructed:

[0060]

[0061] According to the matrix Y, the fusion vector α=[y m1 y m2 ···y mn ], and the fusion vector α is used as the preliminary identification result of the abnormal geological structure in deep space.

[0062] S4. Entropy weight fusion is used to modify the evaluation index:

[0063] Entropy is a macroscopic quantity unique to information theory. Shannon proposed the concept of information entropy to characterize the complexity of system signals. The greater the degree of disorder of the signal, the greater the information entropy; the smaller the degree of disorder, the smaller the information entropy. Entropy weight fusion is an objective weight distribution method. This method derives more objective indicator weights based on the degree of variation of the evaluation indicators. Entropy weight fusion is used to process the preliminary identification results of step S3 to expand the confidence level of the abnormal point data; the greater the entropy value of the data at a certain position in the evaluation result, the lower the credibility of the evaluation result; conversely, the smaller the entropy value, the greater the credibility of the evaluation result; thereby, the preliminary identification results are corrected, specifically:

[0064] When the evaluation index contains multiple states, let the probability of occurrence of the i-th state be p i , i=1,2,…n, then the entropy of the indicator is expressed as:

[0065]

[0066] The frequency p of abnormal multi-source data in the i-th sampling node in the total information volume is i for:

[0067]

[0068] where p i The relationship satisfied is: xi The amount of self-information represented by I(x i ) is defined as: I(x i )=-log a P i ; where I(x i ) is a random variable, for all information I(x i ) to find the average value, that is, to calculate the mathematical expectation of all information self-information, which is the information entropy at the sampling node, denoted as P(x). The specific formula is as follows:

[0069]

[0070] S5. Identification and positioning of abnormal features of deep-earth space structures:

[0071] The information entropy formula obtained in step S4 is used to process the data of the fusion vector α in step S3 to obtain the multi-source heterogeneous data DS-entropy weight identification vector as shown below:

[0072] P=[p 11 p 12 ···p ij ];

[0073] According to the above vectors, the geological conditions at different locations in the deep earth space required for detection are obtained, thereby accurately obtaining the scope and location of geological structural anomalies.

[0074] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for identifying abnormal structures in deep space based on multi-source heterogeneous data fusion, characterized by: The following steps are involved: S1. Obtain multi-source heterogeneous comprehensive survey data: The deep-earth geological structure signals are collected by geophysical prospecting and drilling methods, and the geophysical prospecting data interpretation results and drilling data interpretation results of deep-earth geological structure anomalies are obtained respectively; and the data from different sources are collected at the same sampling interval; S2. Constructing evaluation indicators for multi-source heterogeneous comprehensive survey data: Based on the required deep-earth space comprehensive exploration data, geological structural anomaly evaluation indicators are constructed as follows: Indicator 1: original geological exploration results; Indicator 2: high-density electrical method interpretation results; Indicator 3: seismic detection interpretation results; Indicator 4: borehole transient electromagnetic detection interpretation results. After the evaluation indicators are constructed, the data corresponding to each indicator is divided at the same scale and analyzed and processed, and then the indicators at different locations in the deep-earth space are assigned values. S3. Fusion evaluation index based on improved DS evidence theory: Based on the DS evidence theory, a basic probability matrix M of the multi-source heterogeneous integrated exploration data response is constructed, and an anomaly identification fusion index is established to process each data in the matrix M. Then, the processed data is normalized to construct the judgment matrix Y of the multi-source heterogeneous integrated exploration data response. Then, based on the judgment matrix Y, the preliminary identification results of abnormal geological structures in deep space are obtained. S4. Entropy weight fusion is used to modify the evaluation index: The entropy weight fusion is used to process the preliminary identification results of step S3 to increase the confidence level of the abnormal point data; The larger the entropy value of the data at a certain position in the evaluation result, the lower the credibility of the evaluation result; conversely, the smaller the entropy value, the greater the credibility of the evaluation result; thus, the preliminary identification result can be corrected; S5. Identification and positioning of abnormal features of deep-earth space structures: According to step S3 and step S4, the geological conditions at different locations in the deep earth space required for detection are obtained, thereby accurately obtaining the range and location of geological structural anomalies.

2. The method for identifying abnormal structures in deep space based on multi-source heterogeneous data fusion according to claim 1 is characterized in that: In step S1, data from different sources are collected at a sampling interval of 1 meter; the geophysical data interpretation results include high-density electrical interpretation results and seismic detection interpretation results, and the drilling data interpretation results are borehole transient electromagnetic detection interpretation results.

3. The method for identifying abnormal structures in deep space based on multi-source heterogeneous data fusion according to claim 1 is characterized in that: The step S3 is specifically as follows: Based on the DS evidence theory, the basic probability matrix M of the multi-source heterogeneous integrated survey data response is constructed: Where: They represent the original geological survey results, high-density electrical interpretation results, seismic detection interpretation results, and borehole transient electromagnetic detection interpretation results at sampling node n respectively; Establish anomaly recognition fusion index, which is the basic probability function at sampling node n: Use the above function to process each data in the matrix M; In order to eliminate the influence of different evaluation index dimensions, the matrix M is normalized to obtain the standardized matrix Y. The normalization formula is: Thus, the judgment matrix Y of the multi-source heterogeneous comprehensive survey data response is constructed: According to the matrix Y, the fusion vector α=[y m1 y m2 ···y mn ], and the fusion vector α is used as the preliminary identification result of the abnormal geological structure in deep space.

4. The method for identifying abnormal structures in deep space based on multi-source heterogeneous data fusion according to claim 3 is characterized in that: In step S4, entropy weight fusion is used to process the preliminary identification result of step S3, specifically: When the evaluation index contains multiple states, let the probability of occurrence of the i-th state be p i , i=1,2,…n, then the entropy of the indicator is expressed as: The frequency p of abnormal multi-source data in the i-th sampling node in the total information volume is i for: where p i The relationship satisfied is: x i The amount of self-information represented by I(x i ) is defined as: I(x i )=-log a P i ; where I(x i ) is a random variable, for all information I(x i ) to find the average value, that is, to calculate the mathematical expectation of all information self-information, which is the information entropy at the sampling node, denoted as P(x). The specific formula is as follows:

5. The method for identifying abnormal structures in deep space based on multi-source heterogeneous data fusion according to claim 4 is characterized in that: In step S5, the geological conditions at different locations in the deep earth space required for detection are obtained according to steps S3 and S4, specifically: The information entropy formula obtained in step S4 is used to process the data of the fusion vector α in step S3 to obtain the multi-source heterogeneous data DS-entropy weight identification vector as shown below: P=[p 11 p 12 ···p ij ]; Based on the above vectors, the scope and location of geological structural anomalies can be accurately obtained.

Images