Reservoir parameter dynamic inversion method and device, electronic equipment and medium

Abstract

The invention relates to a reservoir parameter dynamic inversion method and apparatus, an electronic device and a medium. The method comprises the steps of obtaining an initial reservoir parameter of a current period in a time-lapse seismic inversion process; the initial reservoir parameters are generated by adopting a reservoir parameter result obtained by inversion in the previous period; performing next parameter state prediction on the initial reservoir parameters to obtain reservoir parameter prediction data; determining a theoretical seismic record corresponding to the reservoir parameter prediction data; and according to the error between the theoretical seismic record and the actual seismic record, obtaining an update quantity for the reservoir parameters of the current period by fusing an unscented Kalman filter algorithm with an early-stage result to perform iterative update, and determining the target reservoir parameters with the error smaller than a set threshold value as a reservoir parameter result obtained by inversion of the current period. According to the method, the inversion of the physical property state of the high-frequency time-lapse seismic data can be realized, and the inversion result can well reflect the dynamic change of reservoir parameters along with development.

Description

Technical Field

[0001] The present disclosure relates to the technical field of seismic exploration data interpretation, and in particular, to a method, device, electronic device and medium for dynamic inversion of reservoir parameters. Background Art

[0002] Time-lapse seismic technology is to repeat seismic surveys at different times in the same study area. By performing four-dimensional seismic processing, imaging, and differential analysis, attribute extraction, inversion on time-lapse seismic data and supplemented with visualization technology to describe the changes in reservoir physical properties, establish the direct relationship between seismic response differences and factors affecting oil and gas reservoirs, so as to identify the dynamic change information of underground reservoirs and guide the efficient development of oil and gas reservoirs.

[0003] In the process of implementing the concept of the present disclosure, the inventors found that there are at least the following technical problems in the related art: With the progress of seismic data acquisition means, such as in-well or cross-well seismic data acquisition based on distributed fiber optic sensor technology, which has the advantages of high frequency (high density in the time dimension), full well section, low cost, high efficiency, etc., a lot of high-frequency time-lapse seismic data (which can also be described as high-frequency time-lapse seismic data) is generated, which puts higher requirements on the real-time prediction of reservoir dynamic parameters; however, the corresponding time accuracy of the existing related technologies is insufficient, and there are few reservoir parameter inversion technical means compatible with such high-frequency time-lapse seismic data. Summary of the Invention

[0004] In order to solve the above technical problems or at least partially solve the above technical problems, embodiments of the present disclosure provide a method, device, electronic device and medium for dynamic inversion of reservoir parameters.

[0005] In a first aspect, an embodiment of the present disclosure provides a method for dynamic inversion of reservoir parameters. The method includes: obtaining the initial reservoir parameters in the current period during the time-lapse seismic inversion process; the initial reservoir parameters are generated by using the reservoir parameter results obtained from the inversion in the previous period; predicting the next parameter state of the initial reservoir parameters to obtain reservoir parameter prediction data; determining the theoretical seismic record corresponding to the reservoir parameter prediction data; and iteratively updating the reservoir parameters in the current period according to the error between the theoretical seismic record and the actual seismic record, and determining the target reservoir parameters with an error less than a set threshold as the reservoir parameter results obtained from the inversion in the current period.

[0006] In some embodiments, the above-mentioned target reservoir parameters are obtained based on the unscented Kalman filter algorithm. Among them, predicting the next parameter state of the above-mentioned initial reservoir parameters to obtain reservoir parameter prediction data, including: constructing a reservoir parameter set corresponding to the above-mentioned initial reservoir parameters based on the unscented transform; the mean and variance of each group of reservoir parameters in the above-mentioned reservoir parameter set are the same as the mean and variance corresponding to the distribution state of the above-mentioned initial reservoir parameters; calculating the product of a preset reservoir parameter model update amount and a preset random variable, and using the above-mentioned product as a state update amount; calculating the sum of the above-mentioned reservoir parameter set and the above-mentioned state update amount, and using the above-mentioned sum as reservoir parameter prediction data.

[0007] In some embodiments, constructing a reservoir parameter set corresponding to the above-mentioned initial reservoir parameters based on the unscented transform includes:

[0008] Describing the initial reservoir parameter X by using a set of random variables corresponding to the reservoir parameter result obtained by the inversion in the previous period, the number of random variables is L, and the variance is P;

[0009] According to the preset unscented transform parameters α, k, and λ, calculating 2L + 1 derivative values corresponding to the above-mentioned initial reservoir parameter X, and the 2L + 1 groups of reservoir parameters corresponding to the above-mentioned derivative values constitute the above-mentioned reservoir parameter set Xs, and the reservoir parameter set Xs satisfies the following expression:

[0010] Xs (0) = X,

[0011]

[0012]

[0013] λ = α 2 (L + k) - L,

[0014] where i represents the serial number of the reservoir parameter derivative value, and the value range of i is 0 to 2L; Xs (0) represents the first derivative value in the reservoir parameter set Xs, and Xs (i) represents the (i + 1)-th derivative value in the reservoir parameter set Xs.

[0015] In some embodiments, the above-mentioned reservoir parameter prediction data satisfies the following expression:

[0016] X1 = Xs + rs·Δx,

[0017] where X1 represents the reservoir parameter prediction data, which includes a total of 2L + 1 prediction data; Δx represents a preset reservoir parameter model update amount; rs represents a preset random variable;

[0018] The mean of the above-mentioned reservoir parameter prediction data satisfies the following expression:

[0019]

[0020] Among them, represents the mean value of the reservoir parameter prediction data; W m (i) represents the first weight coefficient for the (i + 1)-th data; X1 (i) represents the (i + 1)-th prediction data in the reservoir parameter prediction data;

[0021] The variance of the above-mentioned reservoir parameter prediction data satisfies the following expression:

[0022]

[0023]

[0024] Among them, xdiv (i) represents the difference between the value of the (i + 1)-th prediction data and the mean value of the reservoir parameter prediction data; (xdiv (i) ) T represents the transpose of xdiv (i) ; Wc (i) represents the second weight coefficient for the (i + 1)-th data; Q represents the prediction noise in the state transition process, which is a given value;

[0025] The first weight coefficient W m (i) and the second weight coefficient W c (i) are expressed as follows:

[0026] W m (0) = λ / (L + λ),

[0027]

[0028] W m (i) = W c (i) = λ / 2(L + λ), i = 1 to 2L,

[0029] Among them, β is a preset unscented transform parameter.

[0030] In some embodiments, determining the theoretical seismic record corresponding to the above-mentioned reservoir parameter prediction data includes: calculating the theoretical seismic record corresponding to the above-mentioned reservoir parameter prediction data based on the convolution model.

[0031] In some embodiments, the convolution model of the above-mentioned theoretical seismic record satisfies the following expression:

[0032] S = W * D * X1,

[0033] Among them, S represents the theoretical seismic record; W represents the wavelet matrix; D represents the differential matrix; X1 represents the predicted data of reservoir parameters;

[0034] The mean value of the above-mentioned theoretical seismic record satisfies the following expression:

[0035]

[0036] Among them, represents the mean value of the theoretical seismic record; W m (i) represents the first weight coefficient for the (i + 1)-th measurement data; S (i) represents the (i + 1)-th measurement data in the theoretical seismic record;

[0037] The variance of the above-mentioned theoretical seismic record satisfies the following expression:

[0038]

[0039]

[0040] Among them, Ps represents the variance of the theoretical seismic record; sdiv (i) represents the difference between the value of the (i + 1)-th measurement data and the mean value of the theoretical seismic record; (sdiv (i) ) T represents the transpose of sdiv (i) ; Wc (i) represents the second weight coefficient for the (i + 1)-th measurement data; R represents the measurement noise.

[0041] In some embodiments, the above-mentioned target reservoir parameters are obtained based on the unscented Kalman filter algorithm;

[0042] The error between the above-mentioned theoretical seismic record and the actual seismic record satisfies the following expression:

[0043] error = ∑(dif) 2 = ∑(Sr - S) 2 ,

[0044] Among them, error represents the error between the theoretical seismic record S and the actual seismic record Sr;

[0045] Among them, by calculating the Kalman gain coefficient, the reservoir parameters are iteratively updated and the variance is updated based on the above-mentioned Kalman gain coefficient;

[0046] The expression of the above-mentioned Kalman gain coefficient is as follows:

[0047] K = Psx · Ps -1 ,

[0048]

[0049]

[0050] Among them, K represents the Kalman gain coefficient, and Psx represents the covariance between the predicted data and the measured data; Ps represents the variance of the theoretical seismic record; sdiv (i) represents the difference between the value of the (i + 1)-th measured data and the mean value of the theoretical seismic record; xdiv (i) represents the difference between the value of the (i + 1)-th predicted data point and the mean value of the reservoir parameter prediction data;

[0051] X 迭代后 = X 迭代前 + K(Sr - S),

[0052] P 迭代后 = P 1 - K·(Psx) T ,

[0053] Among them, X 迭代后 represents the reservoir parameter after iterative update; X 迭代前 represents the reservoir parameter before iterative update, specifically the reservoir parameter prediction data; P 迭代后 represents the variance of the reservoir parameter after iterative update; P 1 represents the variance of the reservoir parameter prediction data; (Psx) T represents the transpose of the covariance Psx.

[0054] Second, an embodiment of the present disclosure provides an apparatus for dynamic inversion of reservoir parameters. The above apparatus includes: a parameter acquisition module, a parameter prediction module, a theoretical measurement module, and an iteration module. The above parameter acquisition module is used to acquire the initial reservoir parameters in the current period during the time-lapse seismic inversion process; the above initial reservoir parameters are generated by using the reservoir parameter results obtained from the previous inversion. The above parameter prediction module is used to predict the next parameter state of the above initial reservoir parameters to obtain reservoir parameter prediction data. The above theoretical measurement module is used to determine the theoretical seismic record corresponding to the above reservoir parameter prediction data. The above iteration module is used to iteratively update the reservoir parameters in the current period according to the error between the above theoretical seismic record and the actual seismic record, and determine the target reservoir parameters with an error less than the set threshold as the reservoir parameter results obtained from the current period inversion.

[0055] In a third aspect, an embodiment of the present disclosure provides an electronic device. The above-mentioned electronic device includes a processor, a communication interface, a memory, and a communication bus. Among them, the processor, the communication interface, and the memory complete communication with each other through the communication bus; the memory is used to store a computer program; when the processor executes the program stored on the memory, it implements the method for dynamic inversion of reservoir parameters as described above.

[0056] In a fourth aspect, an embodiment of the present disclosure provides a computer-readable storage medium. A computer program is stored on the above-mentioned computer-readable storage medium, and when the above-mentioned computer program is executed by a processor, it implements the method for dynamic inversion of reservoir parameters as described above.

[0057] The above technical solutions provided by the embodiments of the present disclosure have at least some or all of the following advantages:

[0058] By obtaining the initial reservoir parameters in the current period during the time-lapse seismic inversion process; the above-mentioned initial reservoir parameters are generated using the reservoir parameter results obtained from the inversion in the previous period; predicting the next parameter state of the above-mentioned initial reservoir parameters to obtain reservoir parameter prediction data; determining the theoretical seismic record corresponding to the above-mentioned reservoir parameter prediction data; and iteratively updating the reservoir parameters in the current period according to the error between the above-mentioned theoretical seismic record and the actual seismic record, and determining the target reservoir parameters with an error less than a set threshold as the reservoir parameter results obtained from the inversion in the current period, it is possible to adapt to high-frequency time-lapse seismic data to achieve dynamic inversion of reservoir parameters. As the number of time-lapse periods increases, the reservoir changes in each period contribute to the final reservoir parameter inversion results, realizing the optimization of the overall inversion.

[0059] In some embodiments, the above-mentioned target reservoir parameters are obtained based on the unscented Kalman filter algorithm. By fusing two sets of inversion data from different periods, an optimal estimate of the reservoir parameters is made. These two sets of data are the prediction data and the measurement data respectively. The prediction data is the previous inversion result plus state transformation and update, and the measurement data is the theoretical seismic record (which can also be described as the synthetic seismic record) obtained through the convolution model. Unscented transformation is performed on the prediction data and the measurement data to obtain the posterior probability density of the state transformation. Under the framework of the Kalman filter algorithm, the physical properties parameters of the reservoir in the current period are iteratively updated. When the error between the synthetic seismic record of the reservoir parameters and the actual seismic record meets a certain standard (for example, the error is less than the set threshold), the corresponding target reservoir parameters are output as the inversion results in the current period. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The accompanying drawings here are incorporated into the specification and constitute a part of this specification, showing embodiments consistent with the present disclosure, and are used together with the specification to explain the principles of the present disclosure.

[0061] To more clearly illustrate the technical solutions in the embodiments of the present disclosure or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or related technologies. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0062] Figure 1 Schematically shows a flowchart of a method for dynamic inversion of reservoir parameters according to an embodiment of the present disclosure;

[0063] Figure 2 Schematically shows a detailed implementation flowchart of step S120 according to an embodiment of the present disclosure;

[0064] Figure 3 Schematically shows a schematic diagram of the implementation process of dynamic inversion of reservoir parameters according to an embodiment of the present disclosure;

[0065] Figure 4 Schematically shows a schematic diagram of the process of inverting the reservoir parameters of the current period according to an embodiment of the present disclosure, where (a) is the true value of wave impedance, (b) is the theoretical seismic record, (c) is the initial inversion model, and (d) is the inversion output result of the reservoir parameters of the current period;

[0066] Figure 5 Schematically shows a schematic diagram of theoretical models of reservoir parameters for multiple periods, where (a) is the reservoir parameter model of the basic data, (b) is the reservoir parameter model of the first monitoring period, (c) is the reservoir parameter model of the second monitoring period, (d) is the reservoir parameter model of the third monitoring period, (e) is the reservoir parameter model of the fourth monitoring period, and (f) is the reservoir parameter model of the fifth monitoring period;

[0067] Figure 6 Schematically shows a schematic diagram of the effect of inverting the change process of reservoir parameters for multiple periods based on the unscented Kalman filter algorithm according to an embodiment of the present disclosure, where (a) is the basic data, (b) is the inversion output result of the first monitoring period, (c) is the inversion output result of the second monitoring period, (d) is the inversion output result of the third monitoring period, (e) is the inversion output result of the fourth monitoring period, and (f) is the inversion output result of the fifth monitoring period;

[0068] Figure 7 Schematically shows a structural block diagram of an apparatus for dynamic inversion of reservoir parameters according to an embodiment of the present disclosure;

[0069] Figure 8 Schematically shows a structural block diagram of an electronic device provided by an embodiment of the present disclosure. Detailed implementation manners

[0070] To make the objectives, technical solutions, and advantages of the embodiments of the present disclosure clearer, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present disclosure. Apparently, the described embodiments are some but not all of the embodiments of the present disclosure. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present disclosure without creative efforts shall fall within the scope of protection of the present disclosure.

[0071] The first exemplary embodiment of the present disclosure provides a method for dynamic inversion of reservoir parameters. The method of this embodiment can be executed by an electronic device with computing capabilities.

[0072] Figure 1 The flowchart of the method for dynamic inversion of reservoir parameters according to an embodiment of the present disclosure is schematically shown. Figure 3 The schematic diagram of the implementation process of the dynamic inversion of reservoir parameters according to an embodiment of the present disclosure is schematically shown.

[0073] Referring to Figure 1 and Figure 3 As shown, the method for dynamic inversion of reservoir parameters provided by the embodiments of the present disclosure includes the following steps: S110, S120, S130, and S140.

[0074] In step S110, obtain the initial reservoir parameters of the current period during the time-lapse seismic inversion process; the above initial reservoir parameters are generated using the reservoir parameter results obtained from the previous inversion.

[0075] In step S120, predict the next parameter state of the above initial reservoir parameters to obtain reservoir parameter prediction data.

[0076] In some embodiments, the unscented Kalman filter algorithm is used to iteratively update the reservoir parameters to obtain the target reservoir parameters, and the specific implementation process of the unscented Kalman filter algorithm is applied to steps S120 to S140.

[0077] For ease of understanding, the Kalman Filter algorithm, the Extended Kalman Filter (EKF), and the Unscented Kalman Filter (UKF) algorithm are introduced below first.

[0078] The Kalman Filter algorithm is an algorithm for optimal estimation of the system state by fusing the probabilities of two sets of random data that satisfy the Gaussian distribution from different sources in the system time series, and these two sets of data are described by the state function and the measurement function. The Kalman filter is applicable to linear systems and is commonly used for real-time optimal estimation of time series, such as in the fields of navigation, guidance, and communication.

[0079] For a non - linear system, the Extended Kalman Filter (EKF) performs Taylor series expansion on the state function and the measurement function and omits the high - order terms to obtain an approximate linearized model, and then uses the Kalman filter to implement the optimal estimation of the target. However, the EKF algorithm also has deficiencies. On the one hand, the linearization loses the high - order terms, bringing certain errors. At the same time, when the linearization assumption does not hold, it will lead to a decline in algorithm performance or even non - convergence. On the other hand, when the Taylor series of the state function and the measurement function of some application models are expanded, its Jacobian Matrix is not easy to implement, thus increasing the computational complexity.

[0080] The Unscented Kalman Filter (UKF) algorithm is based on the framework of the linear Kalman filter and uses the Unscented Transform (UT) to handle the non - linear transfer problems of the means and covariances of the state function and the measurement function. The UKF algorithm approximates the probability density distribution of the non - linear function and uses a series of deterministic samples to approximate the posterior probability density of the state. It does not require linear approximation of the state function and the measurement function, does not require determination of the Jacobian matrix, and does not ignore high - order terms, overcoming the defects of low estimation accuracy and poor stability of the EKF.

[0081] Embodiments of the present disclosure integrate the Unscented Kalman Filter algorithm into the dynamic inversion process of reservoir parameters, which not only effectively improves the real - time performance of reservoir parameter inversion for high - frequency time - lapse seismic data, but also helps to expand the application scenarios and is not limited to linear systems. By using the Unscented Kalman Filter algorithm to invert the time - lapse seismic data with high density in time, the purpose and effect of real - time dynamic prediction of reservoir parameters are achieved. While considering the time dimension, it also avoids the deficiencies such as the Kalman filter being only applicable to linear systems.

[0082] Figure 2 Schematically shows a detailed implementation flowchart of step S120 according to an embodiment of the present disclosure.

[0083] In some embodiments, referring to Figure 2 As shown, in the above step S120, predicting the next parameter state of the above initial reservoir parameters to obtain reservoir parameter prediction data includes the following sub - steps: S210, S220, and S230.

[0084] In step S210, based on the unscented transform, construct a reservoir parameter set corresponding to the above initial reservoir parameters; the mean and variance of each group of reservoir parameters in the above reservoir parameter set are the same as the mean and variance corresponding to the distribution state of the above initial reservoir parameters.

[0085] In some embodiments, based on the unscented transform, a reservoir parameter set corresponding to the above initial reservoir parameters is constructed, including:

[0086] Describe the initial reservoir parameters using a set of random variables corresponding to the reservoir parameter results obtained from the inversion of the previous period. The number of random variables is L, and the variance is P; hereinafter, the initial reservoir parameters are represented by X.

[0087] According to the preset unscented transform parameters α, k, and λ, calculate 2L + 1 derivative values corresponding to the above initial reservoir parameters X. The 2L + 1 sets of reservoir parameters corresponding to the derivative values constitute the above reservoir parameter set Xs. Each set of reservoir parameters (here corresponding to the derivative value) is actually in vector form and contains multiple data points. The reservoir parameter set is a set form containing multiple vectors.

[0088] The reservoir parameter set Xs satisfies the following expression:

[0089] Xs( 0 ) = X, (1)

[0090]

[0091]

[0092] λ = α 2 (L + k) - L, (4)

[0093] where i represents the serial number of the reservoir parameter derivative value, and the value range of i is 0 to 2L; Xs (0) represents the first derivative value in the reservoir parameter set Xs, and Xs (i) represents the (i + 1)-th derivative value in the reservoir parameter set Xs.

[0094] In step S220, calculate the product of the preset reservoir parameter model update amount and the preset random variable, and use the product as the state update amount.

[0095] In step S230, calculate the sum of the above reservoir parameter set and the above state update amount, and use the sum as the reservoir parameter prediction data.

[0096] In some embodiments, the above reservoir parameter prediction data satisfies the following expression:

[0097] X1 = Xs + rs·Δx, (5)

[0098] where X1 represents the reservoir parameter prediction data, which includes a total of 2L + 1 prediction data; Δx represents the preset reservoir parameter model update amount; rs represents the preset random variable. For example, in Figure 3 it is represented by the transformation matrix F to update the physical property state corresponding to the reservoir parameters.

[0099] The mean value of the above reservoir parameter prediction data satisfies the following expression:

[0100]

[0101] Wherein, represents the mean value of the reservoir parameter prediction data; W m (i) represents the first weight coefficient for the (i + 1)-th data; X1 (i) represents the (i + 1)-th prediction data in the reservoir parameter prediction data;

[0102] The variance P of the above reservoir parameter prediction data 1 (which can be described as the prediction variance) satisfies the following expression:

[0103]

[0104]

[0105] Wherein, xdiv (i) represents the difference between the value of the (i + 1)-th prediction data and the mean value of the reservoir parameter prediction data; (xdiv (i) ) T represents the transpose of xdiv (i) ; Wc (i) represents the second weight coefficient for the (i + 1)-th data; Q represents the prediction noise in the state transition process and is a given value;

[0106] The expressions of the first weight coefficient W m (i) and the second weight coefficient W c (i) are as follows:

[0107] W m (0) = λ / (L + λ), (9)

[0108]

[0109] W m (i) = W c (i) = λ / 2(L + λ), i = 1 to 2L, (11)

[0110] Wherein, β is a preset unscented transform parameter.

[0111] In some embodiments, the common value ranges of the above unscented transform parameters α, β, and k are: k ≥ 0, 0 < α ≤ 1, β = 2.

[0112] The above process describes the unscented transform of the prediction data.

[0113] In step S130, determine the theoretical seismic record corresponding to the above reservoir parameter prediction data.

[0114] In some embodiments, in the above step S130, determining the theoretical seismic record corresponding to the above reservoir parameter prediction data includes: calculating the theoretical seismic record corresponding to the prediction data of the new reservoir parameter set based on the convolution model. The above theoretical seismic record is also called the synthetic seismic record.

[0115] The convolution model is a model for making theoretical seismic records. It assumes that each seismic record is composed of the convolution of the seismic wavelet and the reflection coefficients of each layer of the underground model, and random noise can be added if necessary.

[0116] In some embodiments, the convolution model of the above theoretical seismic record satisfies the following expression:

[0117] S = W * D * X1, (12)

[0118] Where S represents the theoretical seismic record; W represents the wavelet matrix; D represents the differential matrix; X1 represents the reservoir parameter prediction data. For example, in Figure 3 , the process of measuring the reservoir parameter prediction data by the wavelet matrix and the differential matrix in the convolution model is represented by the measurement matrix H, and the synthetic record (i.e., the theoretical seismic record) is obtained.

[0119] The mean value of the above theoretical seismic record satisfies the following expression:

[0120]

[0121] Where, represents the mean value of the theoretical seismic record; W m (i) represents the first weight coefficient for the (i + 1)-th measurement data; S (i) represents the (i + 1)-th measurement data in the theoretical seismic record.

[0122] The variance corresponding to the above theoretical seismic record satisfies the following expression:

[0123]

[0124]

[0125] Where Ps represents the variance of the theoretical seismic record; sdiv (i) represents the difference between the value of the (i + 1)-th measurement data and the mean value of the theoretical seismic record; (sdiv (i) ) T represents sdiv(i) Transpose of; Wc (i) The second weight coefficient representing the (i + 1)-th measurement data; R represents measurement noise.

[0126] The above process describes the unscented transformation of the measurement data.

[0127] In step S140, according to the error between the above theoretical seismic record and the actual seismic record, the reservoir parameters of the current period are iteratively updated, and the target reservoir parameters with an error less than the set threshold are determined as the reservoir parameter results inverted in the current period.

[0128] In some embodiments, the reservoir parameters of the current period are iteratively updated by using the unscented Kalman filter algorithm to fuse the results of the previous period to obtain the update amount, that is, the above target reservoir parameters are obtained based on the unscented Kalman filter algorithm.

[0129] In some embodiments, referring to Figure 3 as shown, the error between the above theoretical seismic record and the actual seismic record satisfies the following expression:

[0130] error = ∑(dif) 2 = ∑(Sr - S) 2 , (16)

[0131] where error represents the error between the theoretical seismic record S and the actual seismic record Sr, and is abbreviated as e in Figure 3 .

[0132] Among them, by calculating the Kalman gain coefficient, the iterative update of the reservoir parameters and the variance update are performed based on the above Kalman gain coefficient.

[0133] The expression of the above Kalman gain coefficient is as follows:

[0134] K = Psx · Ps -1 , (17)

[0135]

[0136]

[0137] where K represents the Kalman gain coefficient, Psx represents the covariance between the predicted data and the measurement data; Ps represents the variance of the theoretical seismic record; sdiv (i) represents the difference between the value of the (i + 1)-th measurement data and the mean of the theoretical seismic record; xdiv (i) represents the difference between the value of the (i + 1)-th predicted data point and the mean of the reservoir parameter predicted data.

[0138] Based on the above Kalman gain coefficient, iterative update of reservoir parameters and variance update are performed, and the calculation process is as follows:

[0139] X 迭代后 = X 迭代前 + K(Sr - S), (20)

[0140] P 迭代后 = P 1 - K·(Psx) T , (21)

[0141] Among them, X 迭代后 represents the reservoir parameters after iterative update; X 迭代前 represents the reservoir parameters before iterative update, specifically the predicted data of reservoir parameters; P 迭代后 represents the variance after iterative update of reservoir parameters; P 1 represents the variance of the predicted data of reservoir parameters; (Psx) T represents the transpose of the covariance Psx.

[0142] Figure 4 Schematically shows a schematic diagram of the inversion process of the current - period reservoir parameters according to an embodiment of the present disclosure. Among them, (a) is the true wave impedance, (b) is the theoretical seismic record, (c) is the initial inversion model, and (d) is the inversion output result of the current - period reservoir parameters.

[0143] Referring to Figure 4 as shown in (a) - (d) therein, by iteratively updating the reservoir parameters of the current period, the target reservoir parameters with an error less than the set threshold are determined as the reservoir parameter results obtained by inversion in the current period. For example, the target reservoir parameters obtained in the current period are Xb, which are output as the reservoir parameter inversion results in the current period; the corresponding synthetic record is denoted as Sb, and the reservoir parameter inversion results in the current period are as Figure 4 shown in (d) therein.

[0144] In the embodiment including steps S110 - S140, by obtaining the initial reservoir parameters of the current period in the time - lapse seismic inversion process; the above - mentioned initial reservoir parameters are generated using the reservoir parameter results obtained by inversion in the previous period; predicting the next parameter state of the above - mentioned initial reservoir parameters to obtain the predicted data of reservoir parameters; determining the theoretical seismic record corresponding to the above - mentioned predicted data of reservoir parameters; according to the error between the above - mentioned theoretical seismic record and the actual seismic record, iteratively updating the reservoir parameters of the current period, and determining the target reservoir parameters with an error less than the set threshold as the reservoir parameter results obtained by inversion in the current period, it is possible to adapt to high - frequency time - lapse seismic data to achieve dynamic inversion of reservoir parameters. As the number of time - lapse periods increases, the reservoir changes in each period all contribute to the reservoir parameter inversion results in the final period, realizing the optimization of the overall inversion.

[0145] Figure 5 Schematically shows a schematic diagram of a theoretical model of reservoir parameters for multiple periods, where (a) is the reservoir parameter model of the basic data, (b) is the reservoir parameter model of the first monitoring period, (c) is the reservoir parameter model of the second monitoring period, (d) is the reservoir parameter model of the third monitoring period, (e) is the reservoir parameter model of the fourth monitoring period, and (f) is the reservoir parameter model of the fifth monitoring period. If the result of the inversion using the unscented Kalman filter algorithm adopted in the embodiments of the present disclosure is close to the theoretical model, it indicates that the inversion effect is ideal.

[0146] Figure 6 Schematically shows a schematic diagram of the effect of inverting the change process of reservoir parameters for multiple periods based on the unscented Kalman filter algorithm according to an embodiment of the present disclosure, where (a) is the basic data, (b) is the inversion output result of the first monitoring period, (c) is the inversion output result of the second monitoring period, (d) is the inversion output result of the third monitoring period, (e) is the inversion output result of the fourth monitoring period, and (f) is the inversion output result of the fifth monitoring period.

[0147] Comparison Figure 5 in the inversion change processes of each period in (a) - (f) and Figure 6 in the inversion change processes of each period in (a) - (f), it can be seen that since in the inversion iteration process of reservoir parameters for each period, the above-mentioned target reservoir parameters are obtained based on the unscented Kalman filter algorithm, that is, the inversion result of reservoir parameters for each period is the best output of fusing the results of multiple periods by the unscented Kalman filter algorithm. The inversion result of the method provided by the embodiments of the present disclosure is close to the theoretical model and can better reflect the dynamic changes of geological reservoir parameters during the process of oil and gas development.

[0148] In some embodiments, the above-mentioned target reservoir parameters are obtained based on the unscented Kalman filter algorithm. By fusing two sets of inversion data from different periods, an optimal estimate of the reservoir parameters is made. These two sets of data are prediction data and measurement data respectively. The prediction data is the previous inversion result plus state transformation and update, and the measurement data is the theoretical seismic record (which can also be described as the synthetic seismic record) obtained through the convolution model. Unscented transformation is performed on the prediction data and the measurement data to obtain the posterior probability density of the state transformation. Under the framework of the Kalman filter algorithm, the current reservoir physical property parameters are iteratively updated. When the error between the synthetic seismic record of the reservoir parameters and the actual seismic record meets a certain standard (for example, the error is less than the set threshold), the corresponding target reservoir parameters are output as the inversion result of the current period.

[0149] The second exemplary embodiment of the present disclosure provides a device for dynamic inversion of reservoir parameters.

[0150] Figure 7 Schematically shows a block diagram of the structure of the device for dynamic inversion of reservoir parameters according to an embodiment of the present disclosure.

[0151] Referring to Figure 7 As shown, the device 700 for dynamic inversion of reservoir parameters provided by an embodiment of the present disclosure includes: a parameter acquisition module 701, a parameter prediction module 702, a theoretical measurement module 703, and an iteration module 704.

[0152] The above-mentioned parameter acquisition module 701 is used to acquire the initial reservoir parameters of the current period during the time-lapse seismic inversion process; the above-mentioned initial reservoir parameters are generated by using the reservoir parameter results obtained from the inversion of the previous period.

[0153] The above-mentioned parameter prediction module 702 is used to predict the next parameter state of the above-mentioned initial reservoir parameters to obtain reservoir parameter prediction data.

[0154] The above-mentioned theoretical measurement module 703 is used to determine the theoretical seismic record corresponding to the above-mentioned reservoir parameter prediction data.

[0155] The above-mentioned iteration module 704 is used to iteratively update the reservoir parameters of the current period according to the error between the above-mentioned theoretical seismic record and the actual seismic record, and determine the reservoir parameter results obtained from the inversion of the current period by taking the target reservoir parameters with an error less than a set threshold.

[0156] In some embodiments, the reservoir parameters of the current period are iteratively updated by using the unscented Kalman filter algorithm to fuse the results of the previous period to obtain an update amount, that is, the above-mentioned target reservoir parameters are obtained based on the unscented Kalman filter algorithm.

[0157] For more details and beneficial effects of this embodiment, etc., reference can be made to the relevant descriptions of the first embodiment. The entire content of the first embodiment can be incorporated into this embodiment and will not be elaborated here.

[0158] Any one or more of the functional modules included in the above-mentioned device 700 can be combined and implemented in one module, or any one of the modules can be split into multiple modules. Or, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. At least one of the functional modules included in the device 700 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by any other reasonable means such as integrating or packaging the circuit, etc., in hardware or firmware, or implemented in any one of the three implementation manners of software, hardware, and firmware, or in an appropriate combination of any several of them. Or, at least one of the functional modules included in the device 700 can be at least partially implemented as a computer program module, and when the computer program module is run, the corresponding functions can be executed.

[0159] The third exemplary embodiment of the present disclosure provides an electronic device.

[0160] Figure 8 The structural block diagram of the electronic device provided by the embodiment of the present disclosure is schematically shown.

[0161] Referring to Figure 8 As shown, the electronic device 800 provided by the embodiment of the present disclosure includes a processor 801, a communication interface 802, a memory 803, and a communication bus 804. Among them, the processor 801, the communication interface 802, and the memory 803 complete mutual communication through the communication bus 804; the memory 803 is used to store a computer program; the processor 801, when executing the program stored on the memory, realizes the method for dynamic inversion of reservoir parameters as described above.

[0162] The fourth exemplary embodiment of the present disclosure further provides a computer-readable storage medium. A computer program is stored on the above computer-readable storage medium, and when the above computer program is executed by a processor, the method for dynamic inversion of reservoir parameters as described above is realized.

[0163] The computer-readable storage medium may be included in the device or apparatus described in the above embodiment; it may also exist alone without being assembled into the device or apparatus. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present disclosure is realized.

[0164] According to the embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, it may include but is not limited to: portable computer disks, hard disks, random access memories (RAMs), read-only memories (ROMs), erasable programmable read-only memories (EPROMs or flash memories), portable compact disk read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the above. In the present disclosure, the computer-readable storage medium may be any tangible medium that contains or stores a program, and the program may be used by or in combination with an instruction execution system, device, or apparatus.

[0165] It should be noted that in the technical solution provided by the embodiment of the present disclosure, in terms of the collection, collection, update, analysis, processing, use, transmission, storage, etc. of the user's personal information, it complies with the provisions of relevant laws and regulations, is used for legal purposes, and does not violate public order and good customs. Necessary measures are taken for the user's personal information to prevent illegal access to the user's personal information data, and to maintain the security of the user's personal information, network security, and national security.

[0166] It should be noted that, in this document, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprising", "including" or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the said element.

[0167] The above are only specific embodiments of the present disclosure, enabling those skilled in the art to understand or implement the present disclosure. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure will not be limited to these embodiments shown herein, but rather will be accorded the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for dynamic inversion of reservoir parameters, characterized in that, it includes: Obtaining the initial reservoir parameters of the current period during the time-lapse seismic inversion process; The initial reservoir parameters are generated by using the reservoir parameter results obtained from the inversion of the previous period; Predicting the next parameter state of the initial reservoir parameters to obtain reservoir parameter prediction data; Determining the theoretical seismic record corresponding to the reservoir parameter prediction data; According to the error between the theoretical seismic record and the actual seismic record, iteratively update the reservoir parameters of the current period, and determine the target reservoir parameters with an error less than the set threshold as the reservoir parameter results obtained from the inversion of the current period.

2. The method according to claim 1, characterized in that, the target reservoir parameters are obtained based on the unscented Kalman filter algorithm; wherein, predicting the next parameter state of the initial reservoir parameters to obtain reservoir parameter prediction data includes: Based on the unscented transform, constructing a reservoir parameter set corresponding to the initial reservoir parameters; the mean and variance of each group of reservoir parameters in the reservoir parameter set are the same as the mean and variance corresponding to the distribution state of the initial reservoir parameters; Calculating the product of a preset reservoir parameter model update amount and a preset random variable, and using the product as the state update amount; Calculating the sum of the reservoir parameter set and the state update amount, and using the sum as the reservoir parameter prediction data.

3. The method according to claim 2, characterized in that, Based on the unscented transform, constructing a reservoir parameter set corresponding to the initial reservoir parameters includes: Describing the initial reservoir parameter X by using a set of random variables corresponding to the reservoir parameter results obtained from the inversion of the previous period, the number of random variables is L, and the variance is P; According to the preset unscented transform parameters α, k, and λ, calculating 2L + 1 derived values corresponding to the initial reservoir parameter X, and the 2L + 1 groups of reservoir parameters corresponding to the derived values constitute the reservoir parameter set Xs, and the reservoir parameter set Xs satisfies the following expression: Xs( 0 ) = X, λ = α 2 (L + k) - L, where i represents the serial number of the reservoir parameter derivative value, and the value range of i is 0 to 2L; Xs( 0 ) represents the first derivative value in the reservoir parameter set Xs, and Xs (i) represents the (i + 1)-th derivative value in the reservoir parameter set Xs.

4. The method according to claim 3, characterized in that, the reservoir parameter prediction data satisfies the following expression: X1 = Xs + rs·Δx, wherein, X1 represents the reservoir parameter prediction data, which includes a total of 2L + 1 prediction data; Δx represents a preset reservoir parameter model update amount; rs represents a preset random variable; The mean of the reservoir parameter prediction data satisfies the following expression: Among them, represents the mean value of reservoir parameter prediction data; W m (i) represents the first weight coefficient for the (i + 1)-th data; X1 (i) represents the (i + 1)-th prediction data in the reservoir parameter prediction data; The variance of the reservoir parameter prediction data satisfies the following expression: Among them, xdiv (i) represents the difference between the value of the (i + 1)-th predicted data and the mean of the reservoir parameter prediction data; (xdiv (i) ) T represents the transpose of xdiv (i) ; Wc (i) represents the second weight coefficient for the (i + 1)-th data; Q represents the prediction noise of the state transition process, which is a given value; The first weight coefficient W m (i) and the second weight coefficient W c (i) are expressed as follows: W m (0) = λ / (L + λ), W m (i) = W c (i) = λ / 2(L + λ), i = 1 to 2L where β is a preset unscented transform parameter.

5. The method according to claim 1, characterized in that, Determining the theoretical seismic record corresponding to the reservoir parameter prediction data includes: Based on the convolution model, calculating the theoretical seismic record corresponding to the reservoir parameter prediction data.

6. The method according to claim 5, characterized in that, The convolution model of the theoretical seismic record satisfies the following expression: S = W * D * X1, wherein, S represents the theoretical seismic record; W represents the wavelet matrix; D represents the differential matrix; X1 represents the reservoir parameter prediction data; The mean of the theoretical seismic record satisfies the following expression: Among them, represents the mean of the theoretical seismic records; W m (i) represents the first weight coefficient for the (i + 1)-th measurement data; S (i) represents the (i + 1)-th measurement data in the theoretical seismic records; The variance of the theoretical seismic record satisfies the following expression: Among them, Ps represents the variance of the theoretical seismic record; sdiv (i) represents the difference between the value of the (i + 1)-th measurement data and the mean of the theoretical seismic record; (sdiv (i) ) T represents the transpose of sdiv (i) ; Wc (i) represents the second weight coefficient of the (i + 1)-th measurement data; R represents the measurement noise.

7. The method according to claim 1, wherein, the target reservoir parameters are obtained based on the unscented Kalman filter algorithm; the error between the theoretical seismic record and the actual seismic record satisfies the following expression: error = ∑(dif) 2 = ∑(Sr - S) 2 , where error represents the error between the theoretical seismic record S and the actual seismic record Sr; wherein, by calculating the Kalman gain coefficient, the reservoir parameters are iteratively updated and the variance is updated based on the Kalman gain coefficient; the expression of the Kalman gain coefficient is as follows: K = Psx·Ps -1 , Among them, K represents the Kalman gain coefficient, Psx represents the covariance between the predicted data and the measured data; Ps represents the variance of the theoretical seismic record; sdiv (i) represents the difference between the value of the (i + 1)-th measured data and the mean value of the theoretical seismic record; xdiv (i) represents the difference between the value of the (i + 1)-th predicted data point and the mean value of the reservoir parameter prediction data; X 迭代后 = X 迭代前 + K(Sr - S), P 迭代后 = P 1 - K·(Psx) T , Among them, X 迭代后 represents the reservoir parameters after iterative update; X 迭代前 represents the reservoir parameters before iterative update, specifically the predicted data of reservoir parameters; P 迭代后 represents the variance after iterative update of reservoir parameters; P 1 represents the variance of the predicted data of reservoir parameters; (Psx) T represents the transpose of the covariance Psx.

8. An apparatus for dynamic inversion of reservoir parameters, wherein, comprising: a parameter acquisition module, configured to acquire the initial reservoir parameters in the current stage during the time-lapse seismic inversion process; the initial reservoir parameters are generated by using the reservoir parameter results obtained from the previous inversion; a parameter prediction module, configured to predict the next parameter state of the initial reservoir parameters to obtain reservoir parameter prediction data; a theoretical measurement module, configured to determine the theoretical seismic record corresponding to the reservoir parameter prediction data; an iteration module, configured to iteratively update the reservoir parameters in the current stage according to the error between the theoretical seismic record and the actual seismic record, and determine the target reservoir parameters with an error less than a set threshold as the reservoir parameter results obtained from the current inversion.

9. An electronic device, wherein, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory complete communication with each other through the communication bus; the memory is used for storing a computer program; the processor is configured to implement the method according to any one of claims 1-7 when executing the program stored on the memory.

10. A computer-readable storage medium, on which a computer program is stored, wherein, the computer program, when executed by a processor, implements the method according to any one of claims 1-7.

Images