Quantitative analysis method for seismic data quality indexes

Abstract

The invention relates to the technical field of seismic data analysis, and discloses a seismic data quality index quantitative analysis method, which comprises the following steps: acquiring geologic feature parameters of an area to which a target station belongs; acquiring seismic waveform data of an area to which the target station belongs, and performing spike detection processing on the seismic waveform data to obtain power spectrum density and a spike event list corresponding to the seismic waveform data; carrying out coupling feature deep fusion processing on the site fundamental frequency, the fault distance, the power spectrum and the spike event list, and respectively determining a geological resonance related noise index, a fault spike interference index and a coupling nonlinear interaction index; performing correlation analysis processing on the geological resonance correlation noise index, the fault zone spine interference index and the coupling nonlinear interaction index to obtain a joint anomaly score; and comparing the joint anomaly score with a preset anomaly warning value to obtain seismic data quality.

Description

Technical Field

[0001] The present invention relates to the technical field of seismic data analysis, and more specifically, it relates to a method for quantitatively analyzing seismic data quality indicators. Background Art

[0002] Quantitative analysis of seismic data quality indicators is the core technology to ensure the reliability of the earthquake early warning system. A method and device for evaluating seismic acquisition data with the publication number CN102012522B. By obtaining shot-receiver relationship data including shot point and geophone point data and seismic survey network data including wire harness data; determining shot energy, signal-to-noise ratio, and resolution according to the shot-receiver relationship data; generating a single wire harness plane spread diagram and / or a multi-wire harness plane spread diagram based on the shot energy, signal-to-noise ratio, resolution, and seismic survey network data; obtaining a satellite remote sensing map including the geomorphic feature information of the work area; evaluating the seismic acquisition data according to the single wire harness plane spread diagram and / or the multi-wire harness plane spread diagram and the satellite remote sensing map. Through this invention, the accuracy and efficiency of evaluating seismic acquisition data can be improved. However, the above solution lacks the deep coupling between geological features and signal features and has the following problems: In thick sedimentary layer areas, low-frequency resonance noise is easily induced due to the geological structure, and traditional indicators cannot accurately distinguish resonance noise from effective seismic waves, resulting in a relatively high false alarm rate; stations near fault zones are affected by the electrochemical effect of lithological interfaces and often have high-frequency spike interferences. However, the existing methods only count the spike frequencies and do not relate to the fault distance, resulting in a relatively high miss rate; The early warning network adopts a unified threshold standard and does not dynamically adjust according to the geological parameters (sedimentary layer thickness, fault distance) of the stations. For example, the noise characteristics of mountain stations and plain stations are significantly different, but the existing system cannot be automatically adapted, resulting in the threshold adjustment relying on manual experience.

[0003] In summary, the existing seismic data quality assessment methods are insufficient in reliability in complex geological scenarios, and it is urgent to introduce geological environment parameters to construct a quantitative analysis system that couples geology and signals to improve the anomaly recognition accuracy and regional adaptation ability. Summary of the Invention

[0004] The present invention provides a method for quantitatively analyzing seismic data quality indicators to solve the technical problems proposed in the background art.

[0005] The present invention provides a method for quantitatively analyzing seismic data quality indicators, including: Step 1, obtaining the geological feature parameters of the area where the target station is located; wherein, the geological feature parameters include: sedimentary layer thickness, site average shear wave velocity, position coordinates, and active fault coordinates; determining the site fundamental frequency of the target station based on the sedimentary layer thickness and the site average shear wave velocity, and determining the fault distance of the target station based on the position coordinates and the active fault coordinates; Step 2: Obtain the seismic waveform data of the area where the target station is located, and perform spike detection processing on the seismic waveform data to obtain the power spectral density and spike event list corresponding to the seismic waveform data; Step 3: Perform coupled feature depth fusion processing on the site fundamental frequency, fault distance, power spectrum, and spike event list to determine the geological resonance-related noise index, fault zone spike interference index, and coupled nonlinear interaction index respectively; Step 4: Perform correlation analysis processing on the geological resonance-related noise index, fault zone spike interference index, and coupled nonlinear interaction index to obtain the joint anomaly score; Step 5: Compare the joint anomaly score with the preset anomaly warning value to obtain the seismic data quality.

[0006] Furthermore, determining the site fundamental frequency of the target station based on the sediment layer thickness and the average shear wave velocity of the site includes: Determine the site fundamental frequency of the target station based on the sediment layer seismic response theory as follows: Wherein, represents the site fundamental frequency of the target station, represents the average shear wave velocity of the site, represents the sediment layer thickness.

[0007] Furthermore, determining the fault distance of the target station based on the position coordinates and the active fault coordinates includes: Take the Euclidean distance between the position coordinates and the active fault coordinates as the fault distance of the target station.

[0008] Furthermore, performing spike detection processing on the seismic waveform data to obtain the power spectral density and spike event list corresponding to the seismic waveform data includes: The method for obtaining the power spectral density includes: Segment the seismic waveform data to obtain data sub-segments, and determine the power spectral density based on the Welch power spectrum estimation method as follows: Wherein, represents the power spectral density, represents the number of data sub-segments, represents the Hamming window of the k-th segment of data, represents the k-th data sub-segment, represents the Fourier transform; The method for obtaining the spike event list includes: Determine the seismic waveform data at any two adjacent sampling times in the seismic waveform data, and perform a first-order difference to obtain the corresponding waveform amplitude; Calculate the median absolute deviation based on the waveform amplitude as follows: Wherein, represents the median absolute deviation, represents function, represents the waveform amplitude at the i-th adjacent sampling moment, represents the median of the waveform amplitudes at all adjacent sampling moments; If , then is regarded as a spike event; Judge the waveform amplitude at each adjacent sampling moment to form a spike event list.

[0009] Furthermore, the geological resonance related noise index includes: Determine the fundamental frequency of the g-th harmonic frequency band, , select the resonance half-bandwidth of each harmonic frequency band, ; Integrate the power spectral density to obtain the energy of the g-th harmonic frequency band; For the energy of each harmonic frequency band, perform weighted summation and calculate the integral with the power spectral density of all harmonic frequency bands to obtain the geological resonance related noise index; Wherein, represents the geological resonance related noise index, represents the harmonic weight, and represent the upper and lower limits of the effective frequency range, represents the integration variable.

[0010] Furthermore, the fault zone spike interference index is as follows: Wherein, represents the fault zone spike interference index, represents the amplitude weight, Represents the event weight, Represents the spatial weight, Represents the list of spike events, Represents the index of the list of spike events, Represents the waveform amplitude of the Represents the median absolute deviation of the Represents the current time, Represents the time corresponding to the Represents the time decay constant, Represents the fault distance of the target station, Represents the spatial decay constant.

[0011] Furthermore, the coupled non - linear interaction index includes: Synchronously obtain the geological resonance - related noise index and the fault - zone spike interference index at fixed time intervals to obtain the geological resonance - related noise sequence and the fault - zone spike interference sequence ; Calculate the Pearson correlation coefficient of the geological resonance - related noise sequence and the fault - zone spike interference sequence , , , Represents the covariance of the correlation noise sequence and the fault - zone spike interference sequence , Represents the standard deviation of the correlation noise sequence , Represents the standard deviation of the fault - zone spike interference sequence ; Normalize the geological resonance - related noise sequence and the fault - zone spike interference sequence respectively and divide them into 20 equal - width histograms, and calculate the KL divergence, as follows: Among them, Represents the KL divergence, Represents the bins interval of the histogram, Represents the probability distribution of the histogram corresponding to the geological resonance - related noise sequence , Represents the probability distribution of the histogram corresponding to the fault - zone spike interference sequence ; Based on the Pearson correlation coefficient and the KL divergence Calculate the coupled non - linear interaction index as follows: Wherein, represents the coupled non - linear interaction index.

[0012] Furthermore, perform correlation analysis on the geological resonance - related noise index, fault - zone spike interference index, and coupled non - linear interaction index to obtain a joint anomaly score, including: Normalize the geological resonance - related noise index , the fault - zone spike interference index and the coupled non - linear interaction index respectively as follows: Wherein, represents the normalization parameter of the geological resonance - related noise index , represents the normalization parameter of the fault - zone spike interference index , represents the normalization parameter of the coupled non - linear interaction index ; and represent the mean and standard deviation of the geological resonance - related noise sequence , and represent the mean and standard deviation of the fault - zone spike interference sequence , and represent the mean and standard deviation of the coupled non - linear interaction sequence ; Wherein, the coupled non - linear interaction sequence is the sequence formed corresponding to the geological resonance - related noise sequence and the fault - zone spike interference sequence , represents the size of the sliding time window for extracting the geological resonance - related noise sequence and the fault - zone spike interference sequence ; Calculate the joint anomaly score as follows: Wherein, represents the joint anomaly score, represents the index vector, represents , and covariance matrix, represents the inverse matrix of the covariance matrix.

[0013] Furthermore, comparing the joint anomaly score with a preset anomaly warning value to obtain the seismic data quality, including: Based on experts to set the preset anomaly warning value; If the joint anomaly score is greater than or equal to the preset anomaly warning value, then trigger the quality warning of the seismic data quality.

[0014] The beneficial effects of the present invention are as follows: By coupling the site fundamental frequency and the fault distance depth into the power spectrum and spike event analysis, constructing coupling indexes such as the multi-harmonic geological resonance coefficient, the spatio-temporal amplitude three-power spike density factor, and the coupling non-linear interaction index, realizing the accurate quantitative identification of resonance noise and fault interference under complex geological conditions, significantly improving the sensitivity and reliability of seismic data quality assessment, reducing the false alarm and missed alarm rates, and having the ability of automatic regional adaptive threshold adjustment. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flowchart of a method for quantifying and analyzing seismic data quality indicators of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] Now, the subject matter described herein will be discussed with reference to exemplary embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein, and the functions and arrangements of the elements discussed can be changed without departing from the protection scope of the content of this specification. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described relative to some examples can also be combined in other examples.

[0017] As Figure 1 shown, a method for quantifying and analyzing seismic data quality indicators includes: Step 1, obtaining geological characteristic parameters of the area where the target station is located; wherein, the geological characteristic parameters include: sediment layer thickness, average shear wave velocity of the site, position coordinates, and active fault coordinates; determining the site fundamental frequency of the target station based on the sediment layer thickness and the average shear wave velocity of the site, and determining the fault distance of the target station based on the position coordinates and the active fault coordinates; Step 2, obtaining seismic waveform data of the area where the target station is located, and performing spike detection processing on the seismic waveform data to obtain the power spectral density and spike event list corresponding to the seismic waveform data; Step 3: Perform coupled feature depth fusion processing on the site fundamental frequency, fault distance, power spectrum, and spiky event list, and respectively determine the geological resonance-related noise index, fault zone spiky interference index, and coupled non-linear interaction index; Step 4: Perform correlation analysis processing on the geological resonance-related noise index, fault zone spiky interference index, and coupled non-linear interaction index to obtain a joint anomaly score; Step 5: Compare the joint anomaly score with a preset anomaly warning value to obtain the seismic data quality.

[0018] In an embodiment of the present invention, determining the site fundamental frequency of the target station based on the sediment layer thickness and the site average shear wave velocity includes: Based on the sediment layer seismic response theory, determine the site fundamental frequency of the target station as follows: Wherein, represents the site fundamental frequency of the target station, represents the site average shear wave velocity, represents the sediment layer thickness.

[0019] Specifically, through the sediment layer seismic response theory, establish a quantitative relationship between the physical properties (thickness, shear wave velocity) of the sediment layer and the site fundamental frequency, providing support for subsequent analysis of sediment layer resonance noise. By clarifying the site fundamental frequency, the frequency-selective amplification effect of the geological structure below the station on seismic waves can be further evaluated, thereby improving the accuracy of seismic data quality assessment.

[0020] The site fundamental frequency of the target station reflects the inherent resonance frequency of the geological structure below the station and is a key parameter for judging whether the sediment layer will cause low-frequency resonance noise. For example, a loose and thick sediment layer corresponds to a low , which is prone to resonance amplification of low-frequency seismic waves, resulting in noise interference.

[0021] The site average shear wave velocity reflects the elastic properties of the site medium. The higher the shear wave velocity, the faster the seismic wave propagates in the sediment layer. At the same thickness, the site fundamental frequency is higher.

[0022] The sediment layer thickness directly affects the fundamental frequency. The greater the thickness, the lower the fundamental frequency , and the more prone it is to cause resonance effects in the low-frequency band.

[0023] In an embodiment of the present invention, determining the fault distance of the target station based on the position coordinates and the active fault coordinates includes: Take the Euclidean distance between the position coordinates and the active fault coordinates as the fault distance of the target station.

[0024] Specifically, the position coordinates represent the position of the target station in the plane coordinate system (usually converted from longitude and latitude), which are the basic parameters for determining the spatial position of the station.

[0025] The active fault coordinates represent the position coordinates of the active faults in the area in the plane coordinate system, reflecting the spatial distribution of the faults.

[0026] The fault distance is used to measure the spatial proximity between the target station and the active fault, because the geological tectonic activities near the active fault (such as lithological changes, electrochemical effects, etc.) will interfere with the quality of the seismic data of the station (spike noise).

[0027] In an embodiment of the present invention, spike detection processing is performed on the seismic waveform data to obtain the power spectral density and the spike event list corresponding to the seismic waveform data, including: The acquisition method of the power spectral density includes: The seismic waveform data is segmented to obtain data sub-segments, and the power spectral density is determined based on the Welch power spectral estimation method as follows: Among them, represents the power spectral density, represents the number of data sub-segments, represents the Hamming window of the k-th segment of data, represents the k-th data sub-segment, represents the Fourier transform; The acquisition method of the spike event list includes: Determine the seismic waveform data at any two adjacent sampling times in the seismic waveform data, and perform a first-order difference to obtain the corresponding waveform amplitude; Calculate the median absolute deviation based on the waveform amplitude as follows: Among them, represents the median absolute deviation, represents function, represents the waveform amplitude at the i-th adjacent sampling time, represents the median of the waveform amplitudes at all adjacent sampling times; If , then is used as a spike event; Judge the waveform amplitude at each adjacent sampling time to form a spike event list.

[0028] Specifically, the seismic waveform data is segmented and the power spectral density is calculated based on the Welch power spectrum estimation method to reduce spectral leakage and estimation variance, and more accurately analyze the energy distribution of the seismic waveform at different frequencies.

[0029] represents the power spectral density, reflecting the energy distribution of the seismic waveform at the frequency and is used to identify the frequency bands where noise is concentrated. represents the number of data sub - segments, which improves the stability of the power spectrum estimation through multi - segment averaging and reduces the randomness impact of a single data segment. represents the Hamming window of the nth segment of data, which is used to window the data sub - segments to reduce spectral leakage (i.e., the phenomenon that signal energy spreads to adjacent frequencies in spectral analysis), making the power spectrum estimation more accurate.

[0030] Formula Based on the Welch power spectrum estimation method, which is a classic spectral estimation technique. By segmenting, windowing, and averaging the data, it effectively reduces spectral leakage and estimation variance, is suitable for the analysis of non - stationary signals such as seismic waveforms, and can more reliably obtain the power spectral density.

[0031] Perform a first - order difference on the seismic waveform data at adjacent sampling times to highlight the abrupt change in waveform amplitude and facilitate the detection of spike events. Calculate the median absolute deviation to determine the threshold for spike events. This metric is insensitive to outliers and can effectively suppress the interference of random noise. By judging spike events, the amplitude changes that meet the conditions are identified as spikes, forming a list of spike events.

[0032] represents the difference in waveform amplitude at the i - th adjacent sampling time, which is used to detect the abrupt change in waveform amplitude. represents the median absolute deviation, which measures the dispersion of the waveform amplitude difference sequence and serves as the basis for determining the threshold for spike events, being highly robust to outliers. represents the median of the waveform amplitude differences at all adjacent sampling times, which is used to calculate .

[0033] The first - order difference can directly reflect the change in waveform amplitude at adjacent times and is the basic operation for detecting spikes (i.e., abrupt amplitude changes). The median absolute deviation is less sensitive to outliers compared to the standard deviation and is more suitable for signal processing of seismic data that may contain sudden interferences (spikes). Using Determine spike events, where 6 times is the experience value, which can effectively suppress random noise in practical applications and retain real spike events (such as amplitude mutations caused by geological interference or equipment failures), for forming a reliable spike event list.

[0034] In an embodiment of the present invention, the geological resonance-related noise index includes: Determine the site fundamental frequency of the g-th harmonic frequency band, , select the resonance half-bandwidth of each harmonic frequency band , ; Integrate the power spectral density to obtain the energy of the g-th harmonic frequency band ; For the energy of each harmonic frequency band perform weighted summation and calculate the integral with the power spectral density of all harmonic frequency bands to obtain the geological resonance-related noise index; Based on the sediment layer seismic resonance theory, determine the calculation formula of the geological resonance-related noise index as follows: Wherein, represents the geological resonance-related noise index, represents the harmonic weight, and represent the upper and lower limits of the effective frequency range, represents the integration variable.

[0035] Specifically, determine the harmonic frequency band and the half-bandwidth, so as to focus on the 1st to 3rd harmonic frequency bands of the site fundamental frequency (the harmonic order takes 1, 2, 3, representing the fundamental frequency and its 2nd and 3rd harmonics, and the low-frequency harmonics have a significant impact on the resonance of the thick sediment layer). By setting the resonance half-bandwidth , clarify the core frequency range of the sediment layer resonance, avoid interference from other frequency bands, and provide a target interval for subsequent energy calculation. The traditional seismic data quality assessment does not pay attention to the correlation between multi-order harmonics and geological resonance. Select the 1st to 3rd harmonics based on the geological characteristics of the thick sediment layer (this range covers the main resonance frequency bands of the thick sediment layer), and through define the half-bandwidth (determined by combining the frequency-selective amplification characteristics of seismic waves by the geological structure, which can effectively separate the frequency band where the resonance energy is concentrated and avoid the blindness of frequency band division in traditional methods).

[0036] Integrate the power spectral density within each harmonic frequency band to calculate the total energy of this frequency band , quantify the energy distribution of the resonance frequency band. Combine the geological resonance theory with the integral operation in signal processing, and through Calculate the energy of a specific harmonic frequency band, rather than simply analyzing the energy of the entire frequency band. This targeted energy calculation method fully considers the resonance amplification effect of the geological structure on specific frequencies and is an innovative application of the traditional signal energy analysis method. This is a basic application of integration. By integrating the power spectral density within the harmonic frequency band and accumulating the energy of this frequency band, it satisfies the physical principle of energy calculation.

[0037] Calculate the geological resonance-related noise index, and then perform a weighted sum of the energies of each harmonic frequency band (weights ), and then divide by the integral of the power spectral density of the entire frequency band to obtain the geological resonance-related noise index, highlighting the contribution of the low-frequency main resonance and quantifying the proportion of resonance noise in the total energy.

[0038] By introducing as the harmonic weight, based on the characteristics of the resonance physical mechanism of the sedimentary layer, low-frequency harmonics (such as the fundamental frequency plays a dominant role in the resonance of thick sedimentary layers, and the contribution of higher-order harmonics decreases rapidly with the increase of g. The traditional method does not perform differential weighting of the harmonic contributions.

[0039] Formula Combines the geological resonance energy (weighted harmonic energy) with the energy of the entire frequency band to form an index that can reflect the proportion of geological resonance noise, realizing the quantitative correlation between geological resonance characteristics and signal energy.

[0040] In an embodiment of the present invention, the fault zone spike interference index is as follows: Among them, represents the fault zone spike interference index, represents the amplitude weight, represents the event weight, represents the spatial weight, represents the spike event list, represents the index of the spike event list, represents the th waveform amplitude of the spike event, represents the th median absolute deviation of the spike event, represents the current moment, represents the The moment corresponding to a spike event Represents the time decay constant Represents the fault distance of the target station Represents the spatial decay constant

[0041] Specifically, by Normalize the amplitude of the spike event So that the larger the amplitude of the spike, the higher its proportion in the index, highlighting the impact of large spikes on data quality

[0042] In traditional seismic data quality assessment, only the number of spike events is often concerned, while the amplitude difference is ignored. By introducing the amplitude weight, based on (A dispersion index insensitive to outliers) for normalization, so that large spikes (such as strong interference caused by drastic lithological changes in the fault zone) have a higher weight in the index, more accurately reflecting the actual harm degree of spike interference

[0043] Represents the Waveform amplitude of the s-th spike event, directly reflecting the intensity of the spike Represents the median absolute deviation of the s-th spike event, used to measure the dispersion degree of the data sequence where the spike event is located Through normalization, the amplitude Is associated with the dispersion degree of the data sequence

[0044] By Calculate the time weight, so that the spike events that occurred recently have a higher weight, reflecting the timeliness of interference (recent spikes can better reflect the current interference state of the station). Traditional methods usually treat all historical spike events equally and do not consider the time decay effect. By introducing an exponential decay model, simulating the weakening of the interference effect over time (such as spikes caused by short-term electrochemical activities in the fault zone, the more recent ones can better reflect the current geological activity state), making the index more in line with the actual dynamic changes

[0045] Represents the current moment, serving as the time reference Represents the moment corresponding to the s-th spike event, used to calculate the time difference Represents the time decay constant (preferably 12h), controlling the decay speed of the time weight, determined through experiments and experience, so that the time weight does not decay too fast (missing important recent interferences) nor too slow (retaining the influence of historical interferences for too long)

[0046] Exponential decay formula It is widely used in physics and engineering to simulate phenomena that decay over time. Here, it is used for time-weight calculation, which conforms to the law that the interference effect weakens over time, and is easy to calculate and flexible in parameter adjustment.

[0047] By Calculating the spatial weight makes the stations closer to the fault have higher weights for spike events, reflecting the spatial correlation of the fault zone's interference with the stations (stations near the fault are more susceptible to interference such as lithology and electrochemical effects). Traditional seismic data quality assessment does not fully utilize the spatial distance information between stations and the fault. By incorporating spatial factors into the index, based on the geological understanding that the closer to the fault, the higher the likelihood of being interfered, the spatial influence is quantified through an exponential decay model, enabling the index to have spatial resolution ability. Represents the spatial decay constant (preferably 10 km), which controls the decay rate of the spatial weight and is determined by analyzing different geological regions, enabling the spatial weight to reasonably reflect the spatial influence range of fault interference. Is the application of the exponential decay model in spatial weight calculation, satisfying the condition that the closer to the fault, the higher the likelihood of being interfered.

[0048] By Summing the triple-weight products of all spike events, integrating the three dimensions of amplitude, time, and space, comprehensively quantifying the degree of spike interference in the fault zone. By 、 、 Combining the three innovative weights to form a multi-dimensional comprehensive index. Traditional methods only evaluate interference from a single dimension (such as only looking at the number of spikes or only looking at the amplitude), while this index realizes the three-dimensional fusion of "amplitude, time, and space" for the first time, more comprehensively and meticulously depicting the characteristics of spike interference in the fault zone (such as simultaneously capturing high-risk spike events that are near the fault, recent, and large in amplitude).

[0049] Represents the spike interference index of the fault zone, and the larger the value, the more serious the spike interference in the fault zone. Represents the total number of spike events in the spike event list, serving as the upper limit of the summation to ensure that all spike events are involved in the index calculation.

[0050] In an embodiment of the present invention, a coupled non-linear interaction index includes: Synchronously obtaining the geological resonance-related noise index and the fault zone spike interference index at fixed time intervals to obtain the geological resonance-related noise sequence and the fault zone spike interference sequence ; Calculating the Pearson correlation coefficient of the geological resonance-related noise sequence and the fault zone spike interference sequence , , Represents the correlated noise sequence and the fault zone spike interference sequence covariance of Represents the correlated noise sequence standard deviation of Represents the fault zone spike interference sequence standard deviation of; Normalize the geological resonance correlated noise sequence and the fault zone spike interference sequence respectively into 20 equally wide histograms, and calculate the KL divergence, as follows: where represents the KL divergence, represents the bins interval of the histogram, represents the probability distribution of the histogram corresponding to the geological resonance correlated noise sequence represents the probability distribution of the histogram corresponding to the fault zone spike interference sequence ; Based on the Pearson correlation coefficient and the KL divergence calculate the coupling non-linear interaction index, as follows: where represents the coupling non-linear interaction index.

[0051] Specifically, synchronously obtain the geological resonance correlated noise index and the fault zone spike interference index at fixed time intervals (preferably 5 minutes) to form a time series.

[0052] Through measure and linear synchrony of. Covariance reflects the common change trend of the two, and the standard deviation and measure the respective dispersion degrees, and the linear correlation degree is obtained after normalization. Analyzing the linear relationship between geological resonance and fault spike interference by introducing the Pearson correlation coefficient is a breakthrough in traditional single-index analysis. For example, at a station with thick sediment layers and close to the fault, if is close to 1, it indicates that the two change synchronously (such as when the resonance increases, the spike interference also increases), providing clues for the geological signal coupling mechanism. covariance of, the larger the value, the stronger the co-directional change trend. and are respectively and ​The standard deviation reflects the amplitude of sequence fluctuations.

[0053] Normalize and and divide them into 20 equal-width histograms, and calculate , which measures the difference in the probability distributions of the two. By introducing the KL divergence (an index for measuring distribution differences in information theory) into the seismic data quality assessment, capture the and nonlinear differences. For example, if the linear correlation coefficient between the two is high but suddenly increases, it indicates a dislocation in the distribution pattern (such as the resonance continues but the spiky distribution mutates), revealing the complex geological signal interaction mechanism.

[0054] The KL divergence is based on the information entropy theory and quantifies the difference between two distributions through . The larger the value, the more significant the distribution difference. In seismic data, it can effectively capture the subtle differences in the value distributions between geological resonances and fault spiky interferences, providing support for nonlinear analysis.

[0055] By combining linear correlation and distribution difference, amplify the abnormal sensitivity in the scenario of linear synchronization and sudden change in distribution difference. Combine the Pearson correlation coefficient and the KL divergence to form . Traditional methods do not consider both linear and nonlinear characteristics simultaneously, while can comprehensively characterize the complex interaction between geological resonances and fault spiky interferences. For example, during a seismic active period, if the linear correlation between the two increases ( increases) and the distribution difference mutates ( increases), will increase significantly, giving an early warning of abnormal data quality. represents the coupled nonlinear interaction index. The larger the value, the more complex the interaction between geological resonances and fault spiky interferences, and the higher the probability of abnormal data quality.

[0056] In an embodiment of the present invention, correlation analysis processing is performed on the geological resonance-related noise index, the fault zone spiky interference index, and the coupled nonlinear interaction index to obtain a combined anomaly score, including: Normalize the geological resonance-related noise index , the fault zone spiky interference index , and the coupled nonlinear interaction index respectively, as follows: where represents the geological resonance-related noise index Normalization parameter of indicating the fault zone spike interference index Normalization parameter of indicating the coupled non - linear interaction index Normalization parameter; and indicating the mean and standard deviation of the geological resonance - related noise sequence and and indicating the mean and standard deviation of the fault zone spike interference sequence and and indicating the mean and standard deviation of the coupled non - linear interaction sequence; wherein, the coupled non - linear interaction sequence is the sequence formed corresponding to the geological resonance - related noise sequence and the fault zone spike interference sequence ; indicating the size of the sliding time window for extracting the geological resonance - related noise sequence and the fault zone spike interference sequence ; Calculate the joint anomaly score as follows: wherein, indicating the joint anomaly score, indicating the index vector, indicating , and covariance matrix of indicating the inverse matrix of the covariance matrix.

[0057] Specifically, since the dimensions of the geological resonance - related noise index , the fault zone spike interference index and the coupled non - linear interaction index are different, direct analysis cannot accurately reflect the comprehensive influence. Eliminate the dimensional differences through normalization.

[0058] Through , Calculate the combined anomaly score. Using the Mahalanobis distance, comprehensively consider the covariance matrix C of each element (the three normalized indicators) in the indicator vector Z, quantify the anomaly degree of the indicator vector in the multi-dimensional space, and comprehensively reflect the comprehensive anomaly situation of the seismic data quality. The traditional method only simply weights and sums the indicators without considering the correlation between the indicators. By introducing the Mahalanobis distance, capture the correlation between and and through the covariance matrix C (the geological resonance of the thick sediment layer station is positively correlated with the fault spike interference).

[0059] In an embodiment of the present invention, compare the combined anomaly score with a preset anomaly warning value to obtain the seismic data quality, including: Based on the expert setting the preset anomaly warning value; If the combined anomaly score is greater than or equal to the preset anomaly warning value, trigger the quality warning of the seismic data quality.

[0060] Specifically, an expert sets a threshold as the benchmark for judging anomalies according to historical data, station geological conditions and seismic data quality standards, providing a clear basis for subsequent evaluation. The preset anomaly warning value is a threshold that combines expert experience and data characteristics, reflecting the judgment standard for seismic data quality anomalies. By comparing the combined anomaly score J and the preset anomaly warning value, judge whether the seismic data quality is abnormal. If is greater than or equal to the preset anomaly warning value, trigger the quality warning, indicating that the data may be affected by geological resonance, fault spikes and other interferences.

[0061] The above describes the embodiments of this embodiment, but this embodiment is not limited to the above specific implementation manners. The above specific implementation manners are only illustrative and not restrictive. Under the inspiration of this embodiment, those of ordinary skill in the art can also make many forms, all of which fall within the protection scope of this embodiment.

Claims

1. A method for quantitatively analyzing seismic data quality indicators, characterized in that Including: Step 1: Obtain the geological characteristic parameters of the area where the target station is located. Among them, the geological characteristic parameters include: sediment layer thickness, average shear wave velocity of the site, position coordinates, and active fault coordinates. Determine the site fundamental frequency of the target station based on the sediment layer thickness and the average shear wave velocity of the site, and determine the fault distance of the target station based on the position coordinates and the active fault coordinates. Step 2: Obtain the seismic waveform data of the area where the target station is located, and perform spike detection processing on the seismic waveform data to obtain the power spectral density and the spike event list corresponding to the seismic waveform data. Step 3: Perform coupled feature depth fusion processing on the site fundamental frequency, fault distance, power spectrum, and spike event list to respectively determine the geological resonance related noise index, fault zone spike interference index, and coupled non-linear interaction index. Step 4: Perform correlation analysis processing on the geological resonance related noise index, fault zone spike interference index, and coupled non-linear interaction index to obtain the joint anomaly score. Step 5: Compare the joint anomaly score with the preset anomaly warning value to obtain the seismic data quality.

2. The quantitative analysis method for seismic data quality index according to claim 1, wherein Determining the site fundamental frequency of the target station based on the sediment layer thickness and the average shear wave velocity of the site includes: Determine the site fundamental frequency of the target station based on the sediment layer seismic response theory as follows: Among them, represents the site fundamental frequency of the target station, represents the average shear wave velocity of the site, represents the thickness of the sedimentary layer.

3. A method for quantitatively analyzing seismic data quality indicators according to claim 2, characterized in that, Determining the fault distance of the target station based on the position coordinates and the active fault coordinates includes: Take the Euclidean distance between the position coordinates and the active fault coordinates as the fault distance of the target station.

4. A method for quantitatively analyzing seismic data quality indicators according to claim 3, characterized in that, Performing spike detection processing on the seismic waveform data to obtain the power spectral density and the spike event list corresponding to the seismic waveform data includes: The acquisition method of the power spectral density includes: Segment the seismic waveform data to obtain data sub-segments, and determine the power spectral density based on the Welch power spectrum estimation method as follows: Among them, represents the power spectral density, represents the number of data sub - segments, represents the Hamming window of the k - th segment of data, represents the k - th data sub - segment, represents the Fourier transform; The acquisition method of the spike event list includes: Determine the seismic waveform data at any two adjacent sampling times in the seismic waveform data, and perform a first-order difference to obtain the corresponding waveform amplitude. Calculate the median absolute deviation based on the waveform amplitude as follows: Among them, represents the median absolute deviation, represents the function, represents the waveform amplitude at the i-th adjacent sampling moment, represents the median of the waveform amplitudes at all adjacent sampling moments; If , then is regarded as a spike event; Judge the waveform amplitude at each adjacent sampling time to form a spike event list.

5. A method for quantitatively analyzing seismic data quality indicators according to claim 4, characterized in that, The geological resonance related noise index includes: Determine the fundamental frequency of the site of the g-th harmonic frequency band, , select the resonance half-bandwidth of each harmonic frequency band , ; Integrate the power spectral density to obtain the energy in the g-th harmonic frequency band ; The energy for each harmonic frequency band Perform weighted summation and calculate the integral with the power spectral density of all harmonic frequency bands to obtain the geological resonance related noise index; Determine the calculation formula of the geological resonance related noise index based on the sediment layer seismic resonance theory as follows: Among them, represents the geological resonance related noise index, represents the harmonic weight, and represent the upper and lower limits of the effective frequency range, represents the integration variable.

6. The quantitative analysis method for seismic data quality indicators according to claim 5, characterized in that The fault zone spike interference index is as follows: Among them, represents the fault zone spike interference index, represents the amplitude weight, represents the event weight, represents the spatial weight, represents the spike event list, represents the index of the spike event list, represents the waveform amplitude of the th spike event, represents the median absolute deviation of the th spike event, represents the current moment, represents the moment corresponding to the th spike event, represents the fault distance of the target station, represents the spatial decay constant.

7. A method for quantitatively analyzing seismic data quality indicators according to claim 6, characterized in that The coupled non-linear interaction index includes: Synchronously obtain the geological resonance-related noise index and fault zone spike interference index at fixed time intervals to obtain a geological resonance-related noise sequence and a fault zone spike interference sequence ; Calculate the Pearson correlation coefficient of the geological resonance-related noise sequence and the fault zone spike interference sequence ; , , denote the covariance of the relevant noise sequence and the fault zone spike interference sequence ; denote the standard deviation of the relevant noise sequence ; denote the standard deviation of the fault zone spike interference sequence ; Normalize the geological resonance-related noise sequence and the fault zone spike interference sequence into 20 equal-width histograms respectively, and calculate the KL divergence as follows: Among them, represents the KL divergence, represents the bins interval of the histogram, represents the geological resonance-related noise sequence the probability distribution of the corresponding histogram, represents the fault zone spike interference sequence the probability distribution of the corresponding histogram; Based on the Pearson correlation coefficient and the KL divergence Calculate the coupled non-linear interaction index as follows: Among them, represents the coupled non-linear interaction index.

8. A method for quantitatively analyzing seismic data quality indicators according to claim 7, characterized in that, Performing correlation analysis processing on the geological resonance related noise index, fault zone spike interference index, and coupled non-linear interaction index to obtain the joint anomaly score includes: Normalize the geological resonance related noise index , the fault zone spike interference index and the coupled non-linear interaction index respectively, as follows: Among them, represents the normalized parameter of the geological resonance related noise index ; represents the normalized parameter of the fault zone spike interference index ; represents the normalized parameter of the coupled non - linear interaction index ; and represent the mean and standard deviation of the geological resonance related noise sequence ; and represent the mean and standard deviation of the fault zone spike interference sequence ; and represent the mean and standard deviation of the coupled non - linear interaction sequence ; Among them, the coupled nonlinear interaction sequence is the geological resonance related noise sequence and the fault zone spike interference sequence corresponding formed sequence, indicating the extraction of the geological resonance related noise sequence and the fault zone spike interference sequence of the sliding time window size; Calculate the joint anomaly score as follows: Among them, represents the combined anomaly score, represents the index vector, represents , and the covariance matrix of represents the inverse matrix of the covariance matrix.

9. A method for quantitatively analyzing seismic data quality indicators according to claim 8, characterized in that Comparing the joint anomaly score with the preset anomaly warning value to obtain the seismic data quality includes: Set the preset anomaly warning value based on experts; If the combined anomaly score is greater than or equal to the preset anomaly warning value, a quality warning for seismic data quality is triggered.

Images