CN-121995467-A - Seismic signal self-adaptive amplitude compensation method based on multi-domain joint parameter calculation

CN121995467ACN 121995467 ACN121995467 ACN 121995467ACN-121995467-A

Abstract

The invention discloses a seismic signal self-adaptive amplitude compensation method based on multi-domain joint parameter calculation. The method relates to the technical field of amplitude compensation and comprises the following steps of preprocessing an original seismic gather, extracting multi-domain basic information, cooperatively acquiring compensation core parameters, performing self-adaptive amplitude compensation in stages, verifying compensation effects and performing iterative optimization. According to the method, the original seismic trace set is preprocessed, the multi-domain basic information is extracted, the multi-domain compensation core parameters are obtained cooperatively, the seismic signal attenuation and the energy difference are compensated in stages based on the parameters, and finally, the data adaptation subsequent geological interpretation and reservoir prediction are ensured through compensation effect verification and iterative optimization, so that the adaptive amplitude compensation accuracy of the seismic signals is improved, and the problem of low adaptive amplitude compensation accuracy of the seismic signals in the prior art is solved.

Inventors

LIU ZHIYONG
WANG HUI
BIAN SHUTAO
XIONG XIONG
CHEN XIN
WEI PENGHUI
WANG YING
LIANG YIWEI
YANG TIANSHU
ZHANG YUNLONG

Assignees

北京源烃泰克科技有限公司

Dates

Publication Date: 20260508
Application Date: 20260202

Claims (10)

1. The seismic signal self-adaptive amplitude compensation method based on multi-domain joint parameter is characterized by comprising the following steps: step 101, preprocessing an original seismic trace set obtained by seismic exploration acquisition to avoid interference of acquisition errors and near-surface propagation static time difference on seismic signals, obtaining a preprocessed seismic trace set, and synchronously extracting multi-domain basic information of the preprocessed seismic trace set; 102, based on the extracted multi-domain basic information, cooperatively acquiring core parameters required by seismic signal amplitude compensation to reduce the multi-solution property and deviation of single-domain parameter estimation and ensure the accuracy and suitability of the compensation core parameters, wherein the core parameters comprise but are not limited to a common shot point domain, a common detection point domain and a common offset distance domain; Step 103, carrying out staged self-adaptive amplitude compensation on the preprocessed seismic trace set based on core parameters so as to realize the precise compensation of the geometric attenuation, viscoelastic absorption attenuation and transverse energy difference of the seismic signals, and considering the compensation precision and the data stability, wherein the staged self-adaptive amplitude compensation comprises diffusion compensation, absorption compensation and self-adaptive gain compensation; Step 104, performing effect verification and iterative optimization on the seismic trace set subjected to the staged adaptive amplitude compensation to ensure that the compensated seismic data meet the requirements of subsequent geological interpretation and reservoir prediction, and ensure the reliability and suitability of the compensation effect.
2. The method for adaptive amplitude compensation of seismic signals based on multi-domain joint parameter determination as recited in claim 1, wherein the multi-domain includes a time domain, a frequency domain, a spatial domain, and an imaging domain; Extracting multi-domain basic information from the preprocessed seismic trace set by adopting a multi-domain parallel analysis algorithm, extracting amplitude envelope, instantaneous frequency and instantaneous phase characteristics of a seismic signal in the time domain through Hilbert transformation, and counting time-varying energy attenuation rules through a sliding time window; Performing time-frequency decomposition on the preprocessed seismic trace set in the frequency domain through generalized S transformation to obtain energy distribution characteristics including but not limited to full-band amplitude spectrum and phase spectrum; Analyzing the header information of the seismic channel in the space domain, extracting offset, azimuth, common center point coordinates, common shot point marks and common detection point marks, and counting the energy transverse change rules of the seismic signals in different space domains; Preliminary migration imaging is carried out in the imaging domain through a prestack time migration method, and inversion is carried out to obtain the structural form information of the underground stratum velocity field and the reflecting interface, so that synchronous extraction of basic information of the time domain, the frequency domain, the space domain and the imaging domain is completed, and precise matching association of multi-domain information is realized.
3. The method for adaptive amplitude compensation of seismic signals based on multi-domain joint parameter calculation as set forth in claim 1, wherein the specific steps of diffusion compensation are as follows: according to each data point in the seismic trace set, acquiring an instantaneous amplitude envelope and an instantaneous frequency, and continuously comparing the instantaneous amplitude envelope with an amplitude envelope reference value for the data point to be analyzed to obtain an instantaneous amplitude mean square error; Establishing a mapping relation between the instantaneous amplitude mean square error and the instantaneous frequency adjustment quantity, and dynamically adjusting the instantaneous frequency of the seismic signal based on the instantaneous amplitude mean square error, wherein the mapping relation comprises the following specific steps: if the instantaneous amplitude mean square error is smaller than the mean square error critical lower limit, the amplitude attenuation is normal, and the instantaneous frequency adjustment is not performed; If the instantaneous amplitude mean square error is in a mean square error critical interval, the instantaneous amplitude mean square error and the instantaneous frequency adjustment quantity are in positive correlation monotonically increasing relation, the current instantaneous amplitude mean square error is input into the mapping relation between the instantaneous amplitude mean square error and the instantaneous frequency adjustment quantity, the instantaneous frequency gain quantity is output, the instantaneous frequency is increased to compensate high-frequency loss, and the mean square error critical interval represents a closed interval formed by a mean square error critical lower limit and a mean square error critical upper limit.
4. A method for adaptive amplitude compensation of seismic signals based on multi-domain joint parameter estimation as recited in claim 3, wherein said dynamically adjusting the instantaneous frequency of the seismic signals based on instantaneous amplitude mean square error further comprises: If the instantaneous amplitude mean square error is greater than the mean square error critical upper limit, limiting the instantaneous frequency adjustment quantity to prevent excessive and unstable frequency adjustment in a strong noise or abnormal area, inputting the current instantaneous amplitude mean square error into a mapping relation between the instantaneous amplitude mean square error and the instantaneous frequency adjustment quantity, and outputting an instantaneous frequency reduction quantity; Based on the instantaneous frequency adjustment quantity, the instantaneous frequency of the current seismic signal is adjusted, specifically: If the instantaneous frequency gain is output, the current instantaneous frequency of the seismic signal is overlapped with the instantaneous frequency gain to obtain the instantaneous frequency of the target seismic signal; And if the instantaneous frequency reduction amount is output, the current instantaneous frequency of the seismic signal and the instantaneous frequency reduction amount are overlapped to obtain the instantaneous frequency of the target seismic signal.
5. The method for adaptive amplitude compensation of seismic signals based on multi-domain joint parameter calculation as set forth in claim 1, wherein the specific steps of absorption compensation are as follows: Extracting a gather effective time window of the preprocessed seismic gather as a time interval, wherein the time interval represents a closed interval formed by a starting time point corresponding to the ending of a first arrival wave in the gather and a termination time point corresponding to the termination of a deep effective reflection wave in the gather; Dividing the gather effective time window into continuous and non-overlapping sub effective time windows according to a preset time step, respectively obtaining a viscoelastic attenuation coefficient corresponding to each sub effective time window, setting an initial maximum allowable phase rotation amount in the gather effective time window, wherein the initial maximum allowable phase rotation amount is a basic phase rotation threshold value in the seismic wave viscoelastic absorption attenuation compensation process; Establishing a mapping relation between a viscoelastic attenuation coefficient and a phase rotation quantity correction coefficient of each sub-effective time window; If the viscoelastic damping coefficient of each sub-effective time window is smaller than the critical lower limit of the reduction coefficient, inputting the current viscoelastic damping coefficient into a mapping relation between the viscoelastic damping coefficient and the phase rotation quantity correction coefficient of each sub-effective time window, and outputting a phase rotation quantity reduction coefficient; And if the viscoelastic damping coefficient of each sub-effective time window is within a reduced coefficient critical interval, outputting a phase rotation amount correction reference coefficient, wherein the reduced coefficient critical interval represents a closed interval formed by a reduced coefficient critical lower limit and a reduced coefficient critical upper limit.
6. The method for adaptive amplitude compensation of seismic signals based on multi-domain joint parameter resolution as recited in claim 5, wherein said absorption compensation further comprises: If the viscoelastic attenuation coefficient of each sub-effective time window is larger than the critical upper limit of the reduction coefficient, inputting the current viscoelastic attenuation coefficient into a mapping relation between the viscoelastic attenuation coefficient and the phase rotation quantity correction coefficient of each sub-effective time window, and outputting a phase rotation quantity gain coefficient; the maximum allowable phase rotation amount is dynamically adjusted based on the phase rotation amount correction coefficient, specifically: if the phase rotation amount reducing coefficient is output, combining the phase rotation amount reducing coefficient with the maximum allowable phase rotation reference amount to obtain a target maximum allowable phase rotation amount; If the phase rotation amount correction reference coefficient is output, combining the phase rotation amount correction reference coefficient with the maximum allowable phase rotation reference amount to obtain a target maximum allowable phase rotation amount; And if the phase rotation gain coefficient is output, combining the phase rotation gain coefficient with the maximum allowable phase rotation reference quantity to obtain the target maximum allowable phase rotation quantity.
7. The method for adaptive amplitude compensation of seismic signals based on multi-domain joint parameter calculation as set forth in claim 1, wherein the specific steps of adaptive gain compensation are as follows: Extracting the length of an adaptive gain sliding time window of the preprocessed seismic trace set after diffusion compensation and absorption compensation, acquiring an effective signal amplitude mean value of the seismic trace set, and taking the effective signal amplitude mean value as a reference value of a gain effective amplitude threshold value; Presetting a grading critical interval of gain sliding time window length, wherein the grading critical interval comprises a first critical length and a second critical length, the first critical length is a time window length threshold value of an adaptive shallow short reflection event, the second critical length is a time window length threshold value of an adaptive deep long reflection event, and a mapping relation between the gain sliding time window length and an effective amplitude threshold correction coefficient is established; If the length of the gain sliding time window is smaller than or equal to the first critical length, the length of the gain sliding time window is input into a mapping relation between the length of the gain sliding time window and the effective amplitude threshold correction coefficient, an effective amplitude threshold reduction coefficient is output, and the effective amplitude threshold reduction coefficient and the gain effective amplitude threshold are combined to obtain a target gain effective amplitude threshold so as to adapt to the weak signal and noise separation requirement of the shallow short window.
8. The method for adaptive amplitude compensation of seismic signals based on multi-domain joint parameter determination as recited in claim 7, wherein said adaptive gain compensation further comprises: If the length of the window in the sliding of the gain is within a critical length interval, maintaining and outputting an effective amplitude threshold reference coefficient, and combining the effective amplitude threshold reference coefficient with the effective amplitude threshold of the gain to obtain a target effective amplitude threshold of the gain so as to adapt to the effective signal amplitude characteristic of the long window in the middle layer, wherein the critical length interval represents an opening interval formed by a first critical length and a second critical length; If the length of the gain sliding time window is greater than or equal to the second critical length, inputting the length of the gain sliding time window into a mapping relation between the length of the gain sliding time window and an effective amplitude threshold correction coefficient, outputting an effective amplitude threshold gain coefficient, and combining the effective amplitude threshold gain coefficient and the effective amplitude threshold of the gain to obtain a target gain effective amplitude threshold so as to adapt to the strong attenuation signal energy characteristic of the deep long window; After finishing dynamic assignment of the effective gain amplitude threshold values under different gain sliding time window lengths, taking the target effective gain amplitude threshold value after dynamic adjustment as an effective execution judgment basis of the gain lifting operation in the corresponding gain sliding time window, and stopping the gain lifting operation of the window when the signal amplitude in the gain sliding time window is lower than the target effective gain amplitude threshold value corresponding to the window.
9. The method for adaptive amplitude compensation of seismic signals based on multi-domain joint parameter calculation as set forth in claim 1, wherein the specific steps of performing effect verification and iterative optimization on the seismic trace set after the split-stage adaptive amplitude compensation are as follows: extracting effective signal spectrum characteristics of the seismic trace set subjected to staged self-adaptive amplitude compensation to determine the upper limit and the lower limit of the main frequency of the trace set subjected to compensation, wherein the lower limit of the main frequency of the trace set subjected to compensation is the low-frequency cut-off frequency of the effective signal of the trace set subjected to compensation, and the upper limit of the main frequency of the trace set subjected to compensation is the high-frequency cut-off frequency of the effective signal of the trace set subjected to compensation; recording the deviation degree between the main frequency upper limit of the compensated trace set and the main frequency lower limit of the compensated trace set as the actual frequency bandwidth of the compensated trace set, and establishing a mapping relation between the actual frequency bandwidth of the compensated trace set and the energy balance threshold adjustment coefficient of the trace set; If the actual frequency bandwidth of the compensated gather is smaller than or equal to the first critical frequency bandwidth, the current actual frequency bandwidth of the compensated gather is input into a mapping relation between the actual frequency bandwidth of the compensated gather and the gather energy balance threshold adjustment coefficient, the gather energy balance threshold gain coefficient is output, the gather energy balance threshold gain coefficient and the gather energy balance threshold are combined to obtain a target gather energy balance threshold, and the energy balance judgment requirement of the narrow frequency band gather is adapted.
10. The method for adaptive amplitude compensation of seismic signals based on multi-domain joint parameter determination of claim 9, wherein said performing effect verification and iterative optimization further comprises: if the actual frequency bandwidth of the compensated gather is within a critical frequency bandwidth interval, maintaining the current gather energy balance threshold to adapt to the conventional energy balance judging requirement of the middle frequency bandwidth gather, wherein the critical frequency bandwidth interval represents an opening interval between the first critical frequency bandwidth and the second critical frequency bandwidth; If the actual frequency bandwidth of the compensated gather is greater than or equal to the second critical frequency bandwidth, inputting the current actual frequency bandwidth of the compensated gather into a mapping relation between the actual frequency bandwidth of the compensated gather and the gather energy balance threshold adjustment coefficient, outputting a gather energy balance threshold reduction coefficient, and carrying out combination processing on the gather energy balance threshold reduction coefficient and the gather energy balance threshold to obtain a target gather energy balance threshold so as to adapt to the high-resolution energy balance judgment requirement of the broadband gather; And taking the energy balance degree threshold value of the target gather as a judging basis of the energy balance in the compensation effect verification, if the actual energy balance degree of the compensated gather is smaller than the energy balance degree threshold value of the target gather, judging that the energy balance degree reaches the standard, otherwise, judging that the energy balance degree does not reach the standard and starting an iterative optimization flow.

Description

Seismic signal self-adaptive amplitude compensation method based on multi-domain joint parameter calculation Technical Field The invention relates to the technical field of amplitude compensation, in particular to a seismic signal self-adaptive amplitude compensation method based on multi-domain joint parameter calculation. Background Firstly, the prior art relies on single domain parameter estimation (for example, parameter calculation is carried out only based on a time domain or a frequency domain), the relevance of multi-domain information in the seismic wave propagation process is cut off, the physical mechanism of multi-factor coupling actions such as geometric diffusion, viscoelastic absorption, scattering and the like is not fully matched with the seismic wave attenuation, meanwhile, the model is assumed to be too ideal and seriously different from an actual geological propagation scene, particularly, the Q value estimation has obvious multi-resolution due to lack of multi-domain constraints such as time domain time-varying and space domain geometric parameters, the spherical diffusion coefficient calculation ignores actual situations such as complex topography, transverse non-uniformity of a near-surface structure and the like, the problems such as deep energy shortage, energy unbalance of high points and low points of construction and the like still commonly exist after compensation are caused, and the reflection characteristics of underground geological bodies cannot be truly reflected. Secondly, the earth surface consistency compensation technology is insufficient in compensating a frequency band with serious energy loss (such as a deep high-frequency signal), the core root cause is that the compensation parameter inversion is developed only by relying on energy statistics of dominant frequency bands, differential compensation logic is not designed aiming at energy attenuation characteristics of different frequency bands, the energy attenuation mechanism specificity caused by long propagation paths and strong viscoelastic absorption of the deep high-frequency signal is not considered, and finally the problem of transverse energy bead distribution still exists in a compensated section, so that the reservoir prediction accuracy is seriously affected. Meanwhile, the existing compensation technology has the prominent defects of poor compensation stability and easiness in introducing noise or artifacts, the root cause of the problem is caused by inherent defects of algorithm design and the defect of self-adaptive mechanism, the traditional Q compensation adopts an exponential compensation function, the inherent risk of high-frequency range numerical divergence exists, even if an empirical stability factor is added, the stability factor design only depends on manual experience and does not establish a self-adaptive adjustment mechanism based on signal time-frequency characteristics, the attenuation characteristics of all seismic signals in a working area are difficult to adapt, and artifacts such as linear noise, amplitude mutation and the like appear in a compensated section, so that the data continuity is reduced. However, the fixed gain compensation technique lacks signal-noise separation and dynamic gain control logic in algorithm design due to the core physical characteristic of insufficient cognitive signal and noise distribution difference in time-frequency domain, which results in excessive noise compensation in low signal-to-noise ratio region and insufficient compensation in high signal-to-noise ratio region, further deteriorating data quality. Finally, the prior art has the problems of weak adaptability and limited effect under complex geological conditions, the root cause is that the parameter model design is not adapted to the propagation rule of seismic waves under complex geological scenes, the influence of space domain parameters is ignored, in scenes such as complex earth surfaces (such as mountain areas and loess terraces), strong heterogeneous strata (such as igneous rocks and salt domes) and the like, the propagation path of seismic waves is complex, the longitudinal and transverse attenuation characteristics are severely changed, the single parameter model (such as a fixed Q value and uniform earth surface consistency factor) adopted in the prior art cannot be adapted to the multi-dimensional and strong-variation attenuation characteristics, so that compensation failure is caused, meanwhile, in the prestack gather processing process, the traditional method does not cooperatively couple the space domain parameters with the time domain and the frequency domain parameters, so that inherent AVO (Amplitude Versus Offset) characteristics of gathers are destroyed in the compensation process, the amplitude changes along with the offset distance, the AVO characteristics are distorted, the rock character and the accuracy of fluid identification is influenced, and the self-adaptive amplitude