A method, device and equipment for extracting earthquake precursor information and a storage medium

Abstract

The application discloses a kind of earthquake precursor information extraction method, device, equipment and storage medium, it is related to earthquake monitoring technical field, including the following steps: the dynamic gravity signal of the region to be measured is continuously collected;The amplitude of dynamic gravity signal and abnormal early warning threshold are different to obtain the first difference, the amplitude of dynamic gravity signal and abnormal noise threshold are different to obtain the second difference;When the first difference and the second difference are in abnormal interval, start abnormal detection time window to detect the amplitude of dynamic gravity signal, when the duration reaches the set value, the number of times that the amplitude of dynamic gravity signal is in abnormal interval is counted, and the total number of times counted is used as early warning number;If early warning number is greater than set number, calculate the dispersion coefficient of the amplitude of dynamic gravity signal in abnormal interval, and output the amplitude data with dispersion coefficient less than 1 as earthquake precursor information.The application first collects the dynamic gravity signal related to earthquake precursor, then sets multiple control variables, judges the disturbance signal in dynamic gravity signal through multiple control variables, and obtains the judgment result, i.e., earthquake precursor information.The method of the application can accurately obtain earthquake precursor information, and provide research data for earthquake prediction work.

Description

Technical Field

[0001] This invention relates to the field of earthquake monitoring technology, and in particular to a method, apparatus, equipment, and storage medium for extracting earthquake precursor information. Background Technology

[0002] The goal of earthquake prediction is to accurately predict the location, time, and magnitude of an earthquake (the "three elements" of an earthquake) before it occurs, in order to minimize casualties caused by the suddenness of earthquakes. Due to its significant social implications, earthquake prediction has always been a focus of scientific research. Seismologists both domestically and internationally have developed various observation devices by studying geological structures and crustal movements, combined with advancements in information science. These devices support research in seismology, crustal deformation, electromagnetics, and subsurface fluid dynamics, as well as multidisciplinary integrated observation systems.

[0003] Currently, earthquake monitoring and prediction mainly rely on seismic geology, seismic statistics, and earthquake precursor methods. Seismic geology predicts the areas where earthquakes may occur by analyzing geological structures, seismic statistics estimates the probability of earthquakes based on historical data, and earthquake precursor methods predict earthquakes by identifying macroscopic and microscopic signs before an earthquake (such as ground sounds, ground lights, earthquake clouds, groundwater anomalies, and geophysical field anomalies).

[0004] Existing earthquake precursor methods rely heavily on empirical inferences after acquiring macroscopic or microscopic data, resulting in low accuracy of the extracted precursor information in earthquake prediction. Summary of the Invention

[0005] This invention provides a method, apparatus, device, and storage medium for extracting earthquake precursor information, which solves the problem that existing earthquake precursor methods rely mainly on empirical inference for the analysis of macroscopic or microscopic data after acquisition, and this method has limitations in terms of timeliness and accuracy.

[0006] This invention provides a method for extracting earthquake precursor information, comprising the following steps:

[0007] Continuously acquire dynamic gravity signals of the area to be measured;

[0008] Multiple control variables are set, including abnormal warning threshold, abnormal noise threshold, abnormal duration window, number of warnings, and dispersion coefficient;

[0009] The first difference is obtained by subtracting the amplitude of the dynamic gravity signal from the abnormal warning threshold, and the second difference is obtained by subtracting the amplitude of the dynamic gravity signal from the abnormal noise threshold.

[0010] When the first difference and the second difference are in the abnormal range, the abnormal detection time window is opened to detect the amplitude of the dynamic gravity signal. When the duration reaches the set value, the number of times the amplitude of the dynamic gravity signal is in the abnormal range is counted, and the total number of counts is used as the number of warnings.

[0011] If the number of warnings exceeds the set number, the dispersion coefficient of the dynamic gravity signal amplitude in the abnormal range is calculated, and the amplitude data with a dispersion coefficient less than 1 is output as earthquake precursor information.

[0012] Preferably, the control variable further includes a continuous no-abnormality time window. When the first difference and the second difference are not in the abnormal range, the continuous no-abnormality time window is opened for detection. If the duration reaches a set value, the abnormality detection time window is closed.

[0013] Preferably, the abnormal interval is defined as a first difference greater than 1.5 and a second difference less than 28, the set value is 120 minutes, and the set number of times is 700.

[0014] Preferably, dynamic gravity signals are collected by an atmospheric tidal gravimeter with an amplitude of 1–30 mHz.

[0015] Preferably, before obtaining the first difference value by subtracting the amplitude of the dynamic gravity signal from the anomaly warning threshold, the dynamic gravity signal needs to be preprocessed. The preprocessing process includes:

[0016] Perform an integrity check on the dynamic gravity signal and remove any damaged signals;

[0017] Noise values ​​were removed from the dynamic gravity signal after integrity check using the standard fraction method.

[0018] Missing values ​​in the dynamic gravity signal after noise removal are supplemented using a linear interpolation method.

[0019] Correct the dynamic gravity signal after missing values ​​are filled in;

[0020] Data alignment is performed on the corrected dynamic gravity signal;

[0021] The dynamic gravity signal after data alignment is filtered.

[0022] Preferably, the correction of the dynamic gravity signal after the missing values ​​are supplemented includes the correction of gravity parameters based on atmospheric pressure and ocean tides;

[0023] The correction of gravity parameters based on atmospheric pressure and ocean tides is shown in the following formula:

[0024]

[0025]

[0026] in,

[0027]

[0028]

[0029] In the formula, g p To correct for gravity data after atmospheric pressure adjustment, p Atmospheric pressure at the time of observation p n Standard atmospheric pressure g 0 To correct for gravity data after ocean tides, and Here are the latitude and longitude of the observation point, and T is the gravity disturbance. r p This is the radial distance of the observation point from the Earth's center. G It is the gravitational constant. The density of seawater, h Let be a function of the height of ocean tidal waves. As a scaling factor, The spatial relationship between the observation point and the mass. r It is the Earth's average radius or the average distance from the observation point to the Earth's center of mass. h p It is the height of the observation point relative to the Earth's reference surface. The spherical distance angle between the two points is denoted as .

[0030] Preferably, the method further includes extracting time-domain and frequency-domain features from earthquake precursor information to obtain time-domain and frequency-domain features, and then analyzing the earthquake precursor information using these time-domain and frequency-domain features.

[0031] An earthquake precursor information extraction device, comprising:

[0032] The acquisition module is used to continuously acquire dynamic gravity signals of the area to be measured;

[0033] The setting module is used to set multiple control variables, including abnormal warning threshold, abnormal noise threshold, abnormal duration window, number of warnings, and dispersion coefficient;

[0034] The difference module is used to obtain a first difference by subtracting the amplitude of the dynamic gravity signal from the abnormal warning threshold, and to obtain a second difference by subtracting the amplitude of the dynamic gravity signal from the abnormal noise threshold.

[0035] The detection module is used to open an abnormal detection time window to detect the amplitude of the dynamic gravity signal when the first difference and the second difference are in the abnormal range. When the duration reaches a set value, the module counts the number of times the amplitude of the dynamic gravity signal is in the abnormal range and uses the total number of counts as the number of warnings.

[0036] The calculation module is used to calculate the dispersion coefficient of the amplitude of the dynamic gravity signal in the abnormal range if the number of warnings exceeds the set number, and output the amplitude data with a dispersion coefficient of less than 1 as earthquake precursor information.

[0037] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the above-described method for extracting earthquake precursor information.

[0038] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for extracting earthquake precursor information.

[0039] Compared with the prior art, the beneficial effects of the present invention are:

[0040] This invention first continuously acquires dynamic gravity signals related to earthquake precursors. Then, it sets multiple control variables and uses these variables to identify and output abnormal amplitudes in the dynamic gravity signals, thus obtaining earthquake precursor information. The identification results are compared with actual earthquake data, and the control variables are updated accordingly. This extraction method can accurately obtain earthquake precursor information, enhancing the readability and accuracy of earthquake prediction information compared to traditional experience-based inferences, and providing research data for earthquake prediction work. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of a method for extracting earthquake precursor information according to the present invention;

[0043] Figure 2 This is a flowchart of the earthquake precursor signal extraction and processing of the present invention;

[0044] Figure 3 A two-dimensional illustration of a conical region in Continuous Wavelet Transform (CWT);

[0045] Figure 4A three-dimensional illustration of the continuous wavelet transform (CWT). Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] A method for extracting earthquake precursor information, referring to Figure 1 This includes the following steps:

[0048] Step 1: Continuously acquire dynamic gravity signals of the area to be measured.

[0049] Ultra-low frequency dynamic gravity signals are collected using an atmospheric tidal gravimeter. The dynamic atmospheric tidal gravimeter can detect a weak, instantaneous, and variable (1–30 mHz, 1–10 μGal) gravity disturbance signal (dynamic gravity field change signal) to reflect earthquake precursor information. In order to accurately capture earthquake precursor information, interference factors are removed from the collected raw data and effective precursor information reflecting earthquakes is extracted.

[0050] Data preprocessing is performed on the dynamic gravity signals. The data preprocessing stage focuses on ensuring the quality and accuracy of the data collected by the dynamic gravimeter. This includes several key steps such as data cleaning, removal and correction of environmental disturbances, multi-channel data alignment, and high-pass Busterworth filtering. The aim is to prepare a clean, consistent, and accurate dataset for subsequent analysis.

[0051] Data cleaning is performed on the collected dynamic gravity signals. Data cleaning is an important step in data analysis and data science. It involves identifying and correcting (or removing) erroneous and inconsistent data from the raw data to improve its quality and validity. The data cleaning process includes the following steps:

[0052] (1) Check the integrity of the data and remove damaged or missing records.

[0053] (2) Identification and handling of errors and noise values: These may be caused by instrument malfunctions or sudden events (such as power outages or explosions). This includes identifying values ​​that do not conform to the expected pattern of the dataset, such as highly noisy values. This invention uses the Z-score (also known as the standard score) method to detect and remove these noise values. The Z-score is calculated using the following formula:

[0054]

[0055] in:X These are observed values. μ It is the average value (mean) of the dataset. σ It is the standard deviation of the dataset.

[0056] Z-score quantifies the distance of a data point from the mean, measuring this distance in units of standard deviation. This method can identify unusual aspects of a data point relative to the overall dataset. The specific steps are as follows:

[0057] For each data point in the dataset, calculate its Z-score.

[0058] Setting a threshold for outliers: Typically, the threshold for outliers is set to an absolute Z-score greater than 2 or 3. That is, any data point with an absolute Z-score greater than 2 or 3 is considered an outlier. This threshold can be adjusted according to specific circumstances.

[0059] Identify data points whose Z-score exceeds the above threshold to recognize outliers.

[0060] Replace outliers with the mean.

[0061] (3) Handling Missing Values: In a dataset, some records may be missing values ​​for one or more fields. Data cleaning involves deciding how to handle these missing values, such as by filling in missing values, deleting records containing missing values, or using statistical methods for estimation. In cases of missing data, linear interpolation is used to fill in the gaps. At two known points ( x a , y a )and( x b , y b Estimate between ) y The value of , for a given x The formula for linear interpolation is:

[0062]

[0063] This formula creates a path through the point ( x a , y a )and( x b , y b () a straight line, and use it to estimate x corresponding y value.

[0064] (4) Correct environmental disturbance factors for dynamic gravity signals, including correction of gravity parameters based on atmospheric pressure and ocean tides.

[0065] Atmospheric pressure correction: Changes in atmospheric pressure cause changes in gravity, therefore, the gravity parameters need to be corrected according to the atmospheric pressure at the time of observation. This correction is performed using the following formula:

[0066]

[0067] In the formula, g p To correct for gravity data after atmospheric pressure adjustment, p Atmospheric pressure at the time of observation p n Standard atmospheric pressure g 0 To correct gravity data after ocean tides.

[0068] Ocean tidal correction: The gravitational pull of the Sun and Moon causes ocean tides, which in turn cause deformation of the Earth and changes in the gravitational field. This is corrected using the loading theory. This must be done at the periphery. The correction is performed using the following formula:

[0069]

[0070]

[0071]

[0072] In the formula, and Here are the latitude and longitude of the observation point, and T is the gravity disturbance. r p This is the radial distance of the observation point from the Earth's center. G It is the gravitational constant. The density of seawater, h Let be a function of the height of ocean tidal waves. As a scaling factor, The spatial relationship between the observation point and the mass. r It is the Earth's average radius or the average distance from the observation point to the Earth's center of mass. h p It is the height of the observation point relative to the Earth's reference surface. The spherical distance angle between the two points is denoted as .

[0073] (5) The capture of dynamic gravity information requires multi-station and multi-instrument network joint monitoring, and the data captured by multiple instruments need to be aligned under the same timestamp.

[0074] (6) A high-pass Butterworth filter is a type of filter widely used in signal processing. Its main function is to allow high-frequency signals to pass through while suppressing signals below a specific cutoff frequency. It is a linear filter with a smooth frequency response curve, i.e., no ripples in the passband.

[0075] H(s) defines the behavior of the filter in the complex frequency domain (s-domain). For an nth-order high-pass Busterworth filter, the transfer function can be expressed as:

[0076]

[0077] in, s It is a complex frequency variable. a 1 , a 2 ,…, a n The coefficients are calculated based on the filter order and cutoff frequency.

[0078] Step 2: Refer to Figure 2 Multiple control variables are set, including abnormal warning threshold, abnormal noise threshold, abnormal duration window, continuous no-abnormal time window, number of warnings, and dispersion coefficient.

[0079] To identify anomalous wavebands (1–30 mHz) associated with seismic activity, a method and control variables for extracting signals related to earthquake precursors were developed after analyzing and comparing effective dynamic gravity data with historical earthquake data and through continuous experimentation and calibration.

[0080] Step 3: Calculate the first and second differences between the amplitude of the dynamic gravity signal and the anomaly warning threshold and the anomaly noise threshold, respectively. The first difference is the difference between the amplitude of the dynamic gravity signal and the anomaly warning threshold, and the second difference is the difference between the amplitude of the dynamic gravity signal and the anomaly noise threshold.

[0081] Step 4: When the first difference and the second difference are in the abnormal range, open the abnormal detection time window to detect the amplitude of the dynamic gravity signal. When the duration reaches the set value, count the number of times the amplitude of the dynamic gravity signal is in the abnormal range, and use the total number of counts as the number of warnings.

[0082] In this embodiment, the abnormal interval is defined as a first difference greater than 1.5 and a second difference less than 28.

[0083] When the first difference is greater than 1.5 and the second difference is less than 28, the abnormal duration window is defined as 120 minutes. If the duration reaches 120 minutes, the number of warnings is counted for the amplitude of the dynamic gravity signal.

[0084] The detection window is closed when the period of no abnormality (the duration of the value being less than the warning value) is 120 minutes.

[0085] Step 5: If the number of warnings exceeds the set number, calculate the coefficient of variation of the dynamic gravity signal amplitude in the abnormal range, and output the amplitude data with a coefficient of variation less than 1 as earthquake precursor information.

[0086] If the number of warnings exceeds 700 and the coefficient of variation is less than 1.0, it is identified as earthquake precursor information.

[0087] This data is compared with real-time earthquake data to ensure the system can respond quickly to any potential earthquake precursor signals. Furthermore, with technological advancements and the accumulation of new data, the anomaly warning threshold and anomaly noise threshold parameters are continuously verified, updated, and iterated to improve the accuracy and reliability of extracting earthquake precursor signals.

[0088] The obtained earthquake precursor information is subjected to time-domain and frequency-domain feature extraction to obtain time-domain and frequency-domain features, which are then used to analyze the earthquake precursor information.

[0089] Time-domain eigenvalues ​​are important indicators for measuring earthquake precursor information. They are generally divided into dimensional and dimensionless parameters. Dimensional eigenvalues ​​often have intuitive physical meanings and are the most commonly used indicator. These mainly include: Mean: the average of signal values, representing the DC component of the signal; Standard Deviation: the dispersion of the signal value distribution; Energy: the sum of the squares of the signal, reflecting the signal power; Maximum / Minimum: the extreme values ​​of the signal; Peak-to-Peak Value: the difference between the maximum and minimum values ​​of the signal; Root Mean Square (RMS): the square root of the mean of the squares of the signal, reflecting the effective value of the signal.

[0090] Dimensionless eigenvalues, including these parameters, are obtained by combining and normalizing the dimensional parameters of a signal. They describe the shape and distribution characteristics of a signal without containing specific physical units. They mainly include: Skewness: the asymmetry of the signal probability distribution; Kurtosis: the sharpness of the signal probability distribution; Waveform Factor: the ratio of the root mean square value to the mean absolute value, reflecting the steepness of the signal waveform; Crest Factor: the ratio of the peak value to the root mean square value, used to measure the sharpness of the signal peaks; Impulse Factor: the ratio of the peak value to the mean absolute value, describing the impact or prominence of the peak; Margin Factor: the ratio of the peak value to the mean amplitude, assessing the dynamic range of the signal; Shape Factor: the ratio of the root mean square value to the mean absolute value, describing the smoothness of the signal shape; and Clearance Factor: the ratio of the peak value to the mean of the signal cube at the square root.

[0091] After continuous wavelet transform, frequency domain feature engineering is performed to extract frequency domain feature values.

[0092] The purpose of this step is to convert earthquake precursor information to the frequency domain. Frequency domain analysis can reveal features in earthquake precursor information that are not easily observed in the time domain, and it also helps to identify and filter noise, extract the main frequency components of earthquake precursor information, and capture local time-frequency features, etc. Figure 3 , Figure 4 The mathematical expression for the continuous wavelet transform is:

[0093]

[0094] In the formula, This represents the result of continuous wavelet transform; a It is a scaling factor used to change the scale of the wavelet function, which is equivalent to the scaling of the signal. Different scales can capture different frequency features of the signal. b It is a translation factor used to shift the wavelet function along the time axis, thereby capturing the local characteristics of the signal; These are signals that need to be analyzed; It is a wavelet function, and the wavelet function selected here is the Morlet wavelet; It is the complex conjugate of the wavelet function; It is a normalization factor used to maintain energy consistency, which varies with scale. a It changes with the changes.

[0095] By changing the scale a Peaceful relocation bThe value of can be used to obtain the wavelet coefficients of the signal at different scales and locations, which form a two-dimensional coefficient matrix. Statistical features are extracted from the wavelet coefficients to characterize the signal's anomalous properties. The parameters are statistical measures of the wavelet coefficients, and the main frequency domain features include: wavelet energy distribution, wavelet entropy, wavelet coefficient statistical features, frequency band features, and wavelet packet analysis features.

[0096] The specific meanings are as follows: Total Energy: The sum of the energy of the wavelet transform coefficients of the signal; Subband Energy: The energy in a specific wavelet decomposition level or frequency band; Wavelet Energy Entropy: Measures the uniformity of the signal energy distribution; the more uniform the energy distribution, the greater the entropy; Wavelet Spectral Entropy: Related to the uniformity of frequency distribution, similar to energy entropy, but based on the spectrum; Wavelet Coefficient Mean: Reflects the average energy level at a certain scale or subband; Wavelet Coefficient Standard Deviation: Measures the fluctuation of wavelet coefficients around their mean; Wavelet Coefficient Skewness and Kurtosis: Describe the shape and kurtosis of the wavelet coefficient distribution; Center Frequency and Bandwidth: The dominant frequency and frequency range of each wavelet subband; Wavelet Packet Energy Distribution: The energy of each node or subband in wavelet packet decomposition; Wavelet Packet Entropy: Measures the uniformity of the energy distribution of wavelet packet coefficients.

[0097] This invention addresses the dynamic gravity signals associated with earthquake precursors, which have not received much attention from the academic community. It provides specific methods for eliminating environmental disturbance factors and extracting earthquake precursors, effectively acquiring dynamic gravity precursor information before earthquakes and providing research data for earthquake prediction.

[0098] Based on the same concept, the present invention also provides an earthquake precursor information extraction device, including an acquisition module, a setting module, a difference module, a detection module and a calculation module.

[0099] The acquisition module is used to continuously acquire dynamic gravity signals of the area to be measured.

[0100] The setting module is used to set multiple control variables, including abnormal warning threshold, abnormal noise threshold, abnormal duration window, number of warnings, and dispersion coefficient.

[0101] The difference module is used to obtain a first difference by subtracting the amplitude of the dynamic gravity signal from the abnormal warning threshold, and to obtain a second difference by subtracting the amplitude of the dynamic gravity signal from the abnormal noise threshold.

[0102] The detection module is used to open an abnormal detection time window to detect the amplitude of the dynamic gravity signal when the first difference and the second difference are in the abnormal range. When the duration reaches the set value, the module counts the number of times the amplitude of the dynamic gravity signal is in the abnormal range and uses the total number of counts as the number of warnings.

[0103] The calculation module is used to calculate the dispersion coefficient of the amplitude of the dynamic gravity signal in the abnormal range if the number of warnings exceeds the set number, and output the amplitude data with a dispersion coefficient of less than 1 as earthquake precursor information.

[0104] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the aforementioned method for extracting earthquake precursor information.

[0105] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned method for extracting earthquake precursor information.

[0106] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0107] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for extracting earthquake precursor information, characterized in that, Includes the following steps: Continuously acquire dynamic gravity signals of the area to be measured; Multiple control variables are set, including abnormal warning threshold, abnormal noise threshold, abnormal duration window, number of warnings, and dispersion coefficient; The first difference is obtained by subtracting the amplitude of the dynamic gravity signal from the abnormal warning threshold, and the second difference is obtained by subtracting the amplitude of the dynamic gravity signal from the abnormal noise threshold. When the first difference and the second difference are in the abnormal range, the abnormal detection time window is opened to detect the amplitude of the dynamic gravity signal. When the duration reaches the set value, the number of times the amplitude of the dynamic gravity signal is in the abnormal range is counted, and the total number of counts is used as the number of warnings. If the number of warnings exceeds the set number, the dispersion coefficient of the dynamic gravity signal amplitude in the abnormal range is calculated, and the amplitude data with a dispersion coefficient less than 1 is output as earthquake precursor information.

2. The method for extracting earthquake precursor information as described in claim 1, characterized in that, The control variable also includes a continuous no-abnormality time window. When the first difference and the second difference are not in the abnormal range, the continuous no-abnormality time window is opened for detection. If the duration reaches the set value, the abnormality detection time window is closed.

3. The method for extracting earthquake precursor information as described in claim 1, characterized in that, The abnormal interval is defined as a first difference greater than 1.5 and a second difference less than 28, the set value is 120 minutes, and the set number of times is 700.

4. The method for extracting earthquake precursor information as described in claim 1, characterized in that, Dynamic gravity signals were collected using an atmospheric tidal gravimeter, with an amplitude of 1–30 mHz.

5. The method for extracting earthquake precursor information as described in claim 1, characterized in that, Before obtaining the first difference value by subtracting the amplitude of the dynamic gravity signal from the anomaly warning threshold, the dynamic gravity signal needs to be preprocessed. The preprocessing process includes: Perform an integrity check on the dynamic gravity signal and remove any damaged signals; Noise values ​​were removed from the dynamic gravity signal after integrity check using the standard fraction method. Missing values ​​in the dynamic gravity signal after noise removal are supplemented using a linear interpolation method. Correct the dynamic gravity signal after missing values ​​are filled in; Data alignment is performed on the corrected dynamic gravity signal; The dynamic gravity signal after data alignment is filtered.

6. The method for extracting earthquake precursor information as described in claim 5, characterized in that, The correction of the dynamic gravity signal after the missing values ​​are filled includes the correction of gravity parameters based on atmospheric pressure and ocean tides; The correction of gravity parameters based on atmospheric pressure and ocean tides is shown in the following formula: g p -3(pp n )*10 -4 against in, μ = cosψ = ​​sinφ sinφ p +cosφcosφ p cos(λ-λ p ) In the formula, g p To correct for atmospheric pressure in the gravity data, where p is the atmospheric pressure at the time of observation. n The pressure is standard atmosphere, g0 is the gravity data after correction for ocean tides, and φ is... p and λ p Here are the latitude and longitude of the observation point, T is the gravity disturbance, and r is the longitude. p The radial distance of the observation point from the Earth's center, G is the gravitational constant, and ρ is the radial distance of the observation point from the Earth's center. w Let be the density of seawater, h be a function of the height of ocean tidal waves, β be a scaling factor, μ be the spatial relationship between the observation point and the mass, and r be the average radius of the Earth or the average distance from the observation point to the Earth's center of mass. p Ψ is the height of the observation point relative to the Earth's reference surface, and Ψ is the spherical distance angle between the two points.

7. The method for extracting earthquake precursor information as described in claim 1, characterized in that, It also includes extracting time-domain and frequency-domain features from earthquake precursor information to obtain time-domain and frequency-domain features, and then analyzing the earthquake precursor information using these time-domain and frequency-domain features.

8. A device for extracting earthquake precursor information, characterized in that, include: The acquisition module is used to continuously acquire dynamic gravity signals of the area to be measured; The setting module is used to set multiple control variables, including abnormal warning threshold, abnormal noise threshold, abnormal duration window, number of warnings, and dispersion coefficient; The difference module is used to obtain a first difference by subtracting the amplitude of the dynamic gravity signal from the abnormal warning threshold, and to obtain a second difference by subtracting the amplitude of the dynamic gravity signal from the abnormal noise threshold. The detection module is used to open an abnormal detection time window to detect the amplitude of the dynamic gravity signal when the first difference and the second difference are in the abnormal range. When the duration reaches a set value, the module counts the number of times the amplitude of the dynamic gravity signal is in the abnormal range and uses the total number of counts as the number of warnings. The calculation module is used to calculate the dispersion coefficient of the amplitude of the dynamic gravity signal in the abnormal range if the number of warnings exceeds the set number, and output the amplitude data with a dispersion coefficient of less than 1 as earthquake precursor information.

9. A computer device, characterized in that, The method includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the earthquake precursor information extraction method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the earthquake precursor information extraction method according to any one of claims 1-7.

Images