Synthetic record calibration method and device based on three-dimensional model forward modeling

Abstract

The invention belongs to the technical field of petroleum and natural gas exploration and development, and particularly discloses a synthetic record calibration method and device based on three-dimensional model forward modeling. The method comprises the following steps: firstly, obtaining initial depth-time curves of all wells, then carrying out first time shift correction on the initial depth-time curves to obtain all preliminarily calibrated wells, establishing a parameter model, carrying out forward modeling to obtain three-dimensional forward modeling records of different incident angles, and then correcting the depth-time curves after the first time shift according to the obtained three-dimensional forward modeling records to obtain a corrected depth-time curve; fine calibration of target layers of the wells is achieved, fine calibration of all the wells is obtained, and finally the wells with the difference between the fine calibration result and the calibration result for structural interpretation larger than one phase are abandoned; the device comprises a depth-time curve calculation module, a preliminary calibration module, a model construction module, a forward modeling module, a fine calibration module and a screening module. The method is used for improving the quality of well seismic calibration.

Description

Technical Field

[0001] The invention belongs to the technical field of oil and natural gas exploration and development, and in particular relates to a synthetic record calibration method and device based on three-dimensional model forward modeling. Background Art

[0002] Seismic geological calibration is to build a "bridge" between geological data and seismic data, and is the premise and foundation for seismic structural interpretation and reservoir prediction. Through seismic geological calibration, the following problems can be solved: First, the seismic reflection phase axis is given an accurate stratigraphic age meaning, so as to select the target layer for wave group comparison interpretation and structural mapping; second, in areas where the seismic wave group characteristics vary greatly laterally, the drilling layering data is used to control the tracking and comparison of the layer; third, accurate geological layer control and well data control are provided for seismic inversion; fourth, in seismic attribute analysis, a basis is provided for the selection of data analysis time window. According to the different data used, there are mainly three seismic geological calibration methods: VSP (vertical seismic profile) calibration, acoustic wave synthetic record calibration, and outcrop calibration. Among them, the acoustic wave synthetic record calibration method is the most widely used, especially in seismic inversion reservoir prediction, and it is the only seismic geological calibration method currently used.

[0003] Seismic inversion is the most commonly used reservoir prediction method at present. Its process includes extraction of seismic wavelets, calibration of acoustic wave synthetic records, establishment of geological framework model, testing of inversion parameters, sending of inversion operations and checking of inversion results. Compared with structural interpretation, inversion reservoir prediction puts forward higher requirements for calibration of acoustic wave synthetic records, shifting from simple geological layer calibration in structural interpretation to fine reservoir calibration. The current mainstream calibration of acoustic wave synthetic records is still mainly based on manual operation, which is calibrated well by well. Specifically, it is divided into the following steps: First, the initial depth-time relationship is generated using acoustic wave logging curves, and then the related curves such as longitudinal and transverse waves and density are used in combination with seismic wavelets to synthesize one-dimensional seismic records. Finally, according to the wave group characteristics of the synthetic record and the actual seismic data, based on the marker layer, the synthetic record is manually dragged up and down to continuously correct the matching relationship between the synthetic seismic record and the seismic trace near the well to ensure the best match between the synthetic record and the seismic trace near the well, and so on, until the calibration of acoustic wave synthetic records of all wells in the study area is completed.

[0004] For high-density well network blocks, if reservoir prediction is carried out through seismic inversion, the time required to complete the calibration of acoustic wave synthetic records for large quantities of drilling will inevitably increase exponentially compared to other inversion links. At the same time, due to human factors, the quality of well seismic calibration may be uneven, thus affecting the accuracy of subsequent reservoir prediction and even the accuracy of oil and gas reservoir description and remaining oil and gas prediction, and ultimately making it difficult to meet the fast-paced, high-precision production needs of oil fields. Summary of the invention

[0005] In order to solve the above-mentioned deficiencies existing in the known technology, the present invention aims to provide a synthetic record calibration device based on three-dimensional model forward modeling, aiming to save the time spent on completing the calibration of acoustic synthetic records for large-scale drilling, while reducing the influence of human factors and improving the quality of well seismic calibration.

[0006] Another object of the present invention is to provide a synthetic record calibration method based on three-dimensional model forward modeling. The method needs to be implemented using the above-mentioned device. The method saves the time spent on completing the calibration of acoustic wave synthetic records for large-scale drilling, while reducing the influence of human factors and improving the quality of well seismic calibration, thereby improving the accuracy of subsequent reservoir prediction, oil and gas reservoir description, and remaining oil and gas prediction.

[0007] To achieve the above purpose, the technical solution adopted by the present invention is as follows: A synthetic record calibration method based on three-dimensional model forward modeling, the method comprising the following steps performed in sequence: S1, calculate the initial depth-time curves of all wells; S2, performing the first time shift correction on the initial depth-time curves of all wells to obtain the depth-time curves after the first time shift, thereby realizing the preliminary calibration of the target layers of all wells; S3. Using the preliminarily calibrated wells, under the constraints of the stratigraphic framework, a high-precision elastic parameter model in the time domain is established through inter-well difference and extrapolation; S4. Based on the high-precision elastic parameter model in the time domain, carry out forward modeling of the three-dimensional amplitude variation with the offset or the amplitude variation with the incident angle, i.e., three-dimensional AVO / AVA forward modeling, and obtain three-dimensional forward modeling records at different incident angles; S5, selecting the actual partial incident angle stacked 3D seismic data as reference seismic data, and the 3D forward modeling record corresponding to each actual partial incident angle stacked 3D seismic data as adjustment seismic data, obtaining the optimal time shift and maximum correlation 2D plane data of all adjustment seismic data channels, performing a second time shift correction on the deep time curve after the first time shift, obtaining the deep time curve after the second time shift, and realizing fine calibration of the target layers of all wells; The actual partial incident angle stacked 3D seismic data is known from the calibration results for structural interpretation in the early stage, and the calibration results for structural interpretation in the early stage are obtained before executing step S1; S6. Based on the optimal time shift and maximum correlation of each well, the wells whose fine calibration results and the calibration results for structural interpretation differ by more than one phase are discarded to complete the final calibration.

[0008] As a limitation, obtaining the initial deep-time curve in step S1 includes the following steps performed in sequence: a1) Using the acoustic time difference curve and based on the sound velocity conversion formula, the acoustic velocity curves of all wells in the study area are calculated. The velocity of each sampling point on the acoustic velocity curve is the formation velocity. Based on the double-layer travel time formula, the double-layer travel time curves of adjacent sampling points of all wells are obtained in batches; a2) Based on the double-layer travel time curve, the sampling points on the double-layer travel time curve are accumulated and summed to obtain the initial depth time curves of all wells.

[0009] As a second limitation, the process of performing the first time shift correction on the initial deep time curve in step S2 includes the following steps performed in sequence: b1) According to the calibration results of structural interpretation, determine the vertical marker layer and seismic reflection layer interpretation results, and extract the double-layer travel time of the seismic marker layer of all well target layers; Determine the double-layer travel time of the marker layer of the well according to the initial depth-time curve obtained in step S1; b2) The double-layer travel time of the marker layer of the seismic trace where the target layer of the well is located minus the double-layer travel time of the marker layer of the well to obtain the time displacement of each well, which is converted into a time displacement curve with a constant value; b3) Adding the initial depth-time curve obtained in step S1 to the first time-shifted curve to obtain the depth-time curve after the first time-shift, thereby achieving a preliminary calibration of the target layer of each well.

[0010] As a third limitation, in step S3, two or three seismic reflection horizons including a marker layer are used, and a stratigraphic framework is established according to the sedimentary pattern and stratigraphic contact relationship of the study area.

[0011] As a fourth limitation, in step S4, the specific process of carrying out the forward modeling of the three-dimensional amplitude variation with offset AVO or the forward modeling of the amplitude variation with incident angle AVA is as follows: By using a high-precision elastic parameter model in the time domain, forward modeling is carried out in combination with the actual partial incident angle stacking three-dimensional seismic data and the seismic wavelet corresponding to the actual partial incident angle stacking three-dimensional seismic data to obtain four-dimensional gather data, which is then subjected to dimensionality reduction conversion to obtain three-dimensional forward modeling records corresponding to the partial stacking three-dimensional seismic data at different incident angles; among which, the added dimension is the incident angle dimension.

[0012] As a fifth limitation, in step S5, the process of obtaining the optimal time shift and maximum correlation two-dimensional plane data of all adjusted seismic data traces includes the following steps performed in sequence: c1) Open a time window that includes the target layer segment as the analysis time window, include the marker layer in the analysis time window, and set the maximum time shift; c2) For the reference seismic data and the adjusted seismic data within the analysis time window, each time the adjusted seismic data channel is time-shifted, the time-shift amount is recorded and the correlation between the reference seismic data channel and the adjusted seismic data channel within the analysis time window is calculated; Among them, within the maximum time shift range, when the correlation between the reference seismic data track and the adjusted seismic data track in the corresponding analysis time window for a certain time shift is the largest, the time shift is the optimal time shift.

[0013] As a further limitation, in step S5, performing a second time shift correction on the deep time curve after the first time shift includes: selecting the optimal time shift of the seismic marker layer where the target layer segment of all wells is located as the second time shift, converting the secondary time shift into a secondary time shift curve, and then adding it to the deep time curve after the first time shift to obtain the deep time curve after the second time shift.

[0014] The present invention also discloses a synthetic record calibration device based on three-dimensional model forward modeling, which is applied to any one of steps S1-S6, and the device comprises: Deep time curve calculation module, preliminary calibration module, model building module, forward modeling module, fine calibration module and screening module; A depth-time curve calculation module is used to calculate the initial depth-time curve of the well and output it to the preliminary calibration module; A preliminary calibration module is used to perform a first time shift correction on the initial deep-time curve output by the deep-time curve calculation module, obtain a preliminary calibrated well, and output it to the model building module; The model building module, based on the preliminary calibrated wells output by the preliminary calibration module, establishes a high-precision elastic parameter model in the time domain through inter-well difference and extrapolation under the constraint of the stratigraphic grid, and outputs the high-precision elastic parameter model in the time domain to the forward modeling module; The forward modeling module performs forward modeling of three-dimensional amplitude variation with offset or amplitude variation with incident angle based on the high-precision elastic parameter model in the time domain output by the model building module, obtains three-dimensional forward modeling records at different incident angles, and outputs them to the fine calibration module; The fine calibration module obtains the optimal time shift and maximum correlation two-dimensional plane data of all adjusted seismic data channels based on the three-dimensional forward modeling records of different incident angles obtained by the forward modeling module, and performs a second time shift correction on the deep time curve after the first time shift to achieve fine calibration of the target layer of the well; The screening module, based on the wells of each finely calibrated target layer output by the fine calibration module, discards the wells whose calibration results differ from the known calibration results for structural interpretation by more than one phase, and completes the final calibration.

[0015] Due to the adoption of the above technical solution, the present invention has the following beneficial effects compared with the prior art: (1) The method of the present invention saves the time spent on completing the calibration of acoustic wave synthetic records for a large number of drillings, while reducing the influence of human factors, improving the quality of well seismic calibration, and thus improving the accuracy of subsequent reservoir prediction, oil and gas reservoir description, and remaining oil and gas prediction; (2) The method of the present invention effectively improves the quality of well seismic calibration by taking two time-shift corrections, thereby improving the accuracy of subsequent reservoir prediction, oil and gas reservoir description, and remaining oil and gas prediction.

[0016] In summary, the present invention can improve the quality of well seismic calibration, thereby improving the accuracy of subsequent reservoir prediction, oil and gas reservoir description, and remaining oil and gas prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] Figure 1 (a) is a single well velocity curve of Example 1 of the present invention; Figure 1 (b) is a double-layer travel time curve of adjacent sampling points in Example 1 of the present invention; Figure 2 (a) is a double-layer travel time curve of adjacent sampling points in Example 1 of the present invention; Figure 2 (b) is a travel time curve based on a reference reference point according to Example 1 of the present invention; Figure 3 This is a graph of the time-shift data of a single well in Example 1 of the present invention; Figure 4 (a) is a calibration diagram of a synthetic record before a time shift according to Example 1 of the present invention; Figure 4 (b) is a calibration comparison diagram of the synthetic record after a time shift in Example 1 of the present invention; Figure 5 The stratigraphic framework established in Example 1 of the present invention; Figure 6 The density model section in the time domain high-precision elastic parameter model established in Example 1 of the present invention; Figure 7 The longitudinal wave impedance model section in the time domain high-precision elastic parameter model established in Example 1 of the present invention; Figure 8 A cross section of a shear wave impedance model in a time domain high-precision elastic parameter model established in Example 1 of the present invention; Fig. 9 It is the three-dimensional AVO / AVA forward modeling record of Example 1 of the present invention; Fig.10 This is a schematic diagram of data volume correlation analysis in Example 1 of the present invention; Fig.11(a) is a schematic diagram of the optimal time shift amount in Example 1 of the present invention; Fig.11 (b) is a schematic diagram of the maximum correlation coefficient corresponding to the optimal time shift amount in Example 1 of the present invention; Fig.12 This is a calibration comparison diagram of the synthetic record after two time shifts in Example 1 of the present invention. DETAILED DESCRIPTION

[0019] In order to better explain the present invention and facilitate understanding, the preferred embodiments of the present invention are described in detail below through specific implementation modes in conjunction with the accompanying drawings.

[0020] Example 1 Synthetic record calibration method based on 3D model forward modeling This embodiment provides a synthetic log calibration method based on 3D model forward modeling, which has been applied in Block B of the Western Basin of China. This embodiment includes the following steps performed in sequence: S1, calculate the initial depth-time curves of all wells; S2, performing the first time shift correction on the initial depth-time curves of all wells to obtain the depth-time curves after the first time shift, thereby realizing the preliminary calibration of the target layers of all wells; S3. Using the preliminarily calibrated wells, under the constraints of the stratigraphic framework, a high-precision elastic parameter model in the time domain is established through inter-well difference and extrapolation; S4. Based on the high-precision elastic parameter model in the time domain, carry out forward modeling of the three-dimensional amplitude variation with the offset or the amplitude variation with the incident angle, i.e., three-dimensional AVO / AVA forward modeling, and obtain three-dimensional forward modeling records at different incident angles; S5, selecting the actual partial incident angle stacked 3D seismic data as reference seismic data, and the 3D forward modeling record corresponding to each actual partial incident angle stacked 3D seismic data as adjustment seismic data, obtaining the optimal time shift and maximum correlation 2D plane data of all adjustment seismic data channels, performing a second time shift correction on the deep time curve after the first time shift, obtaining the deep time curve after the second time shift, and realizing fine calibration of the target layers of all wells; The actual partial incident angle stacked 3D seismic data is known from the calibration results for structural interpretation in the early stage, and the calibration results for structural interpretation in the early stage are obtained before executing step S1; S6. Based on the optimal time shift and maximum correlation of each well, the wells whose fine calibration results and the calibration results for structural interpretation differ by more than one phase are discarded to complete the final calibration.

[0021] In step S1 of this embodiment, obtaining the initial deep-time curve includes the following steps performed in sequence: a1) Using the acoustic time difference curve and the sound velocity conversion formula, the acoustic velocity curves of all wells in the study area are calculated. The velocity of each sampling point on the acoustic velocity curve is the formation velocity. Based on the double-layer travel time formula, the double-layer travel time curves of adjacent sampling points of all wells are obtained in batches; wherein, the double-layer travel time of each sampling point can be obtained according to the formation velocity and sampling interval of each sampling point; Figure 1 (a) is the velocity curve of a single well. Figure 1 (b) is the double-layer travel time curve of adjacent sampling points. Since the sampling interval of the logging curve is small, usually 0.125m, it can be approximately considered that the formation velocity between adjacent sampling points is constant, and the formation velocity of the sampling point located below is taken as the formation velocity between adjacent sampling points; a2) Based on the double-layer travel time curve, the sampling points on the double-layer travel time curve are accumulated and summed to obtain the initial depth time curve of all wells. Figure 2 (a) is the double-layer travel time curve of adjacent sampling points. Figure 2 (b) is the travel time curve based on the reference point. In the figure, the point before the start point of the acoustic logging is used as the reference point. Figure 2 The sampling points in (a) are accumulated and summed, and the process of accumulation and summation is: the double-layer travel time of the nth, n>1 sampling point relative to the reference point is the accumulation of all sampling points before this sampling point.

[0022] In step S2 of this embodiment, the process of performing the first time shift correction on the initial deep-time curve includes the following steps performed in sequence: b1) According to the calibration results of structural interpretation, determine the interpretation results of the marker layer and seismic reflection layer adjacent to the target layer in the vertical direction, and extract the double-layer travel time T of the marker layer in the seismic trace where the well target layer segment is located. a , the interpretation density in the seismic reflection layer interpretation result is 1 line × 1 trace; According to the initial depth-time curve obtained in step S1, the double-layer travel time T of the well marker layer is determined. b ; b2) T a Subtract T b , and the time displacement of the well is obtained, Figure 3 This is a single well time shift data diagram, where ch73_horizon_time is T a , ch73_top_time is T b , the time shift of each sampling point constitutes a time shift curve; b3) Add the initial depth-time curve obtained in step S1 to the time-shifted curve to obtain the depth-time curve after the time shift, so as to achieve the preliminary calibration of the target layer of a single well. Figure 4 (a) is a calibration diagram of synthetic records before time shift. Figure 4(b) is a calibration comparison diagram of the synthetic record after a time shift, where the horizontal axis corresponds to the plane position and the vertical axis corresponds to the time.

[0023] In step S3 of this embodiment, two or three seismic reflection horizons including the marker layer are used to establish a stratigraphic framework according to the sedimentary pattern and stratigraphic contact relationship of the study area. The seismic reflection horizons are obtained from the calibration results of the structural interpretation before step S1. Figure 5 The stratigraphic framework is established. The five vertical lines are seismic interpretation horizons. Different colors in the figure represent different stratigraphic sections. Figure 6 To establish the rock density model profile in the time domain high-precision elastic parameter model, Figure 7 To establish the rock longitudinal wave impedance model profile in the time domain high-precision elastic parameter model, Figure 8 This is the rock shear wave impedance model profile in the established time domain high-precision elastic parameter model.

[0024] In step S4 of this embodiment, the specific process of carrying out the forward modeling of the three-dimensional amplitude variation with offset AVO or the forward modeling of the amplitude variation with incident angle AVA is as follows: Using the high-precision elastic parameter model in the time domain, combined with the actual partial incident angle stacking 3D seismic data and the seismic wavelet corresponding to the actual partial incident angle stacking 3D seismic data, forward modeling is carried out to obtain 4D gather data, and the 4D gather data is converted into a dimensionality-reducing method to obtain 3D forward modeling records corresponding to the partial stacking 3D seismic data at different incident angles; wherein, the added dimension is the incident angle dimension; Fig. 9 This is a three-dimensional AVO / AVA forward modeling record, in which the near path, middle path, and far path correspond to different incident angles.

[0025] The process of obtaining the optimal time shift and maximum correlation two-dimensional plane data of all adjusted seismic data tracks in step S5 of this embodiment includes the following steps performed in sequence: c1) Open a time window that includes the target layer segment as the analysis time window, include the marker layer in the analysis time window, and set the maximum time shift; c2) For the reference seismic data and the adjusted seismic data in the analysis time window, each time the adjusted seismic data track is time-shifted, the time shift amount is recorded, and the correlation between the reference seismic data track and the adjusted seismic data track in the analysis time window is calculated, where a downward shift relative to the initial position is a positive value, and an upward shift is a negative value; Within the maximum time shift range, when the correlation between the reference seismic data track and the adjusted seismic data track in the corresponding analysis time window is the largest for a certain time shift, the time shift is the optimal time shift. Fig.10 This is a schematic diagram of data volume correlation analysis. The reference layer is the marker layer. The correlation corresponding to time T3 is the largest, so the time shift of T3 is the optimal time shift. Fig.11(a) is the optimal time shift. Warm colors indicate large time shifts, while cold colors indicate small time shifts. Fig.11 (b) is the maximum correlation coefficient corresponding to the optimal time shift. Warm colors indicate large correlation coefficients, while cold colors indicate small correlation coefficients.

[0026] The second time shift correction of the deep time curve after the first time shift includes: selecting the best time shift of the seismic trace marker layer where the target layer segment of all wells is located as the second time shift, converting the secondary time shift into a secondary time shift curve, and then adding it to the deep time curve after the first time shift to obtain the deep time curve after the second time shift. Fig.12 This is a calibration comparison chart of the synthetic records after the second time shift. In the figure, warm colors indicate high correlation, and cold colors indicate low correlation.

[0027] After obtaining the optimal time shift and maximum correlation of all wells and realizing fine calibration of the target layers of all wells, the wells whose fine calibration results differ from the known calibration results for structural interpretation by more than one phase are discarded according to the optimal time shift and maximum correlation of each well to complete the final calibration.

[0028] This embodiment completes the fine calibration of 2458 wells in Block B of the Western Basin of China. If the traditional artificial synthetic record calibration method is used, it is estimated to take 15-20 working days to complete the fine reservoir calibration of all wells, and due to human factors, the quality of calibration may be uneven, affecting the accuracy of subsequent seismic inversion. After adopting this embodiment, while improving the calibration quality, it only takes 3-4 working days to complete, which greatly improves the efficiency of well seismic calibration to 70%, which provides strong support for the fine description of clastic oil and gas reservoirs and the rapid deployment of well locations.

[0029] Embodiment 2 Synthetic record calibration device based on 3D model forward modeling This embodiment is applied to Embodiment 1, and includes a deep-time curve calculation module, a preliminary calibration module, a model building module, a forward modeling module, a fine calibration module, and a screening module.

[0030] A depth-time curve calculation module is used to calculate the initial depth-time curve of the well and output it to the preliminary calibration module; A preliminary calibration module is used to perform a first time shift correction on the initial deep-time curve output by the deep-time curve calculation module, obtain a preliminary calibrated well, and output it to the model building module; The model building module, based on the preliminary calibrated wells output by the preliminary calibration module, establishes a high-precision elastic parameter model in the time domain through inter-well difference and extrapolation under the constraint of the stratigraphic grid, and outputs the high-precision elastic parameter model in the time domain to the forward modeling module; The forward modeling module performs forward modeling of three-dimensional amplitude variation with offset or amplitude variation with incident angle based on the high-precision elastic parameter model in the time domain output by the model building module, obtains three-dimensional forward modeling records at different incident angles, and outputs them to the fine calibration module; The fine calibration module obtains the optimal time shift and maximum correlation two-dimensional plane data of all adjusted seismic data channels based on the three-dimensional forward modeling records of different incident angles obtained by the forward modeling module, and performs a second time shift correction on the deep time curve after the first time shift to achieve fine calibration of the target layer of the well; The screening module, based on the wells of each finely calibrated target layer output by the fine calibration module, discards the wells whose calibration results differ from the known calibration results for structural interpretation by more than one phase, and completes the final calibration.

Claims

1. A synthetic record calibration method based on three-dimensional model forward modeling, characterized in that: The method comprises the following steps performed in sequence: S1, calculate the initial depth-time curves of all wells; S2, performing the first time shift correction on the initial depth-time curves of all wells to obtain the depth-time curves after the first time shift, thereby realizing the preliminary calibration of the target layers of all wells; S3. Using the preliminarily calibrated wells, under the constraints of the stratigraphic framework, a high-precision elastic parameter model in the time domain is established through inter-well difference and extrapolation; S4. Based on the high-precision elastic parameter model in the time domain, carry out forward modeling of the three-dimensional amplitude variation with the offset or the amplitude variation with the incident angle, i.e., three-dimensional AVO / AVA forward modeling, and obtain three-dimensional forward modeling records at different incident angles; S5, selecting the actual partial incident angle stacked 3D seismic data as reference seismic data, and the 3D forward modeling record corresponding to each actual partial incident angle stacked 3D seismic data as adjustment seismic data, obtaining the optimal time shift and maximum correlation 2D plane data of all adjustment seismic data channels, performing a second time shift correction on the deep time curve after the first time shift, obtaining the deep time curve after the second time shift, and realizing fine calibration of the target layers of all wells; The actual partial incident angle stacked 3D seismic data is known from the calibration results for structural interpretation in the early stage, and the calibration results for structural interpretation in the early stage are obtained before executing step S1; S6. Based on the optimal time shift and maximum correlation of each well, the wells whose fine calibration results and the calibration results for structural interpretation differ by more than one phase are discarded to complete the final calibration.

2. The synthetic record calibration method based on three-dimensional model forward modeling according to claim 1 is characterized in that: The step S1 of obtaining the initial deep-time curve includes the following steps performed in sequence: a1) Using the acoustic time difference curve and based on the sound velocity conversion formula, the acoustic velocity curves of all wells in the study area are calculated. The velocity of each sampling point on the acoustic velocity curve is the formation velocity. Based on the double-layer travel time formula, the double-layer travel time curves of adjacent sampling points of all wells are obtained in batches; a2) Based on the double-layer travel time curve, the sampling points on the double-layer travel time curve are accumulated and summed to obtain the initial depth time curves of all wells.

3. The synthetic record calibration method based on three-dimensional model forward modeling according to claim 1 or 2, characterized in that: The process of performing the first time shift correction on the initial deep time curve in step S2 comprises the following steps performed in sequence: b1) According to the calibration results of structural interpretation, determine the vertical marker layer and seismic reflection layer interpretation results, and extract the double-layer travel time of the seismic marker layer of all well target layers; Determine the double-layer travel time of the marker layer of the well according to the initial depth-time curve obtained in step S1; b2) The double-layer travel time of the marker layer of the seismic trace where the target layer of the well is located minus the double-layer travel time of the marker layer of the well to obtain the time displacement of each well, which is converted into a time displacement curve with a constant value; b3) Adding the initial depth-time curve obtained in step S1 to the first time-shifted curve to obtain the depth-time curve after the first time-shift, thereby achieving a preliminary calibration of the target layer of each well.

4. The synthetic record calibration method based on three-dimensional model forward modeling according to claim 1 or 2, characterized in that: In step S3, two or three seismic reflection horizons including the marker layer are used, and a stratigraphic framework is established according to the sedimentary pattern and stratigraphic contact relationship of the study area.

5. The synthetic record calibration method based on three-dimensional model forward modeling according to claim 1 or 2, characterized in that: In step S4, the specific process of carrying out the forward modeling of the three-dimensional amplitude variation with offset AVO or the forward modeling of the amplitude variation with incident angle AVA is as follows: By using a high-precision elastic parameter model in the time domain, forward modeling is carried out in combination with the actual partial incident angle stacking three-dimensional seismic data and the seismic wavelet corresponding to the actual partial incident angle stacking three-dimensional seismic data to obtain four-dimensional gather data, which is then subjected to dimensionality reduction conversion to obtain three-dimensional forward modeling records corresponding to the partial stacking three-dimensional seismic data at different incident angles; among which, the added dimension is the incident angle dimension.

6. The synthetic record calibration method based on three-dimensional model forward modeling according to claim 1 or 2, characterized in that: In step S5, the process of obtaining the optimal time shift and maximum correlation two-dimensional plane data of all adjusted seismic data traces includes the following steps performed in sequence: c1) Open a time window that includes the target layer segment as the analysis time window, include the marker layer in the analysis time window, and set the maximum time shift; c2) For the reference seismic data and the adjusted seismic data within the analysis time window, each time the adjusted seismic data channel is time-shifted, the time-shift amount is recorded and the correlation between the reference seismic data channel and the adjusted seismic data channel within the analysis time window is calculated; Among them, within the maximum time shift range, when the correlation between the reference seismic data track and the adjusted seismic data track in the corresponding analysis time window for a certain time shift is the largest, the time shift is the optimal time shift.

7. The synthetic record calibration method based on three-dimensional model forward modeling according to claim 6 is characterized in that: In step S5, performing a second time shift correction on the deep time curve after the first time shift includes: selecting the optimal time shift of the seismic marker layer where the target layer segment of all wells is located as the second time shift, converting the secondary time shift into a secondary time shift curve, and then adding it to the deep time curve after the first time shift to obtain the deep time curve after the second time shift.

8. A synthetic record calibration device based on three-dimensional model forward modeling, applied to any one of steps S1-S6, characterized in that: The device comprises: a deep time curve calculation module, a preliminary calibration module, a model building module, a forward modeling module, a fine calibration module and a screening module; A depth-time curve calculation module is used to calculate the initial depth-time curve of the well and output it to the preliminary calibration module; A preliminary calibration module is used to perform a first time shift correction on the initial deep-time curve output by the deep-time curve calculation module, obtain a preliminary calibrated well, and output it to the model building module; The model building module, based on the preliminary calibrated wells output by the preliminary calibration module, establishes a high-precision elastic parameter model in the time domain through inter-well difference and extrapolation under the constraint of the stratigraphic grid, and outputs the high-precision elastic parameter model in the time domain to the forward modeling module; The forward modeling module performs forward modeling of three-dimensional amplitude variation with offset or amplitude variation with incident angle based on the high-precision elastic parameter model in the time domain output by the model building module, obtains three-dimensional forward modeling records at different incident angles, and outputs them to the fine calibration module; The fine calibration module obtains the optimal time shift and maximum correlation two-dimensional plane data of all adjusted seismic data channels based on the three-dimensional forward modeling records of different incident angles obtained by the forward modeling module, and performs a second time shift correction on the deep time curve after the first time shift to achieve fine calibration of the target layer of the well; The screening module, based on the wells of each finely calibrated target layer output by the fine calibration module, discards the wells whose calibration results differ from the known calibration results for structural interpretation by more than one phase, and completes the final calibration.

Images