A microseismic event adaptive inversion method based on hierarchical optimization

Abstract

The application relates to a microseismic event adaptive inversion method based on hierarchical optimization. First, a two-dimensional variational mode decomposition is used to extract a mode more conducive to first arrival picking and reconstruct a signal, so that the signal-to-noise ratio of the first arrival picking signal is improved. A Lagrange multiplier method is used to construct a hierarchical optimization inner optimization function based on a simulated first arrival time of a channel, a corrected first arrival time of the channel is used to construct an error weighted coefficient constrained source location objective function, and an inner and outer optimization algorithm is used to jointly iteratively optimize the process of simulating artificial correction of microseismic first arrival. An adaptive differential evolution optimization algorithm is used to adaptively adjust a mutation factor and a crossover rate, so that the global search capability of the outer optimization is improved.

Description

Technical Field

[0001] This application relates to the field of microseismic data processing, and more specifically, to an adaptive inversion method for microseismic events based on hierarchical optimization. Background Technology

[0002] In microseismic event localization, travel time-based methods are a classic and highly reliable approach. However, this method requires a high degree of accuracy in acquiring the first arrival travel time of microseismic events. Currently, while most microseismic events in microseismic monitoring projects are automatically acquired, manual adjustments to the acquired time are still necessary before localization when the event signal-to-noise ratio (SNR) is low, or when the SNR is low on certain channels. This manual adjustment of the first arrival time introduces more workload, and since manual correction lags behind real-time microseismic event localization, it fails to meet the requirements of real-time microseismic monitoring. Summary of the Invention

[0003] To overcome at least one deficiency in the prior art, this application provides an adaptive inversion method for microseismic events based on hierarchical optimization.

[0004] Firstly, a hierarchical optimization-based adaptive inversion method for microseismic events is provided, including:

[0005] Step 1: Obtain the raw data of the microseismic event, which includes data from multiple channels; perform two-dimensional variational mode decomposition on the data of each channel to obtain multiple intrinsic mode components; select the intrinsic mode components that are conducive to first arrival picking for signal reconstruction to obtain the reconstructed signal; obtain the first first arrival time of the channel from the reconstructed signal based on the automatic picking algorithm;

[0006] Step 2: For the current iteration, construct the source location objective function based on the current first arrival time of the channel; use the adaptive differential evolution algorithm to solve the source location objective function to obtain the initial source position of the channel; if the current iteration is the first iteration, the current first arrival time of the channel is the first first arrival time of the channel;

[0007] Step 3: Based on the initial source location of the channel, the simulated first arrival time of the channel is obtained. The Lagrange multiplier method is used to construct a layered optimization function based on the simulated first arrival time of the channel. The layered optimization function is solved to obtain the corrected first arrival time of the channel.

[0008] Step 4: Construct an error-weighted coefficient constrained source location objective function based on the corrected first arrival time of the channel, and use an adaptive differential evolution algorithm to solve the error-weighted coefficient constrained source location objective function to obtain the current source coordinates after error-weighted coefficient constraint;

[0009] Step 5: Determine if the iteration termination condition has been met. If yes, the source coordinates after the error weighting coefficient constraint are the final microseismic event location. If no, return to step 2 and proceed to the next iteration, using the corrected first arrival time of the channel as the current first arrival time of the channel.

[0010] In one embodiment, the objective function for earthquake source location is:

[0011]

[0012] in, Let t0 be the time of earthquake occurrence, (x0, y0, z0) be the coordinates of the earthquake source, and t be the objective function. i Let x be the initial arrival time of the i-th channel. i ,y i ,z i ) represents the device coordinates for obtaining the data of the i-th channel, N is the number of channels, and v is the propagation speed of the seismic wave.

[0013] In one embodiment, the inner optimization function of hierarchical optimization is:

[0014] minQ(t)=min AIC(t)++γ||tt cal ||

[0015] AIC(t)=tlg(var(x[1,t]))+(nt-1)lg(var(x[t+1,n])

[0016]

[0017] Where Q(t) is the inner-layer optimization function of the hierarchical optimization, AIC(t) is the AIC value, t is the initial arrival time, γ is the Lagrange coefficient, and t cal ν is the simulated first arrival time of the channel; var is the variance of the data sequence; x[1,t] is the amplitude from sampling point 1 to t; n is the time window length; x[t+1,n] is the amplitude from sampling point t+1 to n; and ||| represents the norm.

[0018] In one embodiment, the objective function for source location constrained by the error weighting coefficient is:

[0019]

[0020] in, Here, t0 is the seismic source location objective function constrained by error weighting coefficients, (x0, y0, z0) is the seismic origin time, and (x0, y0, z0) are the source coordinates. i Let x be the initial arrival time of the i-th channel. i ,y i ,z iThe coordinates of the device used to obtain data from the i-th channel are given, where N is the number of channels, v is the propagation speed of the seismic wave, and r is the velocity of the seismic wave. i Calculate the error for the data in the i-th channel; E i E represents the error in calculating the initial arrival time of the i-th channel. med Let t be the median of the error calculated from the first arrival times of N channels. i Let be the corrected initial arrival time of the i-th channel. Let be the simulated first arrival time of the i-th channel.

[0021] Secondly, a microseismic event adaptive inversion device based on hierarchical optimization is provided, comprising:

[0022] The first arrival time acquisition module is used to acquire the raw data of microseismic events, which includes data from multiple channels. Two-dimensional variational mode decomposition is performed on the data of each channel to obtain multiple intrinsic mode components. The intrinsic mode components that are conducive to first arrival acquisition are selected for signal reconstruction to obtain the reconstructed signal. The first arrival time of each channel is obtained from the reconstructed signal based on an automatic acquisition algorithm.

[0023] The first arrival time determination module is used to construct the source location objective function based on the current arrival time of the channel for the current iteration; the adaptive differential evolution algorithm is used to solve the source location objective function to obtain the initial source position of the channel; if the current iteration is the first iteration, the current arrival time of the channel is the first arrival time of the channel.

[0024] The first arrival time correction module is used to simulate the simulated first arrival time of the channel based on the initial source position of the channel. The Lagrange multiplier method is used to construct a hierarchical optimization function based on the simulated first arrival time of the channel. The hierarchical optimization function is solved to obtain the corrected first arrival time of the channel.

[0025] The source coordinate determination module is used to construct a source location objective function constrained by error weighting coefficients based on the corrected first arrival time of the channel. The adaptive differential evolution algorithm is used to solve the source location objective function constrained by error weighting coefficients to obtain the current source coordinates after the error weighting coefficient constraint.

[0026] The judgment module is used to determine whether the iteration termination condition has been met. If so, the source coordinates after the error weighting coefficient constraint are the final microseismic event location. If not, the module enters the first arrival time determination module to perform the next iteration, and the corrected arrival time of the channel is used as the current arrival time of the channel.

[0027] Thirdly, a computer-readable storage medium is provided, which stores a computer program that, when executed by a processor, implements the aforementioned microseismic event adaptive inversion method based on hierarchical optimization.

[0028] Fourthly, a computer program product is provided, including a computer program / instruction, which, when executed by a processor, implements the aforementioned microseismic event adaptive inversion method based on hierarchical optimization.

[0029] Compared with the prior art, this application has the following beneficial effects:

[0030] 1. Two-dimensional variational mode decomposition is used to decompose microseismic events, extracting modes with different center frequencies. Based on the dominant frequency band of the first arrival of the microseismic event, several appropriate characteristic mode functions are selected to reconstruct the microseismic signal. The signal-to-noise ratio of the microseismic event signal in the event pickup frequency band is improved through two-dimensional variational mode decomposition.

[0031] 2. Adaptive differential evolution is used for external optimization function inversion. The adaptive differential evolution algorithm solves the problems of traditional differential evolution algorithm, such as reliance on human experience and slow convergence speed, by dynamically adjusting parameters and strategies. Its performance is improved in terms of global search, robustness and applicability.

[0032] 3. By constructing a hierarchical optimization model, the internal optimization model adaptively corrects the picked first arrival based on the source location and the first arrival picking algorithm, and the external optimization model performs source location inversion on the corrected first arrival. By simulating the process of manually modifying the first arrival through the hierarchical optimization model, the correction of the event's first arrival time is combined with the source location inversion to achieve fully automatic correction of the first arrival time of microseismic events, thereby obtaining high-precision positioning results. Attached Figure Description

[0033] This application can be better understood by referring to the description given below in conjunction with the accompanying drawings, which, together with the detailed description below, are incorporated in and form part of this specification. In the drawings:

[0034] Figure 1 A flowchart of the microseismic event adaptive inversion method based on hierarchical optimization is shown. Detailed Implementation

[0035] Exemplary embodiments of the present application will be described below with reference to the accompanying drawings. For clarity and brevity, not all features of the actual embodiments are described in the specification. However, it should be understood that many embodiment-specific decisions can be made in the development of any such actual embodiment to achieve the developer’s specific objectives, and these decisions may vary as the embodiments differ.

[0036] It should also be noted that, in order to avoid obscuring this application with unnecessary details, only the device structure closely related to the solution according to this application is shown in the accompanying drawings, while other details that are not closely related to this application are omitted.

[0037] It should be understood that this application is not limited to the described embodiments by virtue of the following description with reference to the accompanying drawings. In this document, embodiments may be combined with each other, features may be substituted or borrowed between different embodiments, and one or more features may be omitted in one embodiment, where feasible.

[0038] This application provides an adaptive inversion method for microseismic events based on hierarchical optimization. Figure 1 A flowchart of the microseismic event adaptive inversion method based on hierarchical optimization is shown. See [link / reference]. Figure 1 The method mainly includes the following steps:

[0039] Step 1: Obtain the raw data of the microseismic event, which includes data from multiple channels; perform two-dimensional variational mode decomposition on the data of each channel to obtain multiple intrinsic mode components; select the intrinsic mode components that are conducive to first arrival picking for signal reconstruction to obtain the reconstructed signal; obtain the first first arrival time of the channel from the reconstructed signal based on the automatic picking algorithm.

[0040] Here, the intrinsic mode components that are advantageous for first-arrival pickup can be, for example, components with prominent high-frequency characteristics, high signal-to-noise ratio, clear and consistent phase information in time resolution, and / or obvious envelope abrupt changes. That is, the components that most clearly and sharply express the first pulse of the first arrival wave (P-wave) in time and time. They usually have a high center frequency, which can effectively suppress background noise in the frequency band (high signal-to-noise ratio), thereby highlighting the steep rising edge and start time of the first arrival wave, providing the best conditions for accurate pickup.

[0041] Step 2: For the current iteration, construct the source location objective function based on the current first arrival time of the channel; use the adaptive differential evolution algorithm to solve the source location objective function to obtain the initial source position of the channel; if the current iteration is the first iteration, the current first arrival time of the channel is the first first arrival time of the channel.

[0042] Specifically, the objective function for earthquake source location is:

[0043]

[0044] in, Let t0 be the time of earthquake occurrence, (x0, y0, z0) be the coordinates of the earthquake source, and t be the objective function. i Let x be the initial arrival time of the i-th channel. i ,y i ,zi The coordinates of the device used to obtain the data for the i-th channel are given. Here, the device is a sensor that detects seismic waves, N is the number of channels, and v is the propagation speed of the seismic waves.

[0045] Step 3: Based on the initial source location of the channel, the simulated first arrival time of the channel is obtained. The Lagrange multiplier method is used to construct a hierarchical optimization function based on the simulated first arrival time of the channel. The hierarchical optimization function is solved to obtain the corrected first arrival time of the channel.

[0046] Specifically, the inner optimization function of the hierarchical optimization is:

[0047] minQ(t)=min AIC(t)++γ||tt cal ||

[0048] AIC(t)=tlg(var(x[1,t]))+(nt-1)lg(var(x[t+1,n])

[0049]

[0050] Where Q(t) is the inner-layer optimization function of the hierarchical optimization, AIC(t) is the AIC value, t is the initial arrival time, γ is the Lagrange coefficient, and t cal ν is the simulated first arrival time of the channel; var is the variance of the data sequence; x[1,t] is the amplitude from sampling point 1 to t; n is the time window length; x[t+1,n] is the amplitude from sampling point t+1 to n; and ||| represents the norm.

[0051] Here, we introduce the Lagrange factor γ and use the Lagrange multiplier method to construct a new objective function based on the initial arrival picking AIC algorithm, which is the inner optimization function of the hierarchical optimization.

[0052] Using the time difference between the automatically acquired arrival time and the simulated arrival time as a time window, the time within the time window that minimizes Q(t) is the result of correcting the automatically acquired arrival time based on the positioning results. In other words, the problem is transformed into calculating the value of t that minimizes Q(t) within the time window, i.e., t satisfies:

[0053]

[0054] Step 4: Construct an error-weighted coefficient constrained source location objective function based on the corrected first arrival time of the channel, and use an adaptive differential evolution algorithm to solve the error-weighted coefficient constrained source location objective function to obtain the current source coordinates after error-weighted coefficient constraint.

[0055] Specifically, the objective function for source location under error weighting coefficient constraints is:

[0056]

[0057]

[0058] in, Here, t0 is the seismic source location objective function constrained by error weighting coefficients, (x0, y0, z0) is the seismic origin time, and (x0, y0, z0) are the source coordinates. i Let x be the initial arrival time of the i-th channel. i ,y i ,z i The coordinates of the device used to obtain data from the i-th channel are given, where N is the number of channels, v is the propagation speed of the seismic wave, and r is the velocity of the seismic wave. i Calculate the error for the data in the i-th channel; E i E represents the error in calculating the initial arrival time of the i-th channel. med Let t be the median of the error calculated from the first arrival times of N channels. i Let be the corrected initial arrival time of the i-th channel. Let be the simulated first arrival time of the i-th channel.

[0059] The adaptive differential evolution algorithm is used to solve the above objective function. Unlike the conventional differential evolution algorithm, which requires manual input of parameters such as initial population size, scaling factor, crossover probability, and strategy selection, adaptive differential evolution only requires empirical input of the initial population size. The scaling factor, crossover probability, and strategy selection are dynamically optimized by the algorithm, reducing parameter sensitivity and the dependence of different microseismic events on the initial parameters. This maintains the algorithm's strong global optimization capability across different events, increasing the success rate of microseismic event inversion. After optimization by adaptive differential evolution, the source coordinates and origin time are obtained after error weighting coefficient constraints.

[0060] Step 5: Determine if the iteration termination condition has been met. If yes, the source coordinates after the error weighting coefficient constraint are the final microseismic event location. If no, return to step 2 and proceed to the next iteration, using the corrected first arrival time of the channel as the current first arrival time of the channel.

[0061] Here, the iteration termination condition can be: when the error of the source coordinates obtained from two consecutive iterations after being constrained by the error weighting coefficient is less than a set threshold.

[0062] This embodiment first extracts modes that are more conducive to first arrival picking through two-dimensional variational mode decomposition and reconstructs the signal to improve the signal-to-noise ratio of the first arrival picking signal; it then uses inner and outer layer optimization algorithms to jointly iteratively optimize and simulate the process of artificially correcting the first arrival of microseismic events; finally, the adaptive differential evolution optimization algorithm improves the global search capability of the outer layer optimization by adaptively adjusting the mutation factor and crossover rate.

[0063] Employing the same inventive concept as the hierarchical optimization-based adaptive inversion method for microseismic events, this embodiment also provides a corresponding hierarchical optimization-based adaptive inversion device for microseismic events, including:

[0064] The first arrival time acquisition module is used to acquire the raw data of microseismic events, which includes data from multiple channels. Two-dimensional variational mode decomposition is performed on the data of each channel to obtain multiple intrinsic mode components. The intrinsic mode components that are conducive to first arrival acquisition are selected for signal reconstruction to obtain the reconstructed signal. The first arrival time of each channel is obtained from the reconstructed signal based on an automatic acquisition algorithm.

[0065] The first arrival time determination module is used to construct the source location objective function based on the current arrival time of the channel for the current iteration; the adaptive differential evolution algorithm is used to solve the source location objective function to obtain the initial source position of the channel; if the current iteration is the first iteration, the current arrival time of the channel is the first arrival time of the channel.

[0066] The first arrival time correction module is used to simulate the simulated first arrival time of the channel based on the initial source position of the channel. The Lagrange multiplier method is used to construct a hierarchical optimization function based on the simulated first arrival time of the channel. The hierarchical optimization function is solved to obtain the corrected first arrival time of the channel.

[0067] The source coordinate determination module is used to construct a source location objective function constrained by error weighting coefficients based on the corrected first arrival time of the channel. The adaptive differential evolution algorithm is used to solve the source location objective function constrained by error weighting coefficients to obtain the current source coordinates after the error weighting coefficient constraint.

[0068] The judgment module is used to determine whether the iteration termination condition has been met. If so, the source coordinates after the error weighting coefficient constraint are the final microseismic event location. If not, the module enters the first arrival time determination module to perform the next iteration, and the corrected arrival time of the channel is used as the current arrival time of the channel.

[0069] The microseismic event adaptive inversion device based on hierarchical optimization in this embodiment has the same inventive concept as the microseismic event adaptive inversion method based on hierarchical optimization described above. Therefore, the specific implementation of this device can be found in the embodiment section of the microseismic event adaptive inversion method based on hierarchical optimization described above, and its technical effects correspond to the technical effects of the above method, so it will not be repeated here.

[0070] This application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the above-described microseismic event adaptive inversion method based on hierarchical optimization.

[0071] This application provides a computer program product, including a computer program / instructions, which, when executed by a processor, implements the aforementioned microseismic event adaptive inversion method based on hierarchical optimization.

[0072] In summary, this application has the following beneficial effects:

[0073] 1. Two-dimensional variational mode decomposition is used to decompose microseismic events, extracting modes with different center frequencies. Based on the dominant frequency band of the first arrival of the microseismic event, several appropriate characteristic mode functions are selected to reconstruct the microseismic signal. The signal-to-noise ratio of the microseismic event signal in the event pickup frequency band is improved through two-dimensional variational mode decomposition.

[0074] 2. Adaptive differential evolution is used for external optimization function inversion. The adaptive differential evolution algorithm solves the problems of traditional differential evolution algorithm, such as reliance on human experience and slow convergence speed, by dynamically adjusting parameters and strategies. Its performance is improved in terms of global search, robustness and applicability.

[0075] 3. By constructing a hierarchical optimization model, the internal optimization model adaptively corrects the picked first arrival based on the source location and the first arrival picking algorithm, and the external optimization model performs source location inversion on the corrected first arrival. By simulating the process of manually modifying the first arrival through the hierarchical optimization model, the correction of the event's first arrival time is combined with the source location inversion to achieve fully automatic correction of the first arrival time of microseismic events, thereby obtaining high-precision positioning results.

[0076] The above descriptions are merely various embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A microseismic event adaptive inversion method based on hierarchical optimization, characterized in that, include: Step 1: Obtain raw data of microseismic events, which includes data from multiple channels; Two-dimensional variational mode decomposition is performed on the data of each channel to obtain multiple intrinsic mode components; The intrinsic mode components that are advantageous for first-arrival pickup are selected for signal reconstruction to obtain the reconstructed signal; the first first-arrival time of the channel is obtained from the reconstructed signal based on the automatic pickup algorithm; Step 2: For the current iteration, construct the source location objective function based on the current first arrival time of the channel; use the adaptive differential evolution algorithm to solve the source location objective function to obtain the initial source position of the channel; if the current iteration is the first iteration, the current first arrival time of the channel is the first first arrival time of the channel; Step 3: Based on the initial source location of the channel, simulate the simulated first arrival time of the channel. Use the Lagrange multiplier method to construct a hierarchical optimization function based on the simulated first arrival time of the channel. Solve the hierarchical optimization function to obtain the corrected first arrival time of the channel. Step 4: Construct an error-weighted coefficient constrained source location objective function based on the corrected first arrival time of the channel, and use an adaptive differential evolution algorithm to solve the error-weighted coefficient constrained source location objective function to obtain the current source coordinates after error-weighted coefficient constraint; Step 5: Determine whether the iteration termination condition has been met. If yes, the source coordinates after the error weighting coefficient constraint are the final microseismic event location. If no, return to step 2 and proceed to the next iteration, using the corrected first arrival time of the channel as the current first arrival time of the channel. The inner optimization function of the hierarchical optimization is: ) in, For the inner optimization function of hierarchical optimization, AIC value, The initial arrival time, For Lagrange coefficients, This represents the simulated initial arrival time of the channel; The variance of the data sequence. From sampling point 1 to amplitude, The time window length, To sample points arrive amplitude, Represents the norm; The objective function for source location under the error weighting coefficient constraint is: in, The objective function for source location is constrained by error weighting coefficients. The moment of the earthquake, The coordinates of the earthquake source are... To obtain the first The device coordinates for each channel of data, where N is the number of channels. This refers to the propagation speed of seismic waves; For the first Data calculation error for each channel; For the first Error in calculating the initial arrival time of each channel Let the median of the error calculated from the first arrival times of N channels be denoted as . For the first The corrected first arrival times for each channel, For the first Simulated initial arrival time for each channel.

2. The method as described in claim 1, characterized in that, The objective function for earthquake source localization is: in, Let be the objective function. The moment of the earthquake, The coordinates of the earthquake source are... For the first The initial arrival time of each channel, To obtain the first The device coordinates for each channel of data, where N is the number of channels. This represents the propagation speed of seismic waves.

3. A microseismic event adaptive inversion device based on hierarchical optimization, characterized in that, include: The arrival time acquisition module is used to acquire raw data of microseismic events, which includes data from multiple channels. Two-dimensional variational mode decomposition is performed on the data of each channel to obtain multiple intrinsic mode components; The intrinsic mode components that are advantageous for first-arrival pickup are selected for signal reconstruction to obtain the reconstructed signal; the first first-arrival time of the channel is obtained from the reconstructed signal based on the automatic pickup algorithm; The first arrival time determination module is used to construct a source location objective function based on the current arrival time of the channel for the current iteration; solve the source location objective function using an adaptive differential evolution algorithm to obtain the initial source position of the channel; if the current iteration is the first iteration, the current arrival time of the channel is the first arrival time of the channel; The first arrival time correction module is used to simulate the simulated first arrival time of the channel based on the initial source position of the channel, construct a hierarchical optimization function based on the simulated first arrival time of the channel using the Lagrange multiplier method, solve the hierarchical optimization function, and obtain the corrected first arrival time of the channel. The source coordinate determination module is used to construct a source location objective function constrained by error weighting coefficients based on the corrected first arrival time of the channel, and to solve the source location objective function constrained by error weighting coefficients using an adaptive differential evolution algorithm to obtain the current source coordinates after being constrained by error weighting coefficients. The judgment module is used to determine whether the iteration termination condition has been met. If so, the source coordinates after the error weighting coefficient constraint are the final microseismic event location. If not, the module enters the first arrival time determination module to perform the next iteration, and the corrected arrival time of the channel is used as the current arrival time of the channel. The inner optimization function of the hierarchical optimization is: ) in, For the inner optimization function of hierarchical optimization, AIC value, The initial arrival time, For Lagrange coefficients, This represents the simulated initial arrival time of the channel; The variance of the data sequence. From sampling point 1 to amplitude, The time window length, To sample points arrive amplitude, Represents the norm; The objective function for source location under the error weighting coefficient constraint is: in, The objective function for source location is constrained by error weighting coefficients. The moment of the earthquake, The coordinates of the earthquake source are... To obtain the first The device coordinates for each channel of data, where N is the number of channels. This refers to the propagation speed of seismic waves; For the first Data calculation error for each channel; For the first Error in calculating the initial arrival time of each channel Let the median of the error calculated from the first arrival times of N channels be denoted as . For the first The corrected first arrival times for each channel. For the first Simulated initial arrival time for each channel.

4. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the microseismic event adaptive inversion method based on hierarchical optimization as described in claim 1 or 2.

5. A computer program product, characterized in that, Includes a computer program / instruction, which, when executed by a processor, implements the hierarchical optimization-based microseismic event adaptive inversion method as described in claim 1 or 2.

Images