Microseismic event p-wave first arrival picking method and device
By using the Gaussian curve iterative fitting method, the problems of strong subjectivity and large noise interference in the traditional P-wave first arrival picking method are solved, realizing the automated and accurate picking of the P-wave first arrival time in microseismic signals, which is suitable for complex waveform data analysis.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA PETROLEUM & CHEMICAL CORP
- Filing Date
- 2025-07-28
- Publication Date
- 2026-07-24
AI Technical Summary
Traditional P-wave first arrival acquisition methods suffer from high subjectivity and significant noise interference, leading to decreased acquisition accuracy.
The Gaussian curve iterative fitting method is adopted. By selecting a time window within a preset range of the P-wave arrival time, an initial Gaussian model is established and iterative fitting is performed to determine the P-wave arrival time.
It enables automated and accurate acquisition of the first arrival time of P-waves in microseismic signals, has strong noise resistance, is suitable for complex waveform data analysis, and improves acquisition accuracy and consistency.
Smart Images
Figure CN120871243B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of seismic waveform data processing technology, and specifically relates to a method and apparatus for picking up the first arrival of P-waves in microseismic events. Background Technology
[0002] Microseismic monitoring technology has been widely applied in mining, geotechnical engineering, and oil and gas extraction. Accurately acquiring the P-wave first arrival time of microseismic events is crucial for source location, geological structure analysis, and rock mass stability assessment. However, traditional P-wave arrival time acquisition methods suffer from the following problems due to the frequent noise interference accompanying microseismic signals: High subjectivity: The interpretation of the P-wave first arrival time is heavily influenced by the operator's experience, making consistency difficult to guarantee. Random noise affects the identification of waveform characteristics, leading to decreased acquisition accuracy.
[0003] For P-waves, the choice of arrival time (especially the starting point of the initial peak) is often subjective because the shape of the peak typically approaches the baseline rather than intersects it. This problem can be addressed using a simple threshold detection method, where the arrival time is determined by identifying the first arrival of the generated waveform—the time when the amplitude reaches a small fraction of the peak amplitude. However, this method has a drawback: random noise present in the baseline often accounts for a large proportion of the amplitude at specific points, affecting the accuracy of the measurement and rendering it unusable. Summary of the Invention
[0004] This application provides a method and apparatus for picking up the first arrival of P-waves in microseismic events, which improves the analysis accuracy and reliability of microseismic monitoring data.
[0005] In a first aspect, embodiments of this application provide a method for picking up the first arrival of P-waves in microseismic events, including:
[0006] Acquire microseismic monitoring data and select a time window containing the peak within a preset range of P-wave arrival; establish an initial Gaussian model; input the selected time window data into the initial Gaussian model and perform iterative fitting; when the iteration meets the preset conditions, obtain the final Gaussian model curve; determine the P-wave arrival time based on the final Gaussian model curve.
[0007] In some embodiments, acquiring microseismic monitoring data and selecting a time window containing the peak within a preset range of P-wave arrival includes: acquiring microseismic monitoring data and preprocessing it; performing bandpass filtering on the original microseismic signal and then smoothing it using a moving average method; and selecting a time window containing the peak within a preset range of P-wave arrival as the region for fitting analysis.
[0008] In some embodiments, establishing an initial Gaussian model includes:
[0009] Define a single Gaussian model:
[0010]
[0011] Where A is the amplitude, η is the center position, and σ is the standard deviation;
[0012] The model is defined as a superposition of several Gaussian functions, and its mathematical expression is:
[0013]
[0014] Among them, A i Let η be the amplitude of the i-th Gaussian function, and let η represent the height of the wave crest. i The center position represents the mean or expected value, σ. i The standard deviation controls the width of the curve.
[0015] In some embodiments, the selected time window data is input into the initial Gaussian model for iterative fitting. When the iteration meets the preset conditions, the final Gaussian model curve is obtained, including:
[0016] Load the preprocessed data y(t), where y(t) represents the signal amplitude at time t;
[0017] The initial parameters are estimated, and the set of parameter estimates is set as Θ. (1) ={A i (1) ,η i (1) ,σ i (1)}, set the number of iterations, where the meaning of each parameter in the first iteration is, A i (1) Let η be the initial amplitude of the i-th Gaussian function. i (1) Let σ be the initial center position of the i-th Gaussian function. i (1) Let be the initial standard deviation of the i-th Gaussian function, where i ranges from 1 to n, representing each Gaussian function;
[0018] Based on the current parameters, calculate the signal value fitted by the model, which is then compared with the actual signal data. The model output is... k is the number of iterations;
[0019] Based on the model output y mo (k) Compare y(t) with the actual data y(t) and calculate the error;
[0020]
[0021] E (k)Let be the error function value of the k-th iteration, representing the total deviation between the model output and the actual data, M be the total number of sampling points within the selection window, and t be the error function value of the k-th iteration. j For the j-th sampling point, y(t) j ) represents the actual signal at time t j The amplitude, Fit the output of the model for the k-th iteration at time t j The amplitude;
[0022] The objective function is to minimize E (k) :
[0023] Θ (k+1) =Θ (k) +ΔΘ (k)
[0024] Θ (k+1) It is the parameter set for the (k+1)th iteration, ΔΘ (k) It represents the update amount of the parameters in the k-th iteration;
[0025] Determine if the objective function converges. If the convergence condition is met, stop the iteration. The change in the error function is:
[0026] ΔE (k) =|E (k+1) -E (k) |
[0027] If ΔE (k) If the error threshold is less than ε, then convergence is considered achieved, and iteration is stopped. ε is a preset error threshold.
[0028] In some embodiments, the selected time window data is input into the initial Gaussian model for iterative fitting. When the iteration meets a preset condition, the final Gaussian model curve is obtained, including: if the number of iterations reaches the maximum allowed number k. max The iteration process also stops.
[0029] In some embodiments, obtaining the final Gaussian model curve, and determining the P-wave arrival time based on the final Gaussian model curve, includes:
[0030] The parameter set at convergence is used as the optimal parameters to calculate the final model set curve. The first arrival time of the P wave is then determined using the fitted model curve.
[0031] In some embodiments, obtaining the final Gaussian model curve, and determining the P-wave arrival time based on the final Gaussian model curve, includes:
[0032] Using the set of parameters at convergence as the optimal parameters, the final Gaussian model curve is obtained, and the maximum amplitude A of the final Gaussian model curve is calculated. max Set the threshold to A threshold =bAmax b is a preset percentage. Starting from the beginning of the time axis, the algorithm searches for the first time when the final Gaussian model curve reaches or exceeds the threshold A. threshold The time point t0 is the arrival time of the P wave.
[0033] Secondly, this application provides a device for picking up the first arrival of P-waves in microseismic events, comprising:
[0034] Select a unit to acquire microseismic monitoring data, and select a time window containing the peak within the preset range of P-wave arrival;
[0035] Establish a unit to build the initial Gaussian model;
[0036] The determination unit is used to input the selected time window data into the initial Gaussian model for iterative fitting. When the iteration meets the preset conditions, the final Gaussian model curve is obtained. Based on the final Gaussian model curve, the first arrival time of the P wave is determined.
[0037] Thirdly, embodiments of this application provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.
[0038] Fourthly, embodiments of this application provide a computer device, including 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 steps of any of the methods described above.
[0039] The microseismic event P-wave first arrival pickup method and device of the present application have the following beneficial effects:
[0040] This application provides a high-precision method for picking up the first arrival time of P-waves in microseismic events based on Gaussian curve iterative fitting. It realizes the automated and accurate picking of the first arrival time of P-waves in microseismic signals, has strong noise resistance, and is suitable for the analysis of complex waveform data. Attached Figure Description
[0041] Figure 1 This is a schematic diagram of the P-wave first arrival pickup method for microseismic events according to an embodiment of this application;
[0042] Figure 2 This is another flowchart illustrating the method for picking up the first arrival of P-waves in microseismic events according to an embodiment of this application.
[0043] Figure 3 This is a schematic diagram illustrating how the method described in this application is used to determine the specific first arrival time of the P-wave based on the amplitude jump in a three-component waveform monitored by a microseismic observation system.
[0044] Figure 4This is a schematic diagram of the structure of the microseismic event P-wave first arrival pickup device according to an embodiment of this application. Detailed Implementation
[0045] The present application will be further described below with reference to the accompanying drawings and embodiments.
[0046] In the following description, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance. The following description provides multiple embodiments of the invention, which can be substituted or combined with each other. Therefore, this application can also be considered to include all possible combinations of the same and / or different embodiments described. Thus, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then this application should also be considered to include embodiments containing one or more other possible combinations of features A, B, C, and D, even if such embodiments are not explicitly described in the following text.
[0047] Example 1
[0048] like Figure 1 As shown, the method for picking the first arrival of P-waves in microseismic events in this application includes: S101, acquiring microseismic monitoring data and selecting a time window containing the wave crest within a preset range of the first arrival of P-waves; S103, establishing an initial Gaussian model; S105, inputting the selected time window data into the initial Gaussian model and performing iterative fitting. When the iteration meets the preset conditions, the final Gaussian model curve is obtained, and the first arrival time of P-waves is determined based on the final Gaussian model curve.
[0049] This application enables automated and accurate acquisition of the first arrival time of P-waves in microseismic signals, exhibits strong noise resistance, and is suitable for the analysis of complex waveform data.
[0050] Example 2
[0051] like Figure 1 , Figure 2 and Figure 3 As shown, the P-wave first arrival picking method for microseismic events in this application, from the three-component raw data of microseismic monitoring to the final determination of the P-wave first arrival time, includes the following steps:
[0052] Step 1: Acquire microseismic monitoring data and perform data preprocessing: Bandpass filtering is applied to the raw microseismic signal to remove high-frequency and low-frequency noise and retain the main frequency components of the P-wave; then, the moving average method is used to smooth the filtered signal to reduce the influence of random noise.
[0053] Step 2, Select the analysis window: Through the user interface, select a time window containing the main peaks near the first arrival of the P wave (within the preset range) as the area for fitting analysis.
[0054] Step 3, establish the Gaussian model: first set up a single Gaussian model; then set up a superposition model of multiple Gaussian functions to characterize the distribution of complex peaks; give the initial amplitude, center position and standard deviation of each Gaussian function.
[0055] Define a single Gaussian model as shown in formula (1):
[0056]
[0057] Where A is the amplitude, η is the center position, and σ is the standard deviation;
[0058] The model is set as a superposition of several Gaussian functions, and the mathematical expression is formula (2):
[0059]
[0060] Among them, A i Let η be the amplitude of the i-th Gaussian function, and let η represent the height of the wave crest. i The center position represents the mean or expected value, σ. i The standard deviation controls the width of the curve.
[0061] Step 4, Iterative Fitting and Parameter Update: Load the preprocessed data, generate model output based on the initial parameter estimate; calculate the error between the model output and the actual data, and give the error function; minimize the error as the objective function and update the parameters; determine whether it has converged or reached the maximum number of iterations.
[0062] By iteratively fitting a Gaussian model, we find an idealized model that best approximates the actual data distribution.
[0063] The algorithm for iterative fitting first loads the preprocessed data y(t), where y(t) represents the signal amplitude at time t. Then, it estimates the initial parameters, setting the set of parameter estimates as Θ. (1) ={A i (1) ,η i (1) ,σ i (1) Let the iteration number k = 1, where the meaning of each parameter in the first iteration is: A i (1) η is the initial amplitude of the i-th Gaussian function. i (1) It is the initial center position of the i-th Gaussian function, σ i (1) It is the initial standard deviation of the i-th Gaussian function, where i ranges from 1 to n, representing each Gaussian function.
[0064] Based on the current parameters, the signal value fitted by the model is calculated and compared with the actual signal data. According to formula (2), the model output is... In parameter Θ (k) Under the given information, the k-th iteration, is the model's fitting result to the signal.
[0065] Based on the model output y mo (k) Compare y(t) with the actual data y(t) to calculate the error.
[0066] The purpose of calculating the error is to quantify the difference between the model output and the actual data, and to provide a basis for parameter optimization, as shown in formula (3):
[0067]
[0068] E (k) The error function value at the k-th iteration represents the total deviation between the model output and the actual data. M is the total number of sampling points within the selected window. t j It is the j-th sampling point, y(t) j ) is the actual signal at time t j The amplitude, The output of the model fitting in the kth iteration at time t j The amplitude.
[0069] The objective function is to minimize E (k) To improve fitting accuracy, update the parameters, as shown in formula (4):
[0070] Θ (k+1) =Θ (k) +ΔΘ (k) (4)
[0071] Θ (k+1) It is the parameter set for the (k+1)th iteration, ΔΘ (k) It represents the parameter update amount in the k-th iteration.
[0072] Determine whether the objective function has converged. If the convergence condition is met, stop the iteration. The change in the error function is expressed as in formula (5):
[0073] ΔE (k) =|E (k+1) -E (k) | (5)
[0074] If ΔE (k) If the error threshold is less than ε, then convergence is considered achieved, and iteration stops. ε is a preset error threshold, which is set to 10 in this application. -6 If the number of iterations reaches the maximum allowed number k max This also stops the iteration process to prevent infinite loops; in this application, it is set to 1000.
[0075] The optimal parameters and fitting results are output for subsequent P-wave first arrival time determination. Using the convergent parameter set as the optimal parameters, the final model set curve is calculated. The fitted model curve is then used to accurately determine the P-wave first arrival time.
[0076] By iteratively fitting a Gaussian model, an idealized model that best approximates the actual data distribution is found. Obtaining this best-fit model results in a "clean" data representation because the model smooths out the influence of random noise, preserves the essential characteristics of the signal, and unifies scattered data points into a single mathematical expression. Furthermore, model-based calculations mean that instead of directly searching for the starting point from the original data, this idealized mathematical model is first established, and then the starting point is determined from this "clean" model. This ensures that the resulting starting point is not affected by noise in the original data. This iterative curve fitting method provides a new approach for analyzing complex peaks and minimizing the impact of noise in picking up the first arrival of P-waves.
[0077] Step 5, Output the optimal model and determine the first arrival of the P wave: When the iteration meets the convergence condition (or reaches the maximum number of iterations), output the optimal parameters; obtain the final Gaussian model curve and calculate its maximum amplitude; set a threshold as a certain percentage of the maximum amplitude, and search along the time axis for the time point when the model curve first reaches the threshold, which is determined as the first arrival time of the P wave.
[0078] Initial guesses are made for the nonlinear parameters (position and width of overlapping peaks) based on the selected analysis region. For each Gaussian function, the parameter A... i η i σ i Initial estimates are made. These parameters are continuously adjusted until the desired fitting accuracy or the maximum number of iterations is achieved.
[0079] An iterative curve fitting nonlinear least squares optimization algorithm is used to iteratively fit a Gaussian model, setting a maximum number of iterations and adjusting parameter A. i η i σ i This gradually reduces the error between the model curve and the actual data. The iterative process continues until the convergence condition is met, the error is less than the threshold or the maximum number of iterations is reached, and the reason for stopping the iteration is output.
[0080] Determine the first arrival time of the P-wave: Calculate the total maximum amplitude A of the fitted model. max Set the threshold to A threshold =0.01×A max Starting from the beginning of the timeline, find the first time the model curve reaches or exceeds A. threshold The time point t0 is the arrival time of the P wave.
[0081] This application has the following effects:
[0082] Improved picking accuracy: Iterative Gaussian curve fitting effectively suppresses noise interference, making the picking of P-wave first arrival times more accurate. High degree of automation: This method requires no manual intervention, automatically completing parameter optimization and arrival time determination, improving work efficiency. Adaptability to complex signals: It can handle overlapping peaks and has advantages in data analysis of complex waveforms. Unified judgment criteria: Using 1% of the model peak value as a threshold avoids the subjectivity of manually setting thresholds and ensures the consistency of results.
[0083] like Figure 4 As shown, this application provides a microseismic event P-wave first arrival picking device, including: a selection unit 201, used to acquire microseismic monitoring data and select a time window containing the wave peak within a preset range of P-wave first arrival; an establishment unit 202, used to establish an initial Gaussian model; and a determination unit 203, used to input the selected time window data into the initial Gaussian model, perform iterative fitting, and obtain the final Gaussian model curve when the iteration meets the preset conditions, and determine the P-wave first arrival time based on the final Gaussian model curve.
[0084] In this application, the embodiments of the microseismic event P-wave first arrival pickup device and the microseismic event P-wave first arrival pickup method are basically similar. For relevant details, please refer to the description of the microseismic event P-wave first arrival pickup method.
[0085] This application also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes any of the above-described methods for picking up the first arrival of a P-wave in a microseismic event.
[0086] This application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of the above-described method for picking up the first arrival of a P-wave in a microseismic event. The computer-readable storage medium may include, but is not limited to, any type of disk, including floppy disks, optical disks, DVDs, CD-ROMs, microdrives, magneto-optical disks, ROMs, RAMs, EPROMs, EEPROMs, DRAMs, VRAMs, flash memory devices, magnetic cards or optical cards, nanosystems (including molecular memory ICs), or any type of medium or device suitable for storing instructions and / or data.
[0087] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
Claims
1. A method for picking up the first arrival of P-waves in microseismic events, characterized in that, include: Acquire microseismic monitoring data, select a time window containing the peak within a preset range of P-wave arrival; establish an initial Gaussian model; The selected time window data is input into the initial Gaussian model for iterative fitting. When the iteration meets the preset conditions, the final Gaussian model curve is obtained. Based on the final Gaussian model curve, the first arrival time of the P wave is determined. The initial Gaussian model includes: Define a single Gaussian model: in, For amplitude, As the central location, Standard deviation; The model is defined as a superposition of several Gaussian functions, and its mathematical expression is: in, Let be the amplitude of the i-th Gaussian function, and represent the height of the wave crest. The center position represents the mean or expected value. The standard deviation is used to control the width of the curve; The selected time window data is input into the initial Gaussian model for iterative fitting. When the iteration meets the preset conditions, the final Gaussian model curve is obtained, including: Load preprocessed data , This represents the signal amplitude at time t; Estimate the initial parameters, and set the set of parameter estimates as follows: Set the number of iterations, and define the meaning of each parameter in the first iteration. Let be the initial amplitude of the i-th Gaussian function. Let be the initial center position of the i-th Gaussian function. Let be the initial standard deviation of the i-th Gaussian function, where i ranges from 1 to n, representing each Gaussian function; Based on the current parameters, calculate the signal value fitted by the model, which is then compared with the actual signal data. The model output is... k is the number of iterations; Based on the model output Compared with actual data Compare and calculate the error; Let be the error function value of the k-th iteration, representing the total deviation between the model output and the actual data. To select the total number of data sampling points within the selected window, For the j-th sampling point, For the actual signal in time The amplitude, Fit the output of the model for the k-th iteration in time The amplitude; The objective function is to minimize : It is the parameter set for the (k+1)th iteration. It represents the update amount of the parameters in the k-th iteration; Determine if the objective function converges. If the convergence condition is met, stop the iteration. The change in the error function is: if If convergence is achieved, the iteration is stopped. This is the preset error threshold.
2. The method for picking up the first arrival of P-waves in microseismic events according to claim 1, characterized in that, Acquiring microseismic monitoring data and selecting a time window containing the peak within a preset range of P-wave arrival includes: acquiring and preprocessing microseismic monitoring data, performing bandpass filtering on the original microseismic signal, and then smoothing it using the moving average method; selecting a time window containing the peak within a preset range of P-wave arrival as the region for fitting analysis.
3. The method for picking up the first arrival of P-waves in microseismic events according to claim 1, characterized in that, The selected time window data is input into the initial Gaussian model for iterative fitting. When the iterations meet the preset conditions, the final Gaussian model curve is obtained, including: if the number of iterations reaches the maximum allowed number k. max The iteration process also stops.
4. The method for picking up the first arrival of P-waves in microseismic events according to claim 1, characterized in that, The final Gaussian model curve is obtained. Based on the final Gaussian model curve, the P-wave first arrival time is determined, including: The parameter set at convergence is used as the optimal parameters to calculate the final model set curve. The first arrival time of the P wave is then determined using the fitted model curve.
5. The method for picking up the first arrival of P-waves in microseismic events according to claim 1, characterized in that, The final Gaussian model curve is obtained. Based on the final Gaussian model curve, the P-wave first arrival time is determined, including: Using the set of parameters at convergence as the optimal parameters, the final Gaussian model curve is obtained, and the maximum amplitude A of the final Gaussian model curve is calculated. max Set the threshold to A threshold =bA max b is a preset percentage. Starting from the beginning of the time axis, the algorithm searches for the first time when the final Gaussian model curve reaches or exceeds the threshold A. threshold time point that time point This is the arrival time of the P wave.
6. A device for picking up the first arrival of P-waves in microseismic events, characterized in that, The apparatus is used to perform the method as described in any one of claims 1-5, the apparatus comprising: Select a unit to acquire microseismic monitoring data, and select a time window containing the peak within the preset range of P-wave arrival; Establish a unit to build the initial Gaussian model; The determination unit is used to input the selected time window data into the initial Gaussian model for iterative fitting. When the iteration meets the preset conditions, the final Gaussian model curve is obtained. Based on the final Gaussian model curve, the first arrival time of the P wave is determined.
7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method described in any one of claims 1-5.
8. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1-5.