Priori knowledge embedded seismic wave data automatic interpretation method

Abstract

The invention discloses a priori knowledge embedded seismic wave data automatic interpretation method, and relates to the technical field of geological survey. According to the method, value range calibration of classification labels to which feature data belongs is completed in a normal distribution mathematical statistics mode, and fusion of a data decision-making layer is realized by using a D-S theory, so that secondary classification interpretation of seismic wave data on the basis of clustering is automatically realized by taking a KNN algorithm and the D-S evidence theory as cores; by taking a data analysis result of excavated scale data as a reference and combining with calculation of a clustering center distance, embedding of important priori knowledge of real rock mass quality is realized, so that mapping from a clustering label to a rock mass quality label is completed, and rationality and reliability of label classification are fully ensured; according to the method, the interpretation efficiency of the seismic wave data is remarkably improved, the time and the labor cost are greatly saved, the accuracy and the reliability of an interpretation result are ensured, and the method is suitable for large-scale application to interpretation of the seismic wave data in geological exploration.

Description

Technical Field

[0001] The present invention relates to the field of geological survey technology, in particular to a seismic wave data interpretation technology, and in particular to a priori knowledge embedded seismic wave data automatic interpretation method. Background Art

[0002] At present, highway tunnel construction is facing the trend of increasingly complex engineering geological conditions in the tunnel address area. During the construction process, unfavorable geological conditions such as broken zones, karst caves, mud and water gushing are still the focus of construction, which seriously threatens construction safety. Advanced geological prediction is an important means to explore the geological information of the location during tunnel construction, and its accuracy directly affects the development of safety work during the construction process. Among them, the seismic wave detection method has the characteristics of long prediction distance and non-destructive detection, and can timely predict the quality of the rock and soil ahead of excavation. Therefore, it is one of the commonly used advanced geological methods in the tunnel field. Research on its data interpretation has always been the focus of research in this field.

[0003] Seismic wave detection is essentially a geophysical detection method, that is, a method that uses the propagation signal characteristics of elastic waves in different media to predict the geological conditions and rock property changes in front of the tunnel excavation face. According to the differences in principle details, it is specifically divided into TSP, TST, TGS, etc., but in the data interpretation process, various methods currently mainly perform quantitative calculations on the propagation signal data to obtain equivalent rock physical and mechanical parameters including Young's modulus, Poisson's ratio, shear wave and longitudinal wave velocity, and then use the subjective experience of technicians to judge the rock quality and various adverse geological conditions. This interpretation method has the problem that on the one hand, it is highly subjective, and it is easy to make mistakes and omissions due to the technical staff's display, and the interpretation standards are not unified. On the other hand, the amount of seismic wave data is relatively large, and relying on manual high-reliability interpretation may lead to higher costs and longer construction time. Summary of the invention

[0004] The purpose of the present invention is to overcome the problems of poor accuracy and reliability and long interpretation time in existing seismic wave data interpretation methods, and proposes an automatic interpretation method for seismic wave data embedded with prior knowledge.

[0005] In order to achieve the above-mentioned object of the invention, the present invention provides a method for automatic interpretation of seismic wave data with prior knowledge embedded, comprising the following steps:

[0006] (1) Data preprocessing: Preprocess the raw data of seismic wave detection and divide the preprocessed data into excavated scale data and unexcavated scale data;

[0007] (2) Data analysis: The excavated scale data are processed by averaging, and combined with the directly observed rock mass quality, the mean of the excavated scale data is used as the rock mass quality characteristic value, and a rock mass quality label is assigned (a quantitative representation of prior knowledge);

[0008] The unexcavated scale data are clustered by the K nearest neighbor clustering algorithm to obtain the data clustering results of the unexcavated scale data under unsupervised learning and the cluster centers of each cluster data.

[0009] (3) Interpretation label division: Based on the cluster center of the cluster data, the distance from each cluster center to the rock mass quality characteristic value is calculated, and different rock mass quality labels are assigned to the unexcavated data that have completed clustering according to the size of the distance; the cluster data with the smallest distance from the cluster center to the rock mass quality characteristic value is assigned the same label as the excavated rock mass quality, and based on this, other rock mass quality labels are assigned to the cluster data at other distances;

[0010] (4) Range calibration: Perform normal distribution fitting on the data in different rock mass quality labels to obtain the distribution characteristics of the rock mass quality label data, and obtain the value range of the rock mass quality label data according to the set confidence interval;

[0011] (5) Data fusion and interpretation: Based on the value range of different rock mass quality label data, the DS evidence theory is used to fuse and interpret the seismic wave data to obtain the automatic interpretation results of the seismic wave detection data.

[0012] The invention discloses an automatic interpretation method for seismic wave data with prior knowledge embedded, which not only completes the calibration of the value range of characteristic data belonging to each classification label by means of normal distribution mathematical statistics, but also realizes the fusion of data decision layer by means of DS theory, thereby automatically realizing the secondary classification interpretation of seismic wave data based on clustering with KNN algorithm and DS evidence theory as the core; and also realizes the embedding of important prior knowledge of real rock mass quality by means of the concept of prior knowledge embedding, using the data analysis results of excavated scale data as a reference, and combining the calculation of cluster center distance, thereby completing the mapping of cluster labels to rock mass quality labels, and fully ensuring the rationality and reliability of label classification; the automatic interpretation method of the invention significantly improves the interpretation efficiency of seismic wave data, greatly saves time and labor costs, and ensures the accuracy and reliability of the interpretation results, and is suitable for large-scale application in the interpretation of seismic wave data in geological exploration.

[0013] Wherein, in step (1), the preprocessing includes data rarefaction, feature engineering, correlation analysis, and data standardization.

[0014] In the data collection process of seismic wave detection, elastic waves are continuously recorded, so the data density is usually too high. In order to ensure the efficiency of interpretation and the validity of data, it is necessary to thin out the original collected data; data thinning refers to the processing method of dividing the original data according to the specified distance and taking the mean of the segmented segments as the representative value of the segment; it can be represented by the following formula: Where n is the number of samples in the segmented paragraph, x n is the sample value, is the average value (the value after thinning).

[0015] Among them, in the seismic wave detection results, depending on the specific equipment and technical means (TSP, TRT, TGS, etc.) used, the equivalent rock physical and mechanical parameters collected are slightly different, basically including Young's modulus, Poisson's ratio, shear wave velocity, longitudinal wave velocity, wave velocity ratio and other common and important characteristic indicators, and the fluctuation (discreteness) of the above data is also an important basis for subjective interpretation; in order to reflect the discreteness of the characteristics, the standard deviation of the data in each segmented section can be calculated on the basis of equidistant segmentation in data rarefaction, that is: the feature engineering refers to the processing method of calculating the standard deviation of the data in each segmented section based on the equidistant segmentation in data rarefaction.

[0016] Among them, in the field of machine learning, in order to speed up data fitting and reduce data dimensions, it is usually necessary to perform feature correlation analysis to eliminate features with high correlation; preferably, the correlation analysis is a Pearson feature correlation (Pearson correlation coefficient) analysis algorithm, which is a statistical method for measuring the strength and direction of the linear relationship between two variables; the specific formula is: In the formula, a i and b i are the sample values ​​of the two features, and is the sample mean of the two features, ρ is the Pearson correlation coefficient, and n is the number of samples of the feature; ρ is between -1 and +1, indicating the degree of negative correlation and positive correlation respectively, and the closer it is to 0, the weaker the correlation.

[0017] Among them, data standardization is a general process in the field of mathematical statistics, which means scaling the values ​​of indicators of different dimensions to the interval range of (0, 1) to eliminate the impact of dimensions.

[0018] Among them, the excavated scale data are seismic wave detection data that have been verified through excavation or other means that can directly observe the rock mass quality; the unexcavated scale data are seismic wave detection data that cannot or temporarily have not been verified through direct observation of the rock mass quality.

[0019] Among them, in step (2), the method of calculating the feature mean is the mean method.

[0020] The K-Nearest Neighbors (KNN) clustering algorithm is a basic classification and regression method, that is, finding the k training samples closest to the target sample in the feature space, so as to determine the label of the new sample by the label of the training sample (that is, finding the k training samples (neighbors) closest to the new sample; the category that appears most frequently in these k neighbors will be assigned the label of the new sample); k is a positive integer representing the number of "training samples"; preferably, the distance calculation formula is: In the formula, x and y represent the horizontal and vertical coordinates of the sample respectively. i and i are the coordinate values ​​of sample x and sample y in the i-th dimension, respectively, and n is the number of dimensions of the sample.

[0021] Wherein, in step (4), the normal distribution is also called Gaussian distribution, which is a kind of continuous probability distribution; preferably, its probability density function (PDF) is: In the formula, μ is the mean of the distribution, σ is the standard deviation of the distribution, and σ 2 is the variance of the distribution, x is the value of the random variable, and f is the probability density function.

[0022] The confidence interval is an interval used to estimate the unknown value of a parameter. For a normal distribution, a quantile function can be used to calculate the confidence interval at a specific confidence level. Preferably, the confidence interval is: Where z is the quantile of the normal distribution, μ is the mean of the distribution, σ is the standard deviation of the distribution, and n is the sample size; for example, calculating a 90% confidence interval means that the confidence interval covers the middle 90% of the data and excludes 5% of the data at both ends of the normal distribution.

[0023] Among them, in step (5), the appropriate BPA calculation function is selected through the DS evidence theory during the fusion process, and the conflict factor value of the data sample is calculated; according to the value of the conflict factor, the data with high conflict (k is a value of 0 to 1 according to the requirements, the closer to 1, the higher the conflict, and the value is generally not less than 0.9) is processed to weaken the conflict, and finally according to the maximum membership principle (such as: the probability of a sample belonging to labels 1, 2, and 3 is 0.2, 0.5, and 0.3 respectively, then it is judged that its classification result is label 2) to obtain the fusion result and interpret it, so as to obtain the automatic interpretation result of the seismic wave detection data.

[0024] Preferably, the BPA calculation function is an interval Euclidean distance model, specifically: is the value of the jth evaluation index, and F i =[f i- ,f i + ] is the range value of the i-th rock mass quality label, and the Euclidean distance between the two intervals is: Then the basic probability (BPA) calculation function is:

[0025] Preferably, the conflict reduction process refers to using the average evidence method to reduce the conflict between data. The specific steps are: suppose there are n evidence bodies m to be fused. j , j = 1, 2, ..., n has a high degree of conflict, then use the formula Calculate the average body of evidence and complete the average body of evidence m MEA After solving (A), continue to perform fusion processing using the DS fusion rule, and repeat the iterative conflict weakening process n-1 times to obtain the final fusion result.

[0026] Preferably, the DS evidence theory is: let θ = {F1, F2, ..., Fn} and φ = {A1, A2, ..., A n} are the identification framework and evidence set of the multi-source information fusion problem respectively. Now if there is a function m:2 θ →[0,1] satisfies the condition Where: is an empty set; m(A j (F i )) is evidence A j For the subset F in the identification framework i The support probability of ; then m is called the basic probability assignment function of the recognition framework θ, and the calculated result is called the basic assignment probability BPA (Basic probability assignment); after obtaining the BPA of each evidence body in the evidence set from multiple information sources, these evidences need to be fused according to certain rules; the conventional DS fusion rules are as follows:

[0027]

[0028] , where K is called the conflict factor; The conflict factor K is used in DS evidence theory to describe the conflict between evidences. The larger the K, the greater the conflict between the evidence bodies, which also means that the reliability of the decision-making results after fusion is worse.

[0029] Furthermore, the present invention also proposes an interpretation device, comprising at least one processor and a memory communicatively connected to the at least one processor; the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned automatic interpretation method of seismic wave data embedded with prior knowledge.

[0030] Furthermore, the present invention also proposes a computer-readable medium on which instructions executable by a processor are stored. When the instructions are executed by the processor, the processor executes the above-mentioned method for automatic interpretation of seismic wave data with embedded prior knowledge.

[0031] Furthermore, the present invention also proposes an automatic interpretation method of seismic wave data with prior knowledge embedded in it for use in constructing a geological model.

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. The automatic interpretation method of seismic wave data of the present invention completes the calibration of the value range of feature data belonging to each cluster label by means of normal distribution mathematical statistics, and realizes the fusion of data decision layer by using DS theory, thereby automatically realizing the secondary classification interpretation of seismic wave data based on clustering with KNN algorithm and DS evidence theory as the core, significantly improving the interpretation efficiency of seismic wave data and greatly saving time and labor costs.

[0034] 2. The automatic interpretation method of seismic wave data of the present invention adopts the concept of embedding prior knowledge, uses the data analysis results of the excavated scale data as a reference, and combines the calculation of the cluster center distance to realize the embedding of the important prior knowledge of the actual rock mass quality, thereby completing the mapping of cluster labels to rock mass quality interpretation labels, fully ensuring the rationality and reliability of label classification, and ensuring the accuracy and reliability of the interpretation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a flow chart of the method for automatically interpreting seismic wave data in Example 1 of the present invention;

[0036] Figure 2 This is a heat map of the seismic wave data correlation analysis results in Example 1 of the present invention;

[0037] Figure 3 It is a cluster scatter plot of seismic wave data in Example 1 of the present invention;

[0038] Figure 4 This is a normal distribution fitting effect diagram of the cluster data interpretation label (characteristic wave velocity ratio positive) in Example 1 of the present invention;

[0039] Figure 5 This is a normal distribution fitting effect diagram of the cluster data interpretation label (characteristic shear wave velocity) in Example 1 of the present invention;

[0040] Figure 6 This is a normal distribution fitting effect diagram of the cluster data interpretation label (characteristic longitudinal wave velocity) in Example 1 of the present invention;

[0041] Figure 7 The profile and three-dimensional synthetic model are obtained after the automatic interpretation results of seismic wave data in Example 4 of the present invention are used to construct a geological model. DETAILED DESCRIPTION

[0042] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. However, it should not be understood that the scope of the above subject matter of the present invention is limited to the following embodiments, and all technologies realized based on the content of the present invention belong to the scope of the present invention.

[0043] Embodiment 1:

[0044] This embodiment is based on the automatic interpretation of TGS seismic wave detection data obtained during the construction of an actual tunnel project in Guangxi, and the specific steps include:

[0045] (1) According to the actual situation on site, the three types of rock mass quality will be interpreted based on the detection data, that is, the three types of interpretation labels (rock mass quality labels) are complete rock mass, broken rock mass and weak interlayer;

[0046] After data extraction by TGS supporting equipment, the raw data of seismic wave detection has three indicators, namely shear wave vs, longitudinal wave vp and wave velocity ratio vp / vs;

[0047] Among them, in the data preprocessing process, the equidistant segmentation d is selected as 0.5m when the data is thinned, and the standard deviation of the longitudinal wave vp is calculated in the segmented segment to characterize the degree of discrete fluctuation, and a new feature vpStd is derived. The correlation analysis results based on the total of four features (rock mass quality labels: intact rock mass, broken rock mass, weak interlayer, vpStd) are shown in Figure 2 , and according to Figure 2 The correlation results remove the characteristic longitudinal wave vp;

[0048] Among them, the correlation analysis adopts the Pearson characteristic correlation (Pearson correlation coefficient) analysis algorithm, and the specific formula is: In the formula, a i and b i are the sample values ​​of the two features, and is the sample mean of the two features, ρ is the Pearson correlation coefficient, and n is the number of samples of the feature; ρ is between -1 and +1, indicating the degree of negative correlation and positive correlation respectively, and the closer it is to 0, the weaker the correlation;

[0049] When dividing the pre-processed seismic wave data into two parts, the excavated (tunnel face) scale data and the unexcavated scale data, the excavated scale is defined as the first 5m of the detection range, and the rest is the unexcavated scale data.

[0050] (2) The unexcavated scale data are clustered using the K nearest neighbor clustering algorithm (k = 3) to obtain the data clustering results of the unexcavated scale data under unsupervised learning (the cluster labels are Cluster0, Cluster1, and Cluster2. The clustering results are as follows: Figure 3 As shown) and the cluster centers of each cluster data;

[0051] (3) Based on the cluster center of the cluster data, the distance from each cluster center to the rock mass quality characteristic value is calculated, and different rock mass quality labels are assigned to the unexcavated data that have completed clustering according to the size of the distance; among them, the cluster data with the smallest distance from the cluster center to the rock mass quality characteristic value is assigned the same label as the excavated rock mass quality, and based on this, the cluster data with other distances are assigned other rock mass quality labels (the rock mass quality label is "broken rock mass", the calculated D1=1.91, D2=0.72, D3=1.44, so the division results of the rock mass quality interpretation label Cluster 1 is "broken rock mass", Cluster 2 is "intact rock mass", and Cluster 0 is "weak interlayer");

[0052] (4) When performing normal distribution mathematical statistics and value range division on the data in different rock mass quality labels (unexcavated rock mass quality characteristic values), the normal distribution fitting effect is as follows: Figure 4 , Figure 5 , Figure 6 As shown, a 90% confidence interval is selected, and the final calibration results of the value range of each rock mass quality label data are shown in Table 1;

[0053] Among them, the normal distribution is a kind of continuous probability distribution, and its probability density function (PDF) is: In the formula, μ is the mean of the distribution, σ is the standard deviation of the distribution, and σ 2 is the variance of the distribution, x is the value of the random variable, and f is the probability density function;

[0054] The confidence interval is: Where z is the quantile of the normal distribution, μ is the mean of the distribution, σ is the standard deviation of the distribution, and n is the sample size.

[0055] Table 1 The value range of rock mass quality label data under 90% confidence interval

[0056] feature Cluster 2 relatively complete rock mass Cluster 1 Broken rock mass Cluster0 Weak interlayer <![CDATA[v p / in s ]]> -0.07～1.85 -1.92～0.76 -1.43～1.4 <![CDATA[v s ]]> -2.19～0.14 -0.81～1.47 0.23～1.69 <![CDATA[v p Std]]> -1.10～0.13 -1.09～0.54 -0.21～3.53

[0057] (5) During the fusion process, the appropriate BPA calculation function is selected through the DS evidence theory, and the conflict factor value of the data sample is calculated; according to the value of the conflict factor, the data with high conflict (the conflict factor threshold is set to K = 0.90, the proportion of samples corrected by the DS evidence theory is 13% of the total samples, and the average conflict factor K of all samples is 0.92, which belongs to the category of high conflict) is processed to weaken the conflict. Finally, the fusion result is determined and interpreted according to the maximum membership principle, thereby obtaining the automatic interpretation result of the seismic wave detection data.

[0058] Among them, the BPA calculation function is the interval Euclidean distance model, specifically: Assume is the value of the jth evaluation index, and F i =[f i - ,f i + ] is the range value of the i-th rock mass quality label, and the Euclidean distance between the two intervals is: Then the basic probability (BPA) calculation function is:

[0059] The specific steps of conflict reduction processing are as follows: Suppose there are n pieces of evidence m to be fused j , j = 1, 2, ..., n has a high degree of conflict, then use the formula Calculate the average body of evidence and complete the average body of evidence m MEA After solving (A), continue to perform fusion processing using the DS fusion rule, and repeat the iterative conflict weakening process n-1 times to obtain the final fusion result.

[0060] Example 2

[0061] An interpretation device includes at least one processor and a memory connected to the at least one processor in communication; the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute an automatic interpretation method of seismic wave data embedded with prior knowledge in Example 1.

[0062] Example 3

[0063] A computer-readable medium stores instructions executable by a processor, and when the instructions are executed by the processor, the processor executes a method for automatically interpreting seismic wave data with prior knowledge embedded in embodiment 1.

[0064] Example 4

[0065] The final interpretation results obtained in Example 1 are visualized (modeled), including 5 two-dimensional sections selected at the tunnel level and 1 synthetic three-dimensional solid model (such as Figure 7 shown).

[0066] Finally, it should be noted that the embodiments described in detail above are only the best practices of the invention and cannot be used to limit the scope of rights of the invention. Equivalent replacement of the technical solutions recorded in the aforementioned embodiments does not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the invention, and they should all be included in the scope of the claims and description of the invention.

Claims

1. A method for automatic interpretation of seismic wave data with prior knowledge embedded, characterized in that: The following steps are involved: (1) Data preprocessing: Preprocess the raw data of seismic wave detection and divide the preprocessed data into excavated scale data and unexcavated scale data; (2) Data analysis: The excavated scale data are processed by averaging, and combined with the directly observed rock mass quality, the mean of the excavated scale data is used as the rock mass quality characteristic value, and a rock mass quality label is assigned; The unexcavated scale data are clustered by the K nearest neighbor clustering algorithm to obtain the data clustering results of the unexcavated scale data under unsupervised learning and the cluster centers of each cluster data. (3) Interpretation label division: Based on the cluster center of the cluster data, the distance from each cluster center to the rock mass quality characteristic value is calculated, and different rock mass quality labels are assigned to the unexcavated data that have completed clustering according to the size of the distance; the cluster data with the smallest distance from the cluster center to the rock mass quality characteristic value is assigned the same label as the excavated rock mass quality, and based on this, other rock mass quality labels are assigned to the cluster data at other distances; (4) Range calibration: Perform normal distribution fitting on the data in different rock mass quality labels to obtain the distribution characteristics of the rock mass quality label data, and obtain the value range of the rock mass quality label data according to the set confidence interval; (5) Data fusion and interpretation: Based on the value range of different rock mass quality label data, the DS evidence theory is used to fuse and interpret the seismic wave data to obtain the automatic interpretation results of the seismic wave detection data.

2. The method for automatic interpretation of seismic wave data according to claim 1, characterized in that: The preprocessing includes data rarefaction, feature engineering, correlation analysis, and data standardization.

3. The method for automatic interpretation of seismic wave data according to claim 1, characterized in that: In step (4), the normal distribution is a kind of continuous probability distribution; the density function of the probability is: In the formula, μ is the mean of the distribution, σ is the standard deviation of the distribution, and σ 2 is the variance of the distribution, x is the value of the random variable, and f is the density function of the probability.

4. The method for automatic interpretation of seismic wave data according to claim 1, characterized in that: In step (4), the confidence interval is: Where z is the quantile of the normal distribution, μ is the mean of the distribution, σ is the standard deviation of the distribution, and n is the sample size.

5. The method for automatic interpretation of seismic wave data according to claim 1, characterized in that: In step (5), the fusion interpretation includes: selecting a suitable BPA calculation function through the DS evidence theory, and calculating the conflict factor value of the data sample; performing conflict weakening processing on the highly conflicting data according to the value of the conflict factor, and finally determining and interpreting the fusion result according to the maximum membership principle to obtain the automatic interpretation result of the seismic wave detection data.

6. The method for automatic interpretation of seismic wave data according to claim 5, characterized in that: The BPA calculation function is an interval Euclidean distance model, specifically: is the value of the jth evaluation index, and F i =[f i - ,f i + ] is the range value of the i-th rock mass quality label, and the Euclidean distance between the two intervals is: Then the basic probability calculation function is:

7. The method for automatic interpretation of seismic wave data according to claim 5, characterized in that: The conflict reduction process refers to using the average evidence method to reduce the conflict between data. The specific steps are: suppose there are n evidence bodies m to be fused. j , j = 1, 2, ..., n has a high degree of conflict, then use the formula Calculate the average body of evidence and complete the average body of evidence m MEA After solving (A), continue to perform fusion processing using the DS fusion rule, and repeat the iterative conflict weakening process n-1 times to obtain the final fusion result.

8. An automatic interpretation device for seismic wave data, characterized in that: It includes at least one processor and a memory that is communicatively connected to the at least one processor; the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for automatic interpretation of seismic wave data as described in any one of claims 1-7.

9. A computer-readable medium, characterized in that Instructions executable by a processor are stored thereon, and when the instructions are executed by the processor, the processor executes the method for automatic interpretation of seismic wave data as described in any one of claims 1-7.

10. An application of the automatic interpretation method of seismic wave data embedded with prior knowledge as claimed in any one of claims 1 to 7 in constructing a geological model.

Images