Dual-difference first-arrival data processing method and device based on random reference trace selection and statistical synthesis processing, medium and equipment

Abstract

This invention relates to a method, apparatus, medium, and device for processing double-difference first-arrival data based on random reference channel selection and statistical synthesis. The method includes: acquiring the observed first-arrival data of a shot gather and the receiver channel information corresponding to the observed first-arrival data; constructing a corresponding local candidate set according to preset constraints; determining the random selection weight of each candidate reference channel for each local candidate set corresponding to a target channel; for each target channel, performing random selection within its local candidate set according to the random selection weight to obtain the first-arrival of the reference channel, and establishing a probability distribution of the selection of each candidate reference channel within the local candidate set based on the random selection weight; constructing a single double-difference result corresponding to the target channel based on the first-arrival of the target channel and the first-arrival of the randomly selected reference channel; repeating the steps a preset number of times for the same target channel to generate multiple sets of double-difference results; and performing robust statistical synthesis processing on the multiple sets of double-difference results to obtain multiple sets of double-difference results.

Description

Technical Field

[0001] This invention relates to a method, apparatus, medium, and equipment for processing first arrival data with double difference based on random reference trace selection and statistical synthesis, belonging to the field of geophysical exploration data processing technology. Background Technology

[0002] In onshore seismic exploration, first-arrival information is widely used for near-surface velocity modeling, static correction, tomographic inversion, and imaging preprocessing. In actual production, observed first-arrivals are usually acquired by commercial software modules or manual processing workflows. Because different software, different processing personnel, and different parameter settings may employ different first-arrival definition and labeling methods and first-arrival acquisition strategies, such as the first wave initiation point, the first positive peak, the first negative peak, the envelope threshold point, or software-built-in rules, the acquired first-arrivals often have systematic deviations from the ideal first-arrivals.

[0003] The aforementioned systematic biases can manifest as shared time shifts across multiple channels within the same shot gather, or as slowly varying biases that differ with offset, waveform morphology, local topographic relief, near-surface lateral inhomogeneity, and scattering intensity. If absolute first arrivals are directly fitted in subsequent processing or inversion, these systematic biases will directly enter the objective function, leading to measurement errors unrelated to the subsurface medium model being mixed into the residuals, thereby reducing the stability and physical interpretability of the results.

[0004] To reduce the impact of absolute first-arrival errors, existing techniques often employ the double-difference method. This involves constructing relative constraints within the same shot gather using the first-arrival difference between the target trajectory and the reference trajectory, aiming to eliminate some common term errors. However, existing techniques typically construct double-difference by fixing the reference trajectory, pre-setting a single reference trajectory, or fixing a pair of reference trajectories. This approach has the following problems: 1. Fixed reference trace leads to large random errors. If a certain reference trace is used in the same shot collection, the final double difference result will be significantly affected by the quality of the reference trace, waveform morphology, picking stability and local anomalies.

[0005] 2. It cannot effectively adapt to systematic biases caused by inconsistent picking standards. Existing technologies usually assume that the observed first solstices are stable and consistent. However, in actual production, different first solstices definition and labeling methods and first solstices picking strategies can lead to systematic biases between the picked first solstices and the ideal first solstices. Moreover, this bias is not always a global constant.

[0006] 3. Lack of a reference trace candidate mechanism under local constraints. If the range of reference trace candidates is not restricted, the selection of reference traces may cross the switching zone of the first wave event or the region with large waveform differences, reducing the physical consistency of the double difference results.

[0007] 4. Lack of multiple construction and statistical synthesis mechanisms. Existing technologies generally only generate a single set of doubledifference results, and cannot reduce the randomness of reference channel selection through multiple random constructions and statistical synthesis.

[0008] 5. Difficulty in establishing a standardized software implementation process. Reference selection often relies on human experience and lacks a clear, integrated process for data input, candidate set construction, random selection, double-difference generation, and statistical synthesis output. Summary of the Invention

[0009] To address the aforementioned technical problems, this invention provides a method, apparatus, medium, and device for processing double-difference first-arrival data based on random reference channel selection and statistical synthesis. This method constructs a local candidate set of the target channel, randomly selects a reference channel within the local candidate set, repeatedly generates multiple sets of double-difference results, and performs statistical synthesis on the multiple sets of double-difference results to obtain a more robust double-difference output result in a statistical sense.

[0010] To achieve the above objectives, the present invention adopts the following technical solution: A method for processing double-difference first-arrival data based on random reference channel selection and statistical synthesis includes: S1: Obtain the first arrival data of at least one shot set and the receiver channel information corresponding to the first arrival data; S2: Based on the first arrival data and receiver information, construct a corresponding local candidate set for each target trace in the same shot set according to preset constraints; S3: For each target channel, determine the random selection weights for each candidate reference channel within the local candidate set. ; S4: For each target path, within its local candidate set, randomly select weights. Perform random selection to obtain the initial arrival of the reference path, based on the random selection weights. Establish the probability distribution of each candidate reference channel selected within the local candidate set. ; S5: Based on the first arrival of the target trajectory and the first arrival of the randomly selected reference trajectory, construct the first double difference result corresponding to the target trajectory; S6: For the same target path, repeat steps S4 and S5 a preset number of times. ,generate Group of double-difference results; S7: Yes Robust statistical synthesis was performed on the double-difference results to remove random errors, resulting in... Group double difference results ; S8: Will The corresponding residuals are input into the tomographic inversion matrix or the objective function of the wave equation, and iterative inversion calculations are performed to update and output the near-surface velocity model or static correction, thereby transforming the pure data processing results into an accurate characterization of the geological structure of geophysical exploration.

[0011] The double-difference first arrival data processing method based on random reference channel selection and statistical synthesis is preferably described in step S1, where the received channel information includes the received channel number and position, as well as at least one of the following: channel spacing, received coordinates, shot group number, signal-to-noise ratio, pickup quality flag, and waveform data.

[0012] The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis is preferably wherein, in step S2, the candidate reference channels in the local candidate set satisfy at least one of the following conditions: (1) Spatial distance constraint: spatial distance between the target lane and the target lane , D max Distance threshold; (2) Offset distance constraint: the absolute offset distance difference between the target track and the target track. , O max This represents the maximum deviation of the offset distance; (3) Waveform similarity constraint: cross-correlation coefficient of the first window waveform corresponding to the target channel , C min The minimum similarity coefficient; (4) Quality constraints: The initial pickup quality meets the preset conditions and is not a bad path, dead path, missing path, or abnormal path; (5) Physical continuity constraint: Do not cross the preset first wave event switching zone.

[0013] The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis, preferably, in step S3, involves randomly selecting weights. The calculation formula is as follows:

[0014] in, For candidate reference tracks; Spatial distance from the target road; This is the distance attenuation control parameter; The cross-correlation coefficient of the waveform; The normalized pickup quality score; The preset non-negative adjustment coefficient, and .

[0015] The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis, preferably, in step S4, the probability distribution... The calculation formula is as follows:

[0016] in, This represents the total number of reference channels in the local candidate set.

[0017] The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis is preferably described in step S5, where the double-difference result includes at least one of the following: the difference between the first-arrival observed in the target channel and the first-arrival observed in the reference channel, the difference between the composite first-arrival of the target channel and the composite first-arrival of the reference channel, and the double-difference residual formed based on the above two.

[0018] The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis, preferably, in step S6, outlier removal is performed using outlier removal and averaging based on the absolute median difference, as detailed below: (1): Calculation Median of group results ; (2): Calculate the absolute median , This is the k-th double-difference time travel; (3): Eliminate those that satisfy the condition outliers; (4): Calculate the arithmetic mean or weighted average of the remaining valid double-difference results, and output the final double-difference result in a statistical sense. .

[0019] A second aspect of the present invention provides a device for modeling the heat transfer performance mechanism of an open-frame gasifier, comprising: The first processing unit is used to acquire the first arrival data of at least one shot set and the receiver channel information corresponding to the first arrival data; The second processing unit is used to construct a corresponding local candidate set for each target trace in the same shot set according to preset constraints, based on the observation first arrival data and the receiver trace information. The third processing unit is used to determine the random selection weight of each candidate reference channel for each local candidate set corresponding to the target channel. ; The fourth processing unit is used to, for each target channel, randomly select weights within its local candidate set. Perform random selection to obtain the initial arrival of the reference path, based on the random selection weights. Establish the probability distribution of each candidate reference channel selected within the local candidate set. ; The fifth processing unit is used to construct a first-order double-difference result corresponding to the target trace based on the first arrival of the target trace and the first arrival of the randomly selected reference trace. The sixth processing unit is used to repeatedly execute steps S4 and S5 a preset number of times for the same target channel. ,generate Group of double-difference results; The seventh processing unit is used for... Robust statistical synthesis was performed on the double-difference results to remove random errors, resulting in... Group double difference results ; The eighth processing unit is used to process... The corresponding residuals are input into the tomographic inversion matrix or the objective function of the wave equation, and iterative inversion calculations are performed to update and output the near-surface velocity model or static correction, thereby transforming the pure data processing results into an accurate characterization of the geological structure of geophysical exploration.

[0020] A third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing described in any one of the preceding claims.

[0021] A fourth aspect of the present invention 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 the computer program to implement the steps of the double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing described in any one of the above-described methods.

[0022] The present invention has the following advantages due to the adoption of the above technical solutions: 1. This invention constrains the selection of reference channels by using local candidate sets, thus providing a clear physical and data quality basis for the selection and avoiding the irrationality caused by global random selection.

[0023] 2. This invention avoids the over-reliance on the quality of a single track and local anomalies by randomly selecting a reference track.

[0024] 3. This invention reduces the impact of the randomness of reference channel selection on the final double difference result by constructing multiple double differences and statistical synthesis.

[0025] 4. This invention has better adaptability to systematic deviations caused by different initial arrival definition annotation methods and initial arrival picking strategies, and can suppress common term errors and gradual deviations to a certain extent.

[0026] 5. This invention does not rely on a specific tomographic solution model and is compatible with subsequent processing procedures based on ray theory or wave equations.

[0027] 6. The probability sampling mechanism of this invention retains the tendency to select high-quality adjacent channels while also giving the network diversity, thus avoiding excessive reliance on a single local abnormal channel.

[0028] 7. The robust statistics of the present invention (such as MAD-based processing) can effectively immunize against local bursts of picking errors or phase jumps.

[0029] 8. The method of the present invention is easy to implement in engineering and can form a standardized processing flow from the construction of local candidate sets, selection of random reference channels, double difference construction to statistical synthesis output. Attached Figure Description

[0030] Figure 1 This is a flowchart of a double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis, provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the target trajectory and its local candidate set within the same shot collection provided in this embodiment of the present invention; Figure 3 This is a schematic diagram of random reference channel selection and multiple double-difference construction provided in this embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the statistical synthesis of multiple sets of double-difference results provided in this embodiment of the present invention; Figure 5 This is a framework diagram for connecting the output results provided in this embodiment of the invention to a subsequent tomography or fitting system. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention are described clearly and completely below. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0032] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," "third," "fourth," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect.

[0033] For ease of description, spatial relative terms may be used in the text to describe the relationship of one element or feature relative to another element or feature as shown in the figure. These relative terms include, for example, "inside," "outside," "middle," "outer," "below," "above," etc. Such spatial relative terms are intended to include different orientations of the device in use or operation, other than those depicted in the figure.

[0034] Terminology Explanation: "Statistical synthesis processing" refers to the averaging, weighted averaging, median, outlier removal, robust statistics, or a combination thereof, of multiple sets of double-difference results obtained from multiple random reference channel selections to form the final double-difference output result.

[0035] Existing techniques typically construct double difference by using fixed reference traces, pre-defined single reference traces, or fixed reference trace pairs. This approach has the following problems: 1. Fixed reference trace leads to large random errors. If a certain reference trace is used in the same shot collection, the final double difference result will be significantly affected by the quality of the reference trace, waveform morphology, picking stability and local anomalies.

[0036] 2. It cannot effectively adapt to systematic biases caused by inconsistent picking standards. Existing technologies usually assume that the observed first solstices are stable and consistent. However, in actual production, different first solstices definition and labeling methods and first solstices picking strategies can lead to systematic biases between the picked first solstices and the ideal first solstices. Moreover, this bias is not always a global constant.

[0037] 3. Lack of a reference trace candidate mechanism under local constraints. If the range of reference trace candidates is not restricted, the selection of reference traces may cross the switching zone of the first wave event or the region with large waveform differences, reducing the physical consistency of the double difference results.

[0038] 4. Lack of multiple construction and statistical synthesis mechanisms. Existing technologies generally only generate a single set of doubledifference results, and cannot reduce the randomness of reference channel selection through multiple random constructions and statistical synthesis.

[0039] 5. Difficulty in establishing a standardized software implementation process. Reference selection often relies on human experience and lacks a clear, integrated process for data input, candidate set construction, random selection, double-difference generation, and statistical synthesis output.

[0040] To address the aforementioned technical problems, this invention provides a method, apparatus, medium, and equipment for processing double-difference first-arrival data based on random reference trace selection and statistical synthesis. This method can be used for first-arrival tomography, near-surface velocity modeling, static correction model construction, relative travel time fitting, and other seismic data processing scenarios based on double-difference first-arrival constraints. Subsequent solutions can be implemented using either ray theory or wave equations.

[0041] This invention, for each target trajectory within the same shot collection, no longer uses a fixed reference trajectory. Instead, it first constructs a local candidate set that meets preset physical and geometric conditions. Then, based on mathematical models of indicators such as distance, waveform similarity, first-arrival picking quality, and signal-to-noise ratio, it assigns selection weights to candidate reference trajectories in the local candidate set. Next, it performs probability-based random sampling (such as roulette wheel algorithm) within the local candidate set to generate a reference trajectory. Based on the first arrival of the selected reference trajectory and the target trajectory, it constructs a double-difference result. The above random selection and double-difference construction process is repeated multiple times to obtain multiple sets of double-difference results. Finally, it uses robust statistics and other methods to comprehensively process the multiple sets of double-difference results to obtain a statistically significant double-difference output result, which is then used to reconstruct the physical parameter model of the underground medium.

[0042] like Figure 1-5 As shown, the double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis provided by this invention includes the following specific steps: S1: Obtain Data Acquire first-arrival data of at least one shot collection and corresponding receiver channel information. The receiver channel information includes at least the receiver channel number and location; preferably, it also includes at least one of the following: channel spacing, receiver coordinates, shot collection number, signal-to-noise ratio, pickup quality flag, and waveform data.

[0043] S2: Constructing a local candidate set Based on the first arrival data and receiver trace information, a corresponding local candidate set is constructed for each target trace in the same shot set according to preset constraints. The candidate reference traces in the local candidate set must satisfy at least one of the following preset constraints: 1. Spatial distance constraint: Spatial distance from the target lane , D max For distance threshold (e.g.: (Take a distance of 50 to 100 meters). 2. Offset distance constraint: The absolute offset distance difference between the target track and the target track. , O max This represents the maximum deviation of the offset distance; 3. Waveform similarity constraint: Cross-correlation coefficient between the first window waveform corresponding to the target channel and the waveform. , C min The minimum similarity coefficient (e.g.: (Take 0.75). 4. Quality Constraints: The initial pickup quality meets the preset constraints and is not a bad path, dead path, missing path, or abnormal path; 5. Physical continuity constraint: It does not cross the preset first wave event switching zone (such as the junction zone between direct wave and refracted wave), abnormal zone or unstable zone.

[0044] S3: Determine the random selection weights for candidate reference channels. For each target lane corresponding to a local candidate set, the random selection weights of each candidate reference lane are determined using a mathematical evaluation model (which consists of the five preset constraints proposed in step S2). Preferably, the weighting calculation formula integrates distance attenuation, waveform correlation, and pickup quality:

[0045] in, For candidate reference road Spatial distance from the target road; This is the distance attenuation control parameter; The cross-correlation coefficient of the waveform; The normalized pickup quality score; The preset non-negative adjustment coefficient, and .

[0046] S4: Perform random reference track selection For each target path, weights are randomly selected within its local candidate set. Perform a random selection to obtain the initial arrival of the reference path. Calculate the weights based on the weights obtained in step S3. Establish the probability distribution of each candidate reference channel selected within the local candidate set. :

[0047] in, This represents the total number of reference channels in the local candidate set. Subsequently, a channel is randomly selected as the reference channel under this probability distribution using either Roulette Wheel Selection or Monte Carlo sampling.

[0048] S5: Construct a first-order double difference result Based on the first arrival of the target trajectory and the first arrival of a randomly selected reference trajectory, construct a first-order double-difference result corresponding to the target trajectory. Ttarget For the initial arrival of the target path; T ref This is for reference when the Tao first arrives.

[0049] Preferably, for all target traces in the same shot set, a set of double-difference results is formed for that shot set under the current random reference trace configuration. The double-difference results include at least one of the following: the difference between the observed first arrival of the target trace and the observed first arrival of the reference trace, the difference between the synthesized first arrival of the target trace and the synthesized first arrival of the reference trace, and the double-difference residuals formed based on the above two.

[0050] S6: Repeated random selection and double difference construction For the same target path, repeat steps S4 and S5 a preset number of times. (Preferred) ,like ),generate Group double difference results .

[0051] S7: Perform statistical synthesis on multiple sets of double-difference results. Regarding the above Robust statistical synthesis is performed on the double-difference results to eliminate random errors. Preferably, outlier removal and averaging based on the median absolute deviation (MAD) are used. 1. Calculation Median of group results ; 2. Calculate the absolute median. , This is the k-th double-difference time travel; 3. Eliminate those that meet the requirements outliers (such as) ); 4. Calculate the arithmetic mean or weighted average of the remaining valid double-difference results, and output the final double-difference result in a statistical sense. .

[0052] S8: Use statistical synthesis results to reconstruct the subsurface medium model Will The corresponding residuals are input into the tomographic inversion matrix or the objective function of the wave equation, and iterative inversion calculations are performed to update and output the near-surface velocity model or static correction, thereby transforming the pure data processing results into an accurate characterization of the geological structure of geophysical exploration.

[0053] The method of the present invention has the following technical advantages: 1. Local candidate set: Combines physical constraints such as space, waveform or mass to define the range of reference trace candidates (building a defense line, different from global selection), and imposes clear geometric and physical similarity boundaries to prevent crossing different physical phases and avoid arbitrary global selection.

[0054] 2. Weighted probabilistic random selection: The selection of reference channels is no longer blindly random or fixed, but is sampled through a probability distribution jointly driven by "distance-waveform-quality".

[0055] 3. Multiple double-difference construction and robust statistical synthesis: By using multiple sampling and MAD anomaly removal mechanism, the destructive impact of poor quality of a single reference trace (random error) is greatly reduced.

[0056] 4. Measurement error suppression: Directly addresses the systematic biases in the picking standards of commercial software, improving the noise resistance of near-surface velocity modeling.

[0057] The technical solution of the present invention will be described in detail below with reference to specific examples.

[0058] Example 1: A Complete Example of Stochastic Double Difference Processing Based on Numerical Computation Scenario: Targeting a specific artillery barrage. (spatial location) ), and its double difference result needs to be calculated.

[0059] 1. Construct a local candidate set: Set a distance threshold Find a valid receiving channel within this range. .

[0060] 2. Weight and Probability Calculation: Calculate the overall weight according to the formula. Assume the weight of each candidate channel after calculation is... , , , The normalized probabilities are as follows: , , , .

[0061] 3. Multiple random sampling and construction: Set the number of repetitions. Sampling was conducted using roulette wheels, and the following five samples were drawn: The corresponding five double-difference results are as follows (unit: ms): .

[0062] 4. Statistical processing: The median of this data set is If a simple average is used, it may be... Raise (average) Here, median processing or outlier removal is used to determine the final output. This value is then directly input into the inversion equation. This implementation method is suitable for data with small channel spacing and good continuity of the first wave phase axis.

[0063] Example 2: Random double-difference processing based on waveform similarity screening For the target channel, first extract its initial time window; then extract the adjacent time windows. Extract the corresponding time window from the receiving channels within the range and calculate the waveform cross-correlation coefficient. Only retain the correlation coefficient. The receiver channels are used as a local candidate set. Then, the correlation coefficient itself is used as the weight for random selection (i.e., ...). ), perform roulette wheel sampling, generate The results are analyzed, and outliers are removed using the median absolute deviation (MAD), and then the average value is calculated. This implementation method is suitable for areas with extremely complex near-surface structures and drastic waveform changes.

[0064] Example 3: Construction of double-difference input for subsequent tomographic solution Based on Example 1 or Example 2, statistically significant double-difference output results are generated for observed first arrivals. For a given initial model, a synthetic first arrival is calculated using ray tracing or wave equation methods, generating a synthetic double-difference result. The difference between the observed double-difference result and the synthetic double-difference result is used to form double-difference residuals, which are then combined with the absolute first arrival residuals for subsequent tomography or fitting solutions. In this embodiment, the subsequent solution method is not a necessary limitation of the invention; the core of the invention lies in the random reference channel selection, local candidate set constraints, and the statistical synthesis processing flow of multiple double-difference results.

[0065] A second aspect of the present invention provides a device for modeling the heat transfer performance mechanism of an open-frame gasifier, comprising: The first processing unit is used to acquire the first arrival data of at least one shot set and the receiver channel information corresponding to the first arrival data; The second processing unit is used to construct a corresponding local candidate set for each target trace in the same shot set according to preset constraints, based on the observation first arrival data and the receiver trace information. The third processing unit is used to determine the random selection weight of each candidate reference channel for each local candidate set corresponding to the target channel. ; The fourth processing unit is used to, for each target channel, randomly select weights within its local candidate set. Perform random selection to obtain the initial arrival of the reference path, based on the random selection weights. Establish the probability distribution of each candidate reference channel selected within the local candidate set. ; The fifth processing unit is used to construct a first-order double-difference result corresponding to the target trace based on the first arrival of the target trace and the first arrival of the randomly selected reference trace. The sixth processing unit is used to repeatedly execute steps S4 and S5 a preset number of times for the same target channel. ,generate Group of double-difference results; The seventh processing unit is used for... Robust statistical synthesis was performed on the double-difference results to remove random errors, resulting in... Group double difference results ; The eighth processing unit is used to process... The corresponding residuals are input into the tomographic inversion matrix or the objective function of the wave equation, and iterative inversion calculations are performed to update and output the near-surface velocity model or static correction, thereby transforming the pure data processing results into an accurate characterization of the geological structure of geophysical exploration.

[0066] A third aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing described in any one of the preceding claims.

[0067] A fourth aspect of the present invention 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 the computer program to implement the steps of the double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing described in any one of the above-described methods.

[0068] This invention is described based on flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to specific embodiments. It should be understood that each block of the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing device, generate instructions for implementing the flowcharts and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0069] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0070] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0071] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for processing double-difference first-arrival data based on random reference channel selection and statistical synthesis, characterized in that, include: S1: Obtain the first arrival data of at least one shot set and the receiver channel information corresponding to the first arrival data; S2: Based on the first arrival data and receiver information, construct a corresponding local candidate set for each target trace in the same shot set according to preset constraints; S3: For each target channel, determine the random selection weights for each candidate reference channel within the local candidate set. ; S4: For each target path, within its local candidate set, randomly select weights. Perform random selection to obtain the initial arrival of the reference path, based on the random selection weights. Establish the probability distribution of each candidate reference channel selected within the local candidate set. ; S5: Based on the first arrival of the target trajectory and the first arrival of the randomly selected reference trajectory, construct the first double difference result corresponding to the target trajectory; S6: For the same target path, repeat steps S4 and S5 a preset number of times. ,generate Group of double-difference results; S7: Yes Robust statistical synthesis was performed on the double-difference results to remove random errors, resulting in... Group double difference results ; S8: Will The corresponding residuals are input into the tomographic inversion matrix or the objective function of the wave equation, and iterative inversion calculations are performed to update and output the near-surface velocity model or static correction, thereby transforming the pure data processing results into an accurate characterization of the geological structure of geophysical exploration.

2. The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing according to claim 1, characterized in that, In step S1, the receiving channel information includes the receiving channel number and location, as well as at least one of the following: channel spacing, receiving coordinates, shot group number, signal-to-noise ratio, pickup quality flag, and waveform data.

3. The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing according to claim 1, characterized in that, In step S2, the candidate reference channels in the local candidate set satisfy at least one of the following conditions: (1) Spatial distance constraint: spatial distance between the target lane and the target lane , D max Distance threshold; (2) Offset distance constraint: the absolute offset distance difference between the target track and the target track. , O max This represents the maximum deviation of the offset distance; (3) Waveform similarity constraint: cross-correlation coefficient of the first window waveform corresponding to the target channel , C min The minimum similarity coefficient; (4) Quality constraints: The initial pickup quality meets the preset conditions and is not a bad path, dead path, missing path, or abnormal path; (5) Physical continuity constraint: Do not cross the preset first wave event switching zone.

4. The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing according to claim 1, characterized in that, In step S3, weights are randomly selected. The calculation formula is as follows: ； in, For candidate reference tracks; Spatial distance from the target road; This is the distance attenuation control parameter; The cross-correlation coefficient of the waveform; The normalized pickup quality score; The preset non-negative adjustment coefficient, and .

5. The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing according to claim 1, characterized in that, In step S4, the probability distribution The calculation formula is as follows: ； in, This represents the total number of reference channels in the local candidate set.

6. The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing according to claim 1, characterized in that, In step S5, the double difference result includes at least one of the following: the difference between the first arrival observed in the target trace and the first arrival observed in the reference trace, the difference between the composite first arrival of the target trace and the composite first arrival of the reference trace, and the double difference residual formed based on the above two.

7. The double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing according to claim 1, characterized in that, In step S6, random errors are eliminated by outlier removal and averaging based on the absolute median difference. The specific process is as follows: (1): Calculation Median of group results ; (2): Calculate the absolute median , This is the k-th double-difference time travel; (3): Eliminate those that satisfy the condition outliers; (4): Calculate the arithmetic mean or weighted average of the remaining valid double-difference results, and output the final double-difference result in a statistical sense. .

8. A modeling device for the heat transfer performance mechanism of an open-frame gasifier, characterized in that, include: The first processing unit is used to acquire the first arrival data of at least one shot set and the receiver channel information corresponding to the first arrival data; The second processing unit is used to construct a corresponding local candidate set for each target trace in the same shot set according to preset constraints, based on the observation first arrival data and the receiver trace information. The third processing unit is used to determine the random selection weight of each candidate reference channel for each local candidate set corresponding to the target channel. ; The fourth processing unit is used to, for each target channel, randomly select weights within its local candidate set. Perform random selection to obtain the initial arrival of the reference path, based on the random selection weights. Establish the probability distribution of each candidate reference channel selected within the local candidate set. ; The fifth processing unit is used to construct a first-order double-difference result corresponding to the target trace based on the first arrival of the target trace and the first arrival of the randomly selected reference trace. The sixth processing unit is used to repeatedly execute steps S4 and S5 a preset number of times for the same target channel. ,generate Group of double-difference results; The seventh processing unit is used for... Robust statistical synthesis was performed on the double-difference results to remove random errors, resulting in... Group double difference results ; The eighth processing unit is used to process... The corresponding residuals are input into the tomographic inversion matrix or the objective function of the wave equation, and iterative inversion calculations are performed to update and output the near-surface velocity model or static correction, thereby transforming the pure data processing results into an accurate characterization of the geological structure of geophysical exploration.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing as described in any one of claims 1-7.

10. 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 computer program, it implements the steps of the double-difference first-arrival data processing method based on random reference channel selection and statistical synthesis processing as described in any one of claims 1-7.

Images