Micro-seismic event first arrival pickup method and system based on dynamic time warping

Abstract

The invention discloses a micro-seismic event first arrival pickup method and system based on dynamic time warping in the technical field of geophysical exploration, which are used for picking up seismic phase first arrival of a micro-seismic event according to a to-be-processed waveform, and comprises the following steps: step 1, detecting the to-be-processed waveform and carrying out denoising processing; 2, calibrating the time difference of the waveform to be processed; step 3, carrying out superposition processing on the waveform to be processed; 4, picking up the first arrival of the waveform to be processed; 5, comparing relative arrival time; 6, verifying the position of the detector; and 7, adjusting calculation parameters. The method can effectively improve the accuracy of microseism seismic phase first arrival pickup, reference oscillation waves are emitted through multiple oscillation generators at fixed positions, the position of a detector is verified based on a time difference of arrival positioning algorithm, to-be-processed waveforms are superposed after time difference calibration, first arrival pickup is performed on to-be-processed waveform superposition channels through a long-short time energy ratio method, and the accuracy of microseism seismic phase first arrival pickup is improved. Influences of individual inter-channel information errors are reduced, and the reliability of microseism seismic phase identification and first arrival pickup is improved.

Description

Technical Field

[0001] The present invention belongs to the field of geophysical exploration technology, and specifically is a method and system for picking the first arrival of microseismic events based on dynamic time regularization. Background Art

[0002] Microseismic monitoring technology is a widely studied geophysical technology and an important technical means to understand the distribution law of fracturing cracks and evaluate the fracturing effect. With the rapid development of information technology in recent years, this technology has been widely used in underground engineering, such as oil and gas field development, mine safety production, tunnel construction and geological disaster monitoring. The identification and first arrival picking of microseismic events are key steps in the processing of microseismic monitoring data. The accurate identification of microseismic events directly affects the accuracy of microseismic first arrival picking, and its error determines the accuracy of later source location. Therefore, how to improve the accuracy of microseismic event identification and the accuracy of first arrival picking is of key and important significance.

[0003] Traditional artificial microseismic phase identification and first arrival picking mainly rely on manual identification, which is extremely labor-intensive and easily affected by inconsistent picking standards and human experience, resulting in artificial errors. Therefore, methods for automatically identifying and picking microseismic event phases and first arrival picking have been proposed. Common first arrival picking methods can be divided into single-channel feature methods, multi-channel cross-correlation methods, and template matching methods. Single-channel feature methods include the long-short time average ratio method (STA / LTA method), AIC method, polarization analysis method, PAI-S / K method, and fractal dimension method. These single-channel feature algorithms pick up the first arrival of the phase based on the spectral characteristics of the single-channel waveform. They are insensitive to low signal-to-noise ratio events and it is difficult to reduce the impact of individual inter-channel information on the overall microseismic event picking results.

[0004] In actual production, multiple geophones that detect microseismic information may be affected by earthquakes and cause position changes. The accuracy of the geophone position directly affects the calculation parameters and relative time accuracy of the waveform to be processed, which in turn affects the accuracy of microseismic identification and first-arrival picking results.

[0005] Therefore, it is necessary to propose a microseismic event first-arrival picking method and system based on dynamic time warping, which can verify the spatial positioning of multiple detectors, and can perform superposition trace first-arrival picking of seismic phases on the processed waveform based on a single-channel feature algorithm with high computational efficiency, thereby reducing the influence of individual trace information errors and detector displacement errors. Summary of the invention

[0006] In order to solve the above problems, the purpose of the present invention is to provide a method and system for first-arrival picking of microseismic events based on dynamic time warping, in which the time difference of the waveform to be processed is calibrated and then superimposed, and the first arrival of the superimposed waveform to be processed is picked up by the long-short time energy ratio method, and the position change of the detector is compared by the relative time difference between each track, and a reference oscillation wave is emitted by a plurality of oscillation generators at fixed positions. The position of the detector is verified based on the arrival time difference positioning algorithm, and the relative spacing and position parameters of the detector are updated to reduce the influence of information errors between individual tracks, thereby improving the accuracy of microseismic phase identification and first arrival picking.

[0007] In order to achieve the above-mentioned purpose, the technical scheme of the present invention is as follows: a method for picking up the first arrival of a microseismic event based on dynamic time warping, which is used to pick up the first arrival of the seismic phase of a microseismic event according to a waveform to be processed, comprising the steps of: detecting the waveform to be processed by each detector provided in the detection area and performing denoising processing;

[0008] Step 2: Time difference calibration of the waveform to be processed: use the cross-correlation function to obtain the time difference between the channels of the waveform to be processed. Based on the relationship between the time difference between the channels and the relative arrival time, a linear equation group of the time difference between the channels and the relative arrival time of each channel is established and solved to calibrate the time difference of each channel.

[0009] Step 3: Superposition processing of the waveform to be processed, using the horizontal superposition method to perform superposition processing, according to the time delay theorem of Fourier transform, to obtain the amplitude spectrum function and phase spectrum function of the superposition channel;

[0010] Step 4: Pick the first arrival of the waveform to be processed. Use the long-short time energy ratio method to pick the first arrival of the waveform to be processed, and then use the relative arrival time t in step 3 to calculate the first arrival of the waveform. i Perform reverse time difference calibration to obtain the actual first arrival time T of each channel record i ;

[0011] Step 5: Relative arrival time comparison: compare the relative arrival time ti of each trace obtained in step 3 with the relative arrival time of each trace obtained by picking up the first arrival of the previous seismic phase, and mark the detector corresponding to the waveform with a relative arrival time error exceeding ±5% as displacement;

[0012] Step 6: Verify the position of the detector. Use an oscillator generator to send out reference waveforms at several fixed locations. After using the detector marked as displacement for detection, verify the position of the detector based on the arrival time difference positioning algorithm.

[0013] Step 7: Calculate the parameter adjustment, verify the new position of the detector, change the preset detector spacing and position parameters, re-perform the second step of phase first arrival picking, and optimize the calculation accuracy of the next phase first arrival picking. The principle of the basic scheme is: through the preprocessing of the waveform to be processed, the time difference calibration of dynamic time and the waveform superposition, the influence of the information error between individual channels is minimized, and the long-short time energy ratio method is used to simplify the calculation complexity, improve the accuracy and efficiency of the phase first arrival picking, and verify the spatial positioning of multiple detectors based on the arrival time difference positioning algorithm through the reference waveform emitted by the oscillation generator, update the detector position parameters, and ensure the accuracy of the subsequent phase first arrival picking parameters.

[0014] The beneficial effects of the basic scheme are: 1. Preprocessing the waveform to be processed, reducing the impact of noise on the waveform through denoising, making the waveform clearer, which is conducive to subsequent analysis and picking.

[0015] 2. In the time difference calibration process, the time difference between channels is calculated by using the cross-correlation function, and a linear equation group is established to solve it, thus realizing the accurate calibration of the arrival time of each channel. This reduces the error caused by the time difference between channels and improves the accuracy of waveform analysis.

[0016] 3. Waveform superposition processing: The influence of information errors between individual channels is reduced through the horizontal superposition method, the signal-to-noise ratio of the signal is enhanced, and the amplitude spectrum function and phase spectrum function of the superposition channel are obtained, providing more accurate information for subsequent first arrival picking.

[0017] 4. The application of the long-time and short-time energy ratio method simplifies the calculation complexity and improves the accuracy and efficiency of picking the first arrival of the seismic phase.

[0018] 5. The reference waveform sent by the oscillation generator and the arrival time difference positioning algorithm can accurately verify and calibrate the position of the detector. This ensures that the detector position parameters used in the subsequent phase first arrival picking process are accurate, thereby improving the accuracy of the phase first arrival picking results.

[0019] 6. This method is not only applicable to the first arrival picking of microseismic events, but can also be extended to the analysis of other types of seismic events. Its dynamic time warping characteristics enable this method to adapt to waveform data of different time scales and complexities, improving the versatility and practicality of the method.

[0020] Furthermore, the denoising of the waveform to be processed in step 1 includes filtering and denoising based on a frequency domain filter, performing Fourier transform and inverse transform on the waveform to be processed, and performing static correction and deconvolution denoising.

[0021] The beneficial effects of the basic scheme are: 1. The frequency domain filter can effectively remove high-frequency noise and interference in the waveform to be processed, convert the waveform from the spatial domain to the frequency domain through Fourier transform, then filter the specific frequency components, and finally restore it to the spatial domain through inverse Fourier transform. The effective signal components in the waveform can be retained while removing irrelevant background noise. This helps to enhance the characteristics of the waveform, making the first arrival picking of microseismic events more accurate and reliable.

[0022] 2. Static correction is the process of fine-tuning the position of waveform data, and deconvolution is a method used to remove multiple waves and interference waves in seismic data. After denoising, the noise components in the waveform data are effectively removed, which enables static correction and deconvolution algorithms to more accurately identify and adjust the position of waveform data, thereby further improving the stacking effect and the accuracy of data analysis.

[0023] Furthermore, the cross-correlation function of the signals between the two waveform channels to be processed in step 2 is defined as:

[0024]

[0025] In the formula, P is the number of sampling points, x i (p) and x j (p) is two data. When c i,j When (k) is the maximum, it is recorded as c i,j (max), at this time x i (p) and x j The waveform of (p) has the greatest similarity. i (p) and x j The time difference Δt of (p) i,j is regarded as the time difference between the first arrivals of the two paths;

[0026] Time difference between channels Δt i,j and the first arrival time t of the two records i and t j The relationship between them is:

[0027] t i -t j =Δt i，j .

[0028] Furthermore, the five linear equations of the time difference between the channels and the relative arrival time of each channel in step 2 are as follows:

[0029]

[0030] For M-channel records, the above formula needs to perform M(M-1) / 2 cross-correlation operations to obtain M(M-1) / 2 groups of time difference information between channels. The above formula can be simplified to the following form:

[0031] At=Δt

[0032] In the simplified formula, A is the sparse coefficient matrix, Δt is the measured data vector, and t is the relative arrival time vector to be determined. The least squares solution of the simplified formula is:

[0033] t=(A T A) -1 A T Δt.

[0034] The beneficial effects of the basic scheme are: 1. The time difference information between channels calculated by the cross-correlation function can accurately reflect the time difference of signals received by different detectors. By constructing and solving a set of linear equations, the arrival time of each channel signal can be accurately calibrated, thereby improving the accuracy of time synchronization.

[0035] 2. By converting the time difference calibration problem into a linear equation system, it can be solved using efficient numerical calculation methods (such as the least squares method). Compared with the traditional track-by-track comparison and calibration method, this greatly improves the speed and efficiency of data processing.

[0036] 3. Accurate time difference calibration is an important prerequisite for waveform superposition processing. By calibrating the arrival time of each signal, the temporal consistency of each signal during superposition can be ensured, thereby improving the waveform superposition effect and signal-to-noise ratio.

[0037] 4. After the time difference calibration, the arrival time of each signal is more accurate, which helps to more accurately identify the starting time point of the microseismic event in the subsequent first arrival picking step. Improving the accuracy of first arrival picking is of great significance for the location and parameter estimation of earthquake events.

[0038] 5. By constructing and solving a linear equation system, the correlation information between multiple channels of data can be fully utilized to improve the robustness and stability of the system. Even if some channel data is abnormal or missing, it can still be compensated and calibrated through the data of other channels to ensure the accuracy of the overall processing results.

[0039] 6. Accurate time difference calibration and first arrival picking are the basis for complex earthquake event analysis. This method can provide accurate and reliable data support for subsequent seismic wave propagation path analysis, source parameter estimation, etc.

[0040] Furthermore, the amplitude spectrum function of the superposition trace obtained in step 3 is as follows:

[0041]

[0042] Where S(ω) is the jth waveform to be processed after time difference calibration j The spectrum of (t), A j(ω) is the amplitude spectrum of the jth waveform to be processed.

[0043] Furthermore, the phase spectrum function of the superposition trace obtained in step 3 is as follows:

[0044]

[0045] Where φ(ω) is the jth waveform to be processed after time difference calibration j (t) phase spectrum, Φ j (ω) is the phase spectrum of the j-th waveform to be processed.

[0046] The beneficial effects of the basic scheme are: 1. The superposition channel amplitude spectrum function A(ω) effectively reduces the interference of random noise and improves the signal-to-noise ratio of the signal by averaging the amplitude spectrum of multiple waveforms. This helps to more accurately identify and analyze the characteristics of seismic signals in subsequent processing.

[0047] 2. The stacking phase spectrum function Φ(ω) retains the main phase characteristics of the waveform by averaging the phase spectrum of multiple waveforms. Phase information is an important part of seismic signal processing and is of great significance for the location and parameter estimation of seismic events. Through stacking processing, these phase characteristics can be further highlighted and the accuracy of analysis can be improved.

[0048] 3. The amplitude spectrum and phase spectrum functions after superposition processing provide clearer spectrum information, which helps to more accurately identify and analyze the components and characteristics of seismic signals in the frequency domain. This has a positive role in promoting subsequent processing steps such as classification, identification and parameter estimation of seismic events.

[0049] 4. Superposition processing is an efficient signal processing method. By averaging the data of multiple waveforms, it can significantly reduce the amount of calculation and complexity in subsequent processing steps. This helps to improve processing efficiency, shorten analysis time, and provide strong support for real-time monitoring and rapid response of earthquake events.

[0050] 5. Superposition processing can effectively reduce the impact of single waveform data anomalies or errors on the overall processing results by averaging the data of multiple waveforms. This enhances the stability and reliability of the method and improves the accuracy and credibility of the processing results.

[0051] 6. The amplitude spectrum function and phase spectrum function of the stacked trace provide important basic data for subsequent seismic signal processing and analysis. These data can be used in many aspects such as earthquake event location, parameter estimation, waveform classification and identification, and provide strong support for the construction of earthquake monitoring and early warning systems.

[0052] Furthermore, in step 4, the ratio of the short-time window kurtosis average value and the long-time window kurtosis average value of the waveform superposition channel to be processed is calculated, and the first arrival of the waveform superposition channel to be processed is picked according to the ratio. The calculation formula is as follows:

[0053]

[0054] In the formula, x(n) is the stacked trace, M is the long-time window length, and N is the short-time window length. The maximum value point of R is usually approximated as the first arrival point of the seismic wave.

[0055] After obtaining the first arrival time T0 of the superimposed trace, combine the relative arrival time t of each trace recorded in step 3 i , perform time difference calibration in reverse order to obtain the actual first arrival time T of each channel record i , the calculation formula is:

[0056] T i =T0+t i .

[0057] The beneficial effects of the basic scheme are: 1. Kurtosis is a statistic that reflects the shape of data distribution and is highly sensitive to noise and outliers. By calculating the kurtosis ratio, the influence of background noise and interference signals can be more effectively suppressed, and the accuracy of signal recognition can be improved.

[0058] 2. This method avoids the complicated waveform matching and feature extraction process, and directly performs the first arrival picking by calculating the kurtosis ratio, which greatly simplifies the processing flow and improves the processing efficiency.

[0059] 3. By calculating the ratio of the short-time window kurtosis average value and the long-time window kurtosis average value of the stacked trace, the first arrival point of the seismic wave can be identified more effectively, overcoming the disadvantage that the single-channel feature algorithm is greatly affected by the error between individual traces. This method can also overcome the errors and uncertainties existing in the traditional first arrival picking method and improve the accuracy and reliability of the first arrival picking.

[0060] Furthermore, in step six, the number of fixed locations of the oscillation generator is set to be at least 3.

[0061] The beneficial effects of the basic scheme are: 1. At least three oscillator generators at fixed locations can form a basic network for spatial positioning. By measuring the time difference between the reference waveforms emitted by these oscillator generators and reaching each detector, the arrival time difference positioning algorithm can be used to accurately calculate the position of the detector. In three-dimensional space, at least three non-collinear points are required to determine the position of a point. Therefore, at least three oscillator generators are the minimum requirement to ensure the accuracy of the spatial positioning of the detector.

[0062] 2. When at least three oscillator generators are used, if there is a problem with the data of one or two oscillator generators (such as failure, signal interference, etc.), the data of the remaining oscillator generators can still be used to locate the detector. This increases the redundancy of the positioning system and improves the reliability and stability of the system.

[0063] 3. Accurate detector position information is the basis for subsequent processing steps (such as time difference calibration, waveform superposition and first arrival picking). If there is an error in the detector position, it will directly affect the accuracy of these subsequent processing steps. Therefore, ensuring the accuracy of the detector position through at least three oscillator generators can indirectly improve the accuracy and efficiency of the entire processing flow.

[0064] 4. In complex terrain and geological conditions, the propagation path of seismic waves may be affected by many factors (such as terrain undulations, stratum changes, etc.). Using at least three oscillator generators can form a wider spatial coverage and reduce positioning errors caused by changes in terrain and geological conditions. This helps to ensure accurate detector location information in complex environments.

[0065] Furthermore, the arrival time difference positioning algorithm in step six calculates the relative position between the detector and each oscillator generator to perform spatial positioning of the detector based on the time difference between the time point when the detector detects the first peak of the reference waveform and the time point when the oscillator generator emits the first peak of the reference waveform.

[0066] The beneficial effects of the basic scheme are: 1. The algorithm does not rely on specific hardware or equipment layout, and can be positioned under different terrain and geological conditions as long as the time difference can be accurately measured. This makes the algorithm highly flexible and adaptable and can be applied to various complex environments.

[0067] 2. Compared with other positioning methods, such as GPS positioning or landmark-based positioning, this algorithm does not require additional equipment or complex operating procedures. It only needs to set up an oscillation generator at a fixed location and record the time point when the detector detects the peak to calculate and locate.

[0068] 3. Based on the measurement of time difference, the algorithm can perform data processing and positioning calculation in real time or near real time. This helps to quickly obtain the location information of the detector and provide timely data support for subsequent microseismic event analysis.

[0069] A microseismic event first arrival picking system based on dynamic time warping, comprising a plurality of geophones, a plurality of oscillation generators, a processing module and a display and reporting module;

[0070] The detector is used to detect the reference waveform and the waveform to be processed, and the number of detectors is at least 3;

[0071] The oscillation generator is used to simultaneously transmit reference oscillation waves at different fixed locations to verify the position of the detector;

[0072] The processing module is used to calculate the spatial positioning information of the detector according to the reference waveform detected by the detector and the waveform data to be processed, and perform the first arrival picking calculation on the waveform to be processed;

[0073] The display and reporting module is used to display data visualization results and report detector position changes based on the calculation results of the processing module.

[0074] The beneficial effects of the basic scheme are: 1. The system can accurately calculate the relative time difference between the waveforms to be processed by using the dynamic time warping algorithm through the processing module, so as to accurately determine the first arrival of the seismic phase of the waveform superposition trace to be processed. The time difference calibration effectively solves the alignment difference problem existing in the traditional first arrival picking method and significantly improves the accuracy of the first arrival picking.

[0075] 2. The system uses at least three detectors for detection, which can obtain more comprehensive waveform data and further improve the accuracy of first arrival picking. The collaborative work of multiple detectors can also realize cross-validation of waveform data and reduce errors and uncertainties.

[0076] 3. By transmitting reference oscillation waves at different fixed locations simultaneously, the oscillation generator can verify whether the position of the detector is accurate. This helps to ensure that the detector maintains the correct spatial layout during the monitoring process, thereby improving the reliability and stability of the entire system.

[0077] 4. The system realizes the automation of first arrival picking and can quickly complete the collection, processing and analysis of waveform data without manual intervention. This greatly reduces the labor intensity of the staff and improves work efficiency.

[0078] 5. The display report module can display the calculation results of the processing module in the form of data visualization and automatically generate detailed reports. This enables the staff to intuitively understand the monitoring results, find problems in time and take corresponding measures.

[0079] 6. The system has broad application prospects in the field of oil exploration. By accurately obtaining the first arrival time information of microseismic events, it can provide reliable data support for earthquake source location and earthquake source mechanism inversion, thereby helping oil companies to more accurately understand the distribution and characteristics of underground reservoirs and improve the success rate of exploration. In addition to the field of oil exploration, the system can also be applied to other related fields such as earthquake monitoring and geological disaster early warning. By real-time monitoring and analysis of seismic waveform data, timely and accurate earthquake early warning information can be provided to relevant departments, effectively reducing the losses caused by earthquake disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 Schematic diagram of a method for first-arrival picking of microseismic events based on dynamic time warping in an embodiment of the present invention.

[0081] Figure 2 Schematic diagram of the positional relationship between the detector and the oscillation generator in an embodiment of the present invention.

[0082] Figure 3 Schematic diagram of the waveform of the time difference calibration result of the microseismic event first arrival picking method based on dynamic time warping in an embodiment of the present invention.

[0083] Figure 4 Schematic diagram of waveform of the stacked trace seismic phase first arrival picking result of the microseismic event first arrival picking method based on dynamic time warping in an embodiment of the present invention.

[0084] Figure 5 Schematic diagram of waveforms of actual first arrival results of each trace of the microseismic event first arrival picking method based on dynamic time warping in an embodiment of the present invention.

[0085] Figure 6 Schematic diagram of a microseismic event first arrival picking system based on dynamic time warping in an embodiment of the present invention.

[0086] The figure marks in the drawings of the specification include: ★, oscillation generator position; ▼, detector position; ●, first arrival time point of the seismic phase. DETAILED DESCRIPTION

[0087] The following is further described in detail through specific implementation methods:

[0088] Example 1

[0089] Basically as attached Figure 1 , Figure 2 , Figure 3 , Figure 4 and Figure 5 As shown: A method for picking up the first arrival of a microseismic event based on dynamic time warping is used to pick up the first arrival of the seismic phase of a microseismic event according to a waveform to be processed, comprising step 1: detecting the waveform to be processed by each detector provided in the detection area and performing denoising processing, the denoising processing comprising filtering and denoising based on a frequency domain filter, performing Fourier transform and inverse transform on the waveform to be processed, and performing static correction and deconvolution denoising processing;

[0090] Step 2: Time difference calibration of the waveform to be processed. Use the cross-correlation function to obtain the time difference between the channels of the waveform to be processed. Based on the relationship between the time difference between the channels and the relative arrival time, a linear equation group of the time difference between the channels and the relative arrival time of each channel is established and solved to calibrate the time difference of each channel. The cross-correlation function of the channel signal of the waveform to be processed is defined as:

[0091]

[0092] In the formula, P is the number of sampling points, x i (p) and x j (p) is two data. When c i,j When (k) is the maximum, it is recorded as c i,j (max), at this time x i (p) and x j The waveform of (p) has the greatest similarity. i (p) and x j The time difference Δt of (p) i,j is regarded as the time difference between the first arrivals of the two paths;

[0093] Time difference between channels Δt i,j and the first arrival time t of the two records i and t j The relationship between them is:

[0094] t i -t j =Δt i，j ,

[0095] The linear equations of the time difference between the traces and the relative arrival time of each trace and the solution process are as follows, taking 5 traces as an example:

[0096]

[0097] For M-channel records, the above formula needs to perform M(M-1) / 2 cross-correlation operations to obtain M(M-1) / 2 groups of time difference information between channels. The above formula can be simplified to the following form:

[0098] At=Δt

[0099] In the simplified formula, A is the sparse coefficient matrix, Δt is the measured data vector, and t is the relative arrival time vector to be determined. The least squares solution of the simplified formula is:

[0100] t=(A T A) -1 A T Δt;

[0101] Step 3: Superposition processing of the waveform to be processed, using the horizontal superposition method for superposition processing, according to the delay theorem of Fourier transform, the amplitude spectrum function and phase spectrum function of the superposition channel are obtained as follows

[0102]

[0103] Where S(ω) is the jth waveform to be processed after time difference calibration j The spectrum of (t), A j (ω) is the amplitude spectrum of the jth waveform to be processed,

[0104]

[0105] Where φ(ω) is the jth waveform to be processed after time difference calibration j The phase spectrum of (t), Φ j (ω) is the phase spectrum of the j-th waveform to be processed;

[0106] Step 4: Pick the first arrival of the waveform to be processed. Use the long-short time energy ratio method to pick the first arrival of the waveform to be processed. Then perform reverse time difference calibration based on the relative arrival time ti in step 3 to obtain the actual first arrival time Ti of each record. The calculation formula is as follows:

[0107]

[0108] In the formula, x(n) is the stacked trace, M is the long-time window length, and N is the short-time window length. The maximum value point of R is usually approximated as the first arrival point of the seismic wave.

[0109] After obtaining the first arrival time T0 of the superimposed trace, combine the relative arrival time t of each trace recorded in step 3 i , perform time difference calibration in reverse order to obtain the actual first arrival time T of each channel record i , the calculation formula is:

[0110] T i =T0+t i .

[0111] Example 2

[0112] The difference from the above embodiment is that, as shown in the attached Figure 1 , Figure 2 As shown: The method for picking the first arrival of microseismic events based on dynamic time warping also includes step five: relative arrival time comparison, based on comparing the relative arrival time ti of each trace obtained in step three with the relative arrival time of each trace obtained by picking the first arrival of the previous seismic phase, marking the detector corresponding to the waveform with a relative arrival time error exceeding ±5% as displacement;

[0113] Step 6: Verify the detector position. Use an oscillator generator to send out a reference waveform at several fixed locations. After detection using a detector marked as displacement, verify the detector position based on the arrival time difference positioning algorithm. Set the number of fixed locations of the oscillator generator to at least 3. The arrival time difference positioning algorithm calculates the relative position between the detector and each oscillator generator to spatially locate the detector based on the time difference between the time point when the detector detects the first peak of the reference waveform and the time point when the oscillator generator sends out the first peak of the reference waveform.

[0114] Step 7: Calculate parameter adjustment. After verifying the new position of the detector, change the preset detector spacing and position parameters, re-perform the second step of phase first arrival picking, and optimize the calculation accuracy of the next phase first arrival picking.

[0115] Example 3

[0116] The difference from the above embodiment is that, as shown in the attached Figure 1 , Figure 2 and Figure 6 As shown: A microseismic event first arrival picking system based on dynamic time warping, including a plurality of geophones, a plurality of oscillation generators, a processing module and a display and reporting module;

[0117] The detector is used to detect the reference waveform and the waveform to be processed, and the number of detectors is at least 3;

[0118] The oscillation generator is used to simultaneously transmit reference oscillation waves at different fixed locations to verify the position of the detector;

[0119] The processing module is used to calculate the spatial positioning information of the detector according to the reference waveform detected by the detector and the waveform data to be processed, and perform the first arrival picking calculation on the waveform to be processed;

[0120] The display and reporting module is used to display data visualization results and report detector position changes based on the calculation results of the processing module.

[0121] The specific implementation process is as follows: In the process of picking the first arrival of microseismic phases, the microseismic oscillation is first detected by each detector and each waveform to be processed is output. The data of each waveform to be processed enters the processing module for preprocessing to remove noise, and then the processing module uses the cross-correlation function to obtain the time difference between the channels of the waveform to be processed, and establishes a linear equation group of the time difference between the channels and the relative arrival time of each channel. After solving the linear equation group, the arrival time of each channel is calibrated for time difference. After the time difference calibration, the waveforms to be processed are horizontally stacked to improve the signal-to-noise ratio, and the amplitude spectrum function and phase spectrum function of the stacked waveform to be processed are obtained as a reference for subsequent calculations. Finally, the single stacked waveform to be processed is calculated by the long-short time energy ratio method for the first arrival picking. The picked time is calibrated by the reverse time difference to obtain the first arrival time of each waveform to be measured, avoiding the influence of individual channel errors and improving the accuracy and reliability of the first arrival picking. The results after picking are reported visually to the staff through the display report module. The intuitive and clear results can deepen the staff's understanding of the location, parameter estimation, waveform classification and identification of microseismic events, and provide strong data support for the construction of earthquake monitoring and early warning systems. According to the comparison of the relative arrival time of each channel with the relative arrival time of each channel obtained by the first arrival of the previous seismic phase, the detector corresponding to the waveform with a relative arrival time error exceeding ±5% is marked as displacement, and the detector position verification is immediately started.

[0122] During the verification process of the detector position, the oscillation generators emit reference oscillation waves at their respective fixed positions. After the detector detects the reference waveform, the reference waveform data is transmitted to the processing module, including the reference waveform spectrum data and time data. The processing module uses the arrival time difference positioning algorithm to calculate the spatial position of the detector. If the actual position of the detector is different from the preset position, the data is reported to the staff through the display and reporting module, and the detector position parameters are automatically corrected and step two is restarted to optimize and calibrate the first arrival picking effect of the seismic phase.

[0123] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0124] The above is only an embodiment of the present invention. The common sense such as the known specific structure and characteristics in the scheme is not described in detail here. The ordinary technicians in the relevant field know all the common technical knowledge in the technical field of the invention before the application date or priority date, can obtain all the existing technologies in the field, and have the ability to apply the conventional experimental means before that date. The ordinary technicians in the relevant field can improve and implement this scheme in combination with their own abilities under the enlightenment given by this application. Some typical known structures or known methods should not become obstacles for ordinary technicians in the relevant field to implement this application. It should be pointed out that for those skilled in the art, without departing from the structure of the present invention, several deformations and improvements can be made, which should also be regarded as the protection scope of the present invention, which will not affect the effect of the implementation of the present invention and the practicality of the patent. The protection scope required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A method for picking the first arrival of microseismic events based on dynamic time warping, which is used to pick the first arrival of the seismic phase of microseismic events according to the waveform to be processed, characterized in that: include: Step 1: Detect the waveform to be processed through each detector provided in the detection area and perform denoising; Step 2: Time difference calibration of the waveform to be processed: use the cross-correlation function to obtain the time difference between the channels of the waveform to be processed. Based on the relationship between the time difference between the channels and the relative arrival time, a linear equation group of the time difference between the channels and the relative arrival time of each channel is established and solved to calibrate the time difference of each channel. Step 3: Superposition processing of the waveform to be processed, using the horizontal superposition method to perform superposition processing, according to the time delay theorem of Fourier transform, to obtain the amplitude spectrum function and phase spectrum function of the superposition channel; Step 4: Pick the first arrival of the waveform to be processed. Use the long-short time energy ratio method to pick the first arrival of the waveform to be processed, and then use the relative arrival time t in step 3 to calculate the first arrival of the waveform. i Perform reverse time difference calibration to obtain the actual first arrival time T of each channel record i ; Step 5: Relative arrival time comparison: compare the relative arrival time ti of each trace obtained in step 3 with the relative arrival time of each trace obtained by picking up the first arrival of the previous seismic phase, and mark the detector corresponding to the waveform with a relative arrival time error exceeding ±5% as displacement; Step 6: Verify the position of the detector. Use an oscillator generator to send out reference waveforms at several fixed locations. After using the detector marked as displacement for detection, verify the position of the detector based on the arrival time difference positioning algorithm. Step 7: Calculate parameter adjustment. After verifying the new position of the detector, change the preset detector spacing and position parameters, re-perform the second step of phase first arrival picking, and optimize the calculation accuracy of the next phase first arrival picking.

2. The method for first arrival picking of microseismic events based on dynamic time warping according to claim 1, characterized in that: The denoising process of the waveform to be processed in step 1 includes filtering and denoising based on a frequency domain filter, performing Fourier transform and inverse transform on the waveform to be processed, and performing static correction and deconvolution denoising processes.

3. The method for first arrival picking of microseismic events based on dynamic time warping according to claim 2, characterized in that: The cross-correlation function of the signals between the two waveform channels to be processed in step 2 is defined as: In the formula, P is the number of sampling points, x i (p) and x j (p) is two data. When c i,j When (k) is the maximum, it is recorded as c i,j (max), at this time x i (p) and x j The waveform of (p) has the greatest similarity. i (p) and x j The time difference Δt of (p) i,j It is regarded as the time difference between the first arrival of the two paths; Time difference between channels Δt i,j and the first arrival time t of the two records i and t j The relationship between them is: t i -t j =Δt i，j 。 4. The method for first arrival picking of microseismic events based on dynamic time warping according to claim 3 is characterized in that: The five linear equations of the time difference between channels and the relative arrival time of each channel in step 2 are as follows: For M-channel records, the above formula needs to perform M(M-1) / 2 cross-correlation operations to obtain M(M-1) / 2 groups of time difference information between channels. The above formula can be simplified to the following form: At=Δt In the simplified formula, A is the sparse coefficient matrix, Δt is the measured data vector, and t is the relative arrival time vector to be calculated. The least squares solution of the simplified formula is: t=(A T A) -1 A T Δt。 5. The method for first arrival picking of microseismic events based on dynamic time warping according to claim 4 is characterized in that: The amplitude spectrum function of the superposition trace obtained in step 3 is as follows: Where S(ω) is the jth waveform to be processed after time difference calibration j The spectrum of (t), A j (ω) is the amplitude spectrum of the jth waveform to be processed.

6. The method for first arrival picking of microseismic events based on dynamic time warping according to claim 5, characterized in that: The phase spectrum function of the superposition trace obtained in step 3 is as follows: Where φ(ω) is the jth waveform to be processed after time difference calibration j (t) phase spectrum, Φ j (ω) is the phase spectrum of the j-th waveform to be processed.

7. The method for first arrival picking of microseismic events based on dynamic time warping according to claim 6, characterized in that: In step 4, the ratio of the short-time window kurtosis average value and the long-time window kurtosis average value of the waveform superposition channel to be processed is calculated, and the first arrival of the waveform superposition channel to be processed is picked according to the ratio. The calculation formula is as follows: In the formula, x(n) is the stacked trace, M is the long-time window length, and N is the short-time window length. The maximum value point of R is usually approximated as the first arrival point of the seismic wave. After obtaining the first arrival time T0 of the superimposed trace, combine the relative arrival time t of each trace recorded in step 3 i , perform time difference calibration in reverse order to obtain the actual first arrival time T of each channel record i , the calculation formula is: T i =T0+t i 。 8. The method for first arrival picking of microseismic events based on dynamic time warping according to claim 7, characterized in that: In step 6, the number of fixed locations of the oscillation generator is set to at least 3.

9. The method for first arrival picking of microseismic events based on dynamic time warping according to claim 8, characterized in that: The arrival time difference positioning algorithm in step six calculates the relative position between the detector and each oscillator generator to spatially locate the detector based on the time difference between the time point when the detector detects the first peak of the reference waveform and the time point when the oscillator generator emits the first peak of the reference waveform.

10. A microseismic event first arrival picking system based on dynamic time warping, characterized in that: It includes several detectors, several oscillation generators, a processing module and a display and reporting module; The detector is used to detect the reference waveform and the waveform to be processed, and the number of detectors is at least 3; The oscillation generator is used to simultaneously transmit reference oscillation waves at different fixed locations to verify the position of the detector; The processing module is used to calculate the spatial positioning information of the detector according to the reference waveform detected by the detector and the waveform data to be processed, and perform the first arrival picking calculation on the waveform to be processed; The display and reporting module is used to display data visualization results and report detector position changes based on the calculation results of the processing module.

Images