Shale high-quality reservoir prediction method and system based on whale optimization algorithm

Abstract

The invention discloses a shale high-quality reservoir prediction method and system based on a whale optimization algorithm, and solves the technical problems of strong hyper-parameter dependence, easy falling into local optimum, insufficient noise immunity and the like in a traditional wave impedance inversion method by combining a chaotic mapping strategy and a global optimization mechanism of the whale optimization algorithm. The method specifically comprises the steps that a nonlinear objective function driven by seismic data is established, chaotic mapping is adopted to optimize initial population distribution, wave impedance parameters are dynamically updated by simulating a self-adaptive search strategy of whale preying behaviors, and quantitative prediction of reservoir porosity is achieved in combination with rock physical analysis. According to the method, the uncertainty of manual adjustment of hyper-parameters is avoided, the recognition precision and inversion efficiency of the transverse boundary of the complex reservoir are remarkably improved, the method is particularly suitable for high-resolution exploration of low-porosity and low-permeability shale and compact sandstone reservoirs, and reliable technical support is provided for oil and gas reservoir fine description and well location deployment.

Description

Technical Field

[0001] The present invention relates to the technical field of geophysical exploration, and specifically relates to a method and system for predicting high-quality shale reservoirs based on a whale optimization algorithm. Background Art

[0002] Currently, in the fields of oil and gas exploration and fine description of shale reservoirs, the wave impedance inversion technology is the core means to reveal reservoir lithology, porosity, and fluid saturation. However, the existing methods face the following technical bottlenecks in practical applications:

[0003] 1. Hyperparameter dependence and local optimum problems of traditional inversion methods

[0004] Traditional methods represented by sparse pulse inversion and model-constrained inversion need to manually adjust multiple regularization hyperparameters (such as weight coefficients) to balance the stability and resolution of the inversion results. Under complex geological conditions with large lateral variations in reservoirs and scarce logging data, the prior estimation error of hyperparameters will cause the inversion results to deviate from the true geological model. In addition, the heterogeneity of underground media, seismic data noise, and thin interbed interference make the traditional methods insufficient in resolving weak reservoir features (such as sand body / shale pinch-out points), and the inversion results are prone to fall into local optima.

[0005] 2. Application limitations of deep learning methods

[0006] Although deep learning technology has been introduced to improve inversion accuracy, it depends on a large amount of logging data to train the model, but the number of wells in the actual work area is limited, resulting in insufficient generalization ability of the model. At the same time, the complexity of high-dimensional feature extraction is prone to cause overfitting problems, and the physical meaning of the inversion results is not clear, making it difficult to guide actual exploration decisions.

[0007] 3. Defects in the efficiency and convergence accuracy of intelligent optimization algorithms

[0008] Although existing intelligent optimization algorithms (such as simulated annealing and genetic algorithms) avoid the dependence on the initial model through a global search mechanism, their convergence speed is slow, the computational cost is high, and the uneven distribution of the initial population is prone to cause the algorithm to fall into local extreme value traps. For example, in the application of low-porosity and low-permeability reservoirs, traditional intelligent algorithms often result in a porosity prediction error exceeding 15% due to insufficient iteration times.

[0009] In view of the above problems, there is an urgent need for an inversion method that can simultaneously meet the following requirements:

[0010] Avoid hyperparameter dependence: Avoid manually adjusting weight coefficients and achieve a completely data-driven inversion process;

[0011] Global optimization ability: Break through the local optimum limit and accurately identify complex reservoir features (such as low-impedance gas layers and thin sand body boundaries);

[0012] High efficiency convergence and anti-noise performance: It converges stably within a finite number of iterations and has strong robustness to noise and lateral inhomogeneity;

[0013] Low data dependence: It can still achieve high-precision prediction under the condition of scarce logging samples.

[0014] In recent years, swarm intelligence optimization algorithms have shown potential in the field of non-linear inversion due to their simple parameters and strong global optimization ability. However, the deficiencies in the initial population distribution strategy and convergence mechanism of existing algorithms limit their accuracy and efficiency in complex reservoir prediction.

[0015] Based on the above technical gaps, the present invention proposes an innovative solution. By integrating the chaos mapping initialization strategy and the adaptive parameter update mechanism, it overcomes the hyperparameter sensitivity, local optimal trap, and computational efficiency bottleneck of traditional methods, providing reliable technical support for the high-resolution description of high-quality shale reservoirs. Summary of the Invention

[0016] Aiming at the defects of the existing technology, the present invention provides a method and system for predicting high-quality shale reservoirs based on the whale optimization algorithm. By establishing a data-driven objective function based on the least squares principle and using the whale optimization algorithm to globally optimize the wave impedance model, the problems of hyperparameter dependence and local optimality in traditional inversion methods are solved. At the same time, the Circle chaos mapping is used to optimize the initial population distribution, further improving the global search ability and inversion accuracy of the algorithm. Finally, by combining the inverted wave impedance with logging data and rock physics analysis results, a statistical regression model of wave impedance and porosity is established to achieve high-resolution prediction of high-quality shale reservoirs.

[0017] In order to achieve the above invention objectives, the technical solutions adopted by the present invention are as follows:

[0018] A method for predicting high-quality shale reservoirs based on the whale optimization algorithm, comprising the following steps:

[0019] S1: Establish an objective function based on the residual norm of post-stack seismic data and synthetic seismograms, expressed as:

[0020] Cost(Z) = ‖(d - f(Z))‖

[0021] where d is the observed seismic data, and f(Z) is the synthetic seismic record corresponding to the wave impedance Z, which is obtained by calculation through the convolution model;

[0022] S2: Initialize the wave impedance parameter population using the Circle chaos mapping strategy, and the specific formula is:

[0023] Z i,j = (Ub i,j - Lbi,j )· rand [0, 1] + Lb i,j

[0024] where i and j represent the population and dimension coordinate indices respectively, and Ubi ,j and Lbi ,j represent the upper and lower boundaries of the search window.

[0025] Then, use the Circle chaotic map shown in the following formula to further initialize the randomly generated population:

[0026]

[0027] where represents the new initialization position, and mod() represents the remainder operation.

[0028] S3: Update the wave impedance parameters through three stages of the whale optimization algorithm:

[0029] Search for prey stage: When |A| > 1, update the position according to the random search strategy:

[0030] Z(t + 1) = Z rand (t) - A·|C·Z rand (t) - Z(t)|

[0031] Encircle prey stage: When |A| ≤ 1, update the position according to the optimal individual guidance strategy:

[0032] Z(t + 1) = Z best (t) - A·|C·Z best (t) - Z(t)|

[0033] Bubble net attack stage: Select the update strategy with probability p:

[0034] Z(t + 1) = {|Z best (t) - Z(t)|·e bl ·cos(2πl) + Z best (t)

[0035] where A represents the adaptive coefficient, A = 2a· rand [0, 1] - a, C represents the random coefficient, C = 2· rand [0, 1], a represents the control parameter attenuation factor, T is the maximum number of iterations, t is the number of iterations, l is a random number between 0 and 1, b is a constant used to define the logarithmic spiral shape, Z rand (t) represents the randomly selected individual position, Z(t) represents the current individual position, Z best(t) represents the position of the optimal individual in the current population, and e bl represents the exponential function;

[0036] S4: During the iteration process, the population is dynamically updated according to the objective function until the preset convergence criterion is met or the maximum number of iterations is reached, and the wave impedance model corresponding to the minimum value of the objective function is output;

[0037] S5: Combine the wave impedance model with well logging data and the results of rock physics analysis to achieve quantitative prediction of high-quality shale reservoirs.

[0038] Furthermore, the search window boundaries Ub i,j and Lb i,j are determined as follows:

[0039] Based on the wave impedance logging curves of at least one well in the target work area, its mean value and standard deviation are statistically analyzed;

[0040] Set the search window boundaries to the wave impedance value range of the mean value ± 20% to constrain the population initialization.

[0041] Furthermore, the parameter settings of the whale optimization algorithm in step S3 include: the population size is 50 - 100; the maximum number of iterations T = 200; the probability p = 0.5, which is used to balance the encircling shrinking and spiral bubble net attack strategies.

[0042] Furthermore, the reservoir prediction in step S5 includes:

[0043] Divide the shale reservoir and non-reservoir through the wave impedance threshold: the gas layer and the poor gas layer correspond to low wave impedance, and the dry layer and mudstone correspond to high wave impedance;

[0044] Based on the linear regression relationship between wave impedance and porosity, convert the wave impedance inversion result into a porosity distribution profile.

[0045] Furthermore, the convergence condition in step S3 is: the change rate of the objective function value in 10 consecutive iterations is less than 1%, or the maximum number of iterations T is reached.

[0046] The present invention also discloses a high-quality shale reservoir prediction system based on the whale optimization algorithm, which can be used to implement the above-mentioned high-quality shale reservoir prediction method. Specifically, it includes the following modules:

[0047] Objective function construction module: configured to establish an objective function based on the residual norm of the post-stack seismic data and the synthetic seismogram

[0048] Population initialization module: configured to initialize the wave impedance parameter population by using the Circle chaotic mapping strategy, specifically including:

[0049] Initial population generation unit: Generate a random population based on the boundaries of the search window:

[0050] Chaotic mapping optimization unit: Further optimize the initial population using the Circle chaotic mapping:

[0051] Wave impedance optimization module: Configured to update the wave impedance parameters through three stages of the whale optimization algorithm, including: Search for prey sub-module: Update the position according to the random search strategy;

[0052] Encircle prey sub-module: Update the position according to the optimal individual guidance strategy;

[0053] Bubble net attack sub-module: Select the update strategy with probability p;

[0054] Iterative control and output module: Configured to dynamically update the population according to the objective function during the iteration process until the preset convergence criterion is met or the maximum number of iterations is reached, and output the wave impedance model corresponding to the minimum value of the objective function;

[0055] Reservoir prediction module: Configured to combine the wave impedance model with well logging data and the results of rock physics analysis;

[0056] Result display and interaction module: Configured to visualize the inversion results and provide user interaction functions.

[0057] The present invention also discloses a computer device, including a memory, a processor, and a computer program stored on the memory and executable by the processor. When the processor executes the program, the above-mentioned shale high-quality reservoir prediction method is implemented.

[0058] The present invention also discloses a computer-readable storage medium, in which a computer program is stored, and when the program is executed by a processor, the above-mentioned shale high-quality reservoir prediction method is implemented.

[0059] Compared with the prior art, the advantages of the present invention are as follows:

[0060] 1. Traditional wave impedance inversion methods need to rely on manually adjusting hyperparameters such as the regularization weight coefficient, and their values have a significant impact on the inversion results. Especially in areas with insufficient well logging data or large lateral changes in reservoirs, it is easy to cause model deviations. Through the global optimization mechanism of the whale optimization algorithm, the present invention completely abandons the adjustment of hyperparameters and only drives the objective function based on seismic data (Cost(Z) = ∥d - f(Z)∥), effectively solving the hyperparameter sensitivity problem in traditional methods and significantly improving the inversion stability under complex geological conditions.

[0061] 2. The present invention optimizes population initialization by introducing the Circle chaotic mapping strategy, enabling the initial solutions to evenly cover the search space. Combining the adaptive search, encirclement, and bubble net attack mechanisms of the whale optimization algorithm significantly reduces the risk of the algorithm falling into local extrema and speeds up the convergence rate. Compared with traditional intelligent optimization algorithms, the present invention greatly reduces the number of iterations while ensuring high precision, significantly improving the computational efficiency.

[0062] 3. The present invention has strong robustness to low signal-to-noise ratio seismic data, can effectively suppress noise interference, and stably obtain high-precision wave impedance inversion results. At the same time, its ability to depict formation boundaries is significantly better than traditional linear inversion methods, especially suitable for the fine identification of thin interbedded reservoirs and pinch-out points of sand-shale bodies, providing a reliable technical means for complex reservoir description.

[0063] 4. Based on the wave impedance-porosity cross-plot relationship established by rock physics analysis, the present invention can accurately convert the wave impedance inversion results into reservoir physical properties parameters (such as porosity), realizing the quantitative prediction of reservoir physical properties. Compared with traditional statistical interpolation methods, the present invention shows higher reliability and consistency in the spatial distribution prediction of reservoir parameters.

[0064] 5. The present invention can be modularly integrated into a reservoir prediction system, supporting 3D visualization, dynamic parameter interaction, and multi-format data export, and is compatible with mainstream seismic interpretation and geological modeling software. By adjusting the search window boundary and rock physics relationship model, it can quickly adapt to various reservoir types such as tight sandstone and shale gas, and has broad engineering application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 is the logging curve of Well A;

[0066] Figure 2 is based on the rock physics analysis results of Well A;

[0067] Figure 3 is the cross-plot of porosity and wave impedance of Well A;

[0068] Figure 4 is the flow chart of the shale high-quality reservoir prediction method based on the whale optimization algorithm in the embodiment of the present invention;

[0069] Figure 5 is the true wave impedance value map of part of the Marmousi 2 model in the embodiment of the present invention;

[0070] Figure 6 is the synthetic post-stack seismic record map in the embodiment of the present invention; in the figure, (a) has no noise, and (b) has a signal-to-noise ratio of 1;

[0071] Figure 7This is the post-stack wave impedance inversion result based on the whale optimization algorithm in the embodiment of the present invention; in the figure, (a) has no noise, and (b) has a signal-to-noise ratio of 2.

[0072] Figure 8 This is the inversion result diagram corresponding to CDP 100 in the embodiment of the present invention; in the figure, (a) has no noise, and (b) has a signal-to-noise ratio of 2

[0073] Figure 9 This is the post-stack seismic data diagram passing through the sandstone pinch-out area in the embodiment of the present invention;

[0074] Figure 10 This is the inversion result diagram of the sparse pulse wave impedance inversion method in the embodiment of the present invention;

[0075] Figure 11 This is the inversion result diagram of the shale high-quality reservoir prediction method in the embodiment of the present invention;

[0076] Figure 12 This is the cross-sectional diagram of the wave impedance inversion result of the shale high-quality reservoir prediction method in the embodiment of the present invention;

[0077] Figure 13 This is the cross-sectional diagram of the porosity prediction result of the shale high-quality reservoir prediction method in the embodiment of the present invention. Detailed implementation manners

[0078] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the following further elaborates on the present invention in detail according to the accompanying drawings and by way of examples.

[0079] I. Rock physics analysis

[0080] Rock physics research is to study the physical properties of rocks and their variation laws, such as the longitudinal wave velocity, transverse wave velocity, and density of shale. Generally, different rocks have different physical characteristics. The physical characteristics of rocks are directly related to seismic characteristics. Therefore, the physical properties of underground rocks and their saturated fluids can be extracted from seismic data through seismic inversion. So rock physics is a bridge connecting seismic characteristics and formation characteristics. Its specific role is to clarify the physical properties of various rocks and their corresponding seismic characteristics through rock physics research, so as to guide seismic inversion, interpret the inversion results, and evaluate the quality of inversion. Reservoir A is developed with typical low-porosity and low-permeability silty shale and muddy sandstone reservoirs in the southwestern region of the Sichuan Basin. Efficiently, accurately, and with high resolution to identify favorable reservoirs and clarify their thickness and distribution are important tasks for the fine exploration of this reservoir. In this embodiment, first from the perspective of rock physics, the necessity of carrying out wave impedance inversion on Reservoir A is further analyzed and explained. Figure 1It is the actual logging curve of Well A running through Reservoir A. From the logging interpretation results shown therein, it can be seen that the gas-bearing zones and poor gas-bearing zones mainly exhibit low impedance characteristics, while the dry layers and mudstone layers show high impedance characteristics. Figure 2 It is the petrophysical analysis result obtained based on the actual logging curve of this well. From the histograms and crossplots shown in the figure, it can be seen that the favorable shale reservoirs and non-reservoirs can be well distinguished by wave impedance. Figure 3 It is the crossplot of wave impedance and porosity of Reservoir A obtained based on the data statistics of this well. It can be seen that there is a very good correlation between the porosity and wave impedance of this reservoir. From the above analysis results, it can be seen that wave impedance can not only effectively distinguish favorable reservoirs and non-reservoirs, but also further characterize the porosity of the reservoir, fully demonstrating the necessity of carrying out high-precision and high-resolution wave impedance inversion in Reservoir A.

[0081] II. Theory and Process of the Shale High-Quality Reservoir Prediction Method Based on Whale Optimization Algorithm in the Present Invention

[0082] 2.1 Conventional Poststack Wave Impedance Inversion

[0083] Currently, the commonly used wave impedance inversion methods in actual production are mainly the sparse pulse wave impedance inversion method and the wave impedance inversion method based on model constraints.

[0084] Based on the convolution model, the forward equation of poststack seismic data can be expressed as:

[0085] d = f(Z) + n (1)

[0086] where d is the observed seismic data, f(Z) is the non-linear function of seismic data with respect to wave impedance Z, and n is the noise.

[0087] (1) Sparse Pulse Wave Impedance Inversion Method

[0088] The sparse pulse inversion algorithm is a single-channel model-constrained inversion method based on the convolution model shown in formula (1). Assuming that the reflection coefficient sequence of the formation is sparsely distributed, the purpose of inversion is to minimize the following objective function:

[0089] E = ∑(r) p + λ q ∑(d - f(Z)) q + α 2 ∑(t - Z) 2 (2)

[0090] Among them, \(E\) is the objective function, \(r\) is the reflection coefficient sequence, \(Z\) is the wave impedance sequence, \(d\) is the observed seismic record; \(f(Z)\) is the synthetic seismic record; \(t\) is the artificially set impedance trend, which can generally be replaced by the initial model, \(p\) and \(q\) are norms, generally taking 1 and 2 respectively, and \(\lambda\) and \(\alpha\) are weight coefficients, that is, the hyperparameters mentioned above.

[0091] Based on formula (2), an iterative method can be used for solution. First, experiments are carried out with a small number of pulses to obtain an initial reflection coefficient model, and then the model is continuously modified to make the objective function value smaller and smaller. Continuously increase the number of pulses and repeat the iteration until the inverted wave impedance model can no longer be significantly optimized, then the iteration can be terminated, and the final wave impedance model is output as the inversion result.

[0092] (2) Wave impedance inversion method based on model constraints

[0093] Based on Bayesian theory, the following objective function can be constructed:

[0094]

[0095] Among them, are the hyperparameters mentioned above, represents the variance of wave impedance parameters, represents the noise variance, and \(\mu\) is the mean vector of wave impedance parameters, usually taking the value of the initial model. The solution of the above nonlinear inversion objective function can be realized by means of Taylor series expansion.

[0096] It is not difficult to see that both common methods introduce hyperparameters, and in actual inversion, reasonable values of these parameters are required, otherwise it is impossible to obtain a wave impedance inversion result with satisfactory accuracy. To avoid the occurrence of this problem, intelligent optimization algorithms are very good alternative choices.

[0097] 2.2 Wave impedance inversion based on whale optimization algorithm

[0098] Post-stack seismic data has the advantages of high signal-to-noise ratio, high resolution, and good quality. Therefore, for complete nonlinear inversion, it is only necessary to establish the following objective function driven by seismic data based on the least squares principle:

[0099] Cost(Z) = ‖(d - f(Z))‖ (4)

[0100] Among them, Cost represents the value function.

[0101] The whale optimization algorithm is a swarm intelligence optimization algorithm developed by simulating the predation behavior of humpback whales. The algorithm mainly includes three stages: searching for prey, surrounding prey, and bubble net attack.

[0102] (1) Search for prey: In the whale optimization algorithm, each whale represents a feasible solution Z of the objective function i , corresponding to the wave impedance value of a sampling point in the wave impedance sequence to be inverted in the objective function shown in formula (4). The value range of i is from 1 to the maximum number of sampling points, that is, the dimension D of the population. In the initial stage, based on the random walk mechanism of whale hunting, individual whales randomly search and update the position of the next generation. When |A|>1, the search behavior of a certain whale individual can be expressed as:

[0103] Z(t + 1) = Z rand (t) - A·|C·Z rand (t) - Z(t)| (5)

[0104] where Z(t + 1) represents the position that needs to be updated at the current iteration moment of the whale, and Z(t) represents the position of the whale at the current iteration moment. They correspond to the wave impedance values at different moments of a sampling point in the wave impedance sequence to be inverted. t represents the current number of iterations, Z rand (t) is the current position of the whale individual and is a random position selected from the current population. The calculation of coefficients A and C is as follows:

[0105] A = 2a· rand [0, 1] - a (6)

[0106] C = 2· rand [0, 1] (7)

[0107] where, T is the maximum number of iterations.

[0108] (2) Encircle the prey: When |A|≤1, the whale group begins to enter the stage of encircling the prey. The whale group gradually approaches the position of the whale with the optimal fitness value (corresponding to the minimum value of the cost function). By gradually encircling, shrinking, and approaching the prey and determining the target position, this process can be expressed as:

[0109] Z(t + 1) = Z best (t) - A·|C·Z best (t) - Z(t)| (8)

[0110] where, Z best (t) represents the optimal position of the whale at the current number of iterations.

[0111] (3) Bubble net attack method: This is a unique hunting strategy of whales. While exhaling bubbles, they swim upward in a spiral, surrounding and approaching the prey to the sea surface, and finally catching the prey. Considering that the behaviors of different whales are not synchronous, a random probability p is set to balance the two behaviors of encircling and shrinking and spiral bubble net hunting, and is expressed as:

[0112]

[0113] Among them, l is a random number with a value between 0 and 1, and b is a constant used to define the logarithmic spiral shape, usually taken as 1.

[0114] Based on formulas (5)-(9), the global optimization based on the whale optimization algorithm can be achieved. Similar to other optimization algorithms, before starting to implement the optimization algorithm, the population size needs to be defined and the population needs to be initialized. A good population initialization strategy can further improve the convergence progress and speed of the algorithm. For this reason, this embodiment introduces a more reasonable Circle chaotic mapping initialization strategy. First, the initial population can be randomly generated using the following formula:

[0115] Z i,j =(Ub i,j -Lb i,j )· rand [0,1]+Lb i,j (10)

[0116] Among them, i and j respectively represent the population and dimension coordinate indices, Ub i,j and Lb i,j represent the upper and lower boundaries of the search window.

[0117] Then, the randomly generated population can be further initialized using the Circle chaotic mapping shown in the following formula:

[0118]

[0119] Among them, represents the new initialization position, and mod() represents the remainder operation.

[0120] Based on the above position update formula, the global optimization solution of the wave impedance parameters to be inverted can be achieved through multiple iterations. The corresponding flowchart of the present invention is as Figure 4 shown, including the following steps:

[0121] S1: Establish an objective function based on the residual norm of the post-stack seismic data and the synthetic seismogram;

[0122] S2: Initialize the wave impedance parameter population using the Circle chaotic mapping strategy;

[0123] S3: Update the wave impedance parameters through three stages of the whale optimization algorithm;

[0124] S4: During the iteration process, dynamically update the population according to the objective function until the preset convergence criterion is met or the maximum number of iterations is reached, and output the wave impedance model corresponding to the minimum value of the objective function;

[0125] S5: Combine the wave impedance model with well logging data and rock physics analysis results to achieve quantitative prediction of high-quality shale reservoirs.

[0126] III. Application Test of the Method of the Present Invention

[0127] 3.1 Synthetic Data Test

[0128] To test the feasibility and effectiveness of the research method, this embodiment conducts a synthetic data test on the method of the present invention using a part of the Marmousi 2 model. Figure 5 is the true P-wave impedance value corresponding to the intercepted part of the Marmousi 2 model, and the sampling interval is 2 ms. The reflection coefficient is calculated using the illustrated true impedance value and convolved with a Ricker wavelet with a dominant frequency of 40 Hz to obtain the synthetic post-stack seismic response profile as shown in Figure 6 a. To test the stability and noise resistance of the proposed post-stack wave impedance non-linear inversion method based on the whale optimization algorithm, random noise with a root mean square signal-to-noise ratio of 1 is added to the noise-free seismic data shown in Figure 6 a, as shown in Figure 6 b. Based on the noise-free data and noisy data shown in Figure 6 , inversion tests based on the method of the present invention are respectively carried out, and the inversion results are as shown in Figure 7 . It can be seen from the comparison between Figure 5 and Figure 7 that the method of the present invention can obtain the wave impedance inversion result with high precision, and also has good stability and noise resistance. Figure 8 is the comparison between the wave impedance inversion result corresponding to CDP 100 and the true wave impedance inversion result. It can be seen that the inversion result of the method of the present invention has a good match with the true value. In summary, the post-stack wave impedance non-linear inversion method based on the whale optimization algorithm studied in this embodiment can stably and accurately obtain the wave impedance inversion result, and performs excellently in terms of stability and noise resistance.

[0129] 3.2 Actual Data Test

[0130] The overall reservoir of Target Research Area A is in a shelf facies sedimentary environment. The reservoir of silty mudstone and black shale in the deep-water shelf top sedimentary environment is better. The reservoir is mainly composed of tight shale and siltstone, showing low porosity and low permeability characteristics. To demonstrate the advantages of the method of the present invention, first, a comparison between the sparse pulse inversion method and the method of the present invention is carried out based on the post-stack seismic data shown in Figure 9 . Figure 10 And Figure 11They are the inversion results obtained by the sparse pulse inversion method and the method of the present invention respectively. It can be seen from the comparison that the wave impedance inversion result based on the method of the present invention depicts the sand body pinch-out point marked by the elliptical area more clearly and has a better effect, fully verifying the effectiveness and advantages of the method of the present invention in the prediction of Reservoir A. Based on the above conclusions, this embodiment combines the rock physics analysis results and conducts the prediction of Reservoir A. Figure 12 It is a well-connected profile of the wave impedance inversion result obtained based on the method of the present invention. It can be seen from the comparison with the logging curves inserted therein that the method of the present invention can obtain wave impedance values with satisfactory rationality and accuracy. The reservoir space of Reservoir A is mainly intragranular dissolved pores and micropores. The effective reservoir porosity is distributed between 4.0% and 13.9%, with an average value of 6.6% and a median value of 6.4%. Based on Figure 3 the porosity-wave impedance cross-plot shown, this embodiment further conducts the prediction of reservoir porosity based on the high-precision wave impedance inversion result. The prediction result is as shown in Figure 13 The inserted curve on the profile is the porosity curve. From the analysis of the prediction result, the porosity curve on the well is in good agreement with the prediction result. Overall, the porosity of Reservoir A is about 4.% - 8.2%. The physical properties of the first and second sections are relatively good. It can be seen from the well logging curve of the third section that the porosity is a bit worse. From the comparison of the physical properties of the upper, middle and lower sections, the physical properties of the first and second sections are good, and the average porosity can be about 5% - 10%. However, the physical properties of the third section are relatively poor, and the predicted porosity in the main drilling platform area is about 4.5% - 6.1%. The above prediction results are highly consistent with the actual drilling results, fully verifying the feasibility and effectiveness of the method of the present invention.

[0131] IV. Conclusion

[0132] Post-stack wave impedance inversion is an important means for fine reservoir exploration. The rock physics analysis results show that wave impedance can not only effectively distinguish the favorable reservoirs and non-reservoirs in the Qiongzhusi Formation strata, but also further characterize the reservoir porosity, highlighting the necessity of conducting post-stack wave impedance inversion for Reservoir A. Therefore, in order to improve the rationality and accuracy of the prediction results of Reservoir A, this embodiment develops a high-precision post-stack wave impedance non-linear inversion method based on the whale optimization algorithm for the actual problems faced by the target reservoir parameter inversion, and introduces the circle chaotic mapping population initialization strategy to further improve the convergence accuracy and speed of the inversion method. The synthetic data test results show that the method of the present invention can stably obtain the wave impedance inversion result with high precision. The actual data test results show that the research method can effectively overcome the defects of the traditional method and obtain more reasonable and satisfactory wave impedance inversion results, and finally realizes the fine characterization of Reservoir A based on the corresponding inversion results, fully verifying the feasibility and effectiveness of the research method.

[0133] In another embodiment of the present invention, a shale high-quality reservoir prediction system based on the whale optimization algorithm is proposed, which can be used to implement the above-mentioned shale high-quality reservoir prediction method. Specifically, it includes the following modules:

[0134] Objective function construction module: configured to establish an objective function based on the residual norm of post-stack seismic data and synthetic seismograms

[0135] Population initialization module: configured to initialize the wave impedance parameter population using the Circle chaotic mapping strategy, specifically including:

[0136] Initial population generation unit: generate a random population based on the search window boundary

[0137] Chaotic mapping optimization unit: further optimize the initial population using the Circle chaotic mapping

[0138] Wave impedance optimization module: configured to update the wave impedance parameters through three stages of the whale optimization algorithm, including: Search for prey sub-module: when |A|>1, update the position according to the random search strategy

[0139] Surround prey sub-module: when |A|≤1, update the position according to the optimal individual guidance strategy

[0140] Bubble net attack sub-module: select the update strategy with probability p

[0141] Iterative control and output module: configured to dynamically update the population according to the objective function during the iteration process until the preset convergence criterion is met or the maximum number of iterations is reached, and output the wave impedance model corresponding to the minimum value of the objective function

[0142] Reservoir prediction module: configured to combine the wave impedance model with well logging data and rock physics analysis results to achieve quantitative prediction of shale high-quality reservoirs through the following steps:

[0143] Divide low-impedance reservoirs (gas layers, poor gas layers) and high-impedance non-reservoirs (dry layers, mudstones) based on wave impedance thresholds

[0144] Use the wave impedance-porosity crossplot relationship to convert the wave impedance inversion result into a porosity distribution profile

[0145] Result display and interaction module: configured to visualize the inversion result and provide user interaction functions, specifically including:

[0146] Visualization sub-module: generate wave impedance inversion profile diagrams (corresponding to Figure 7 、 Figure 9 ), porosity distribution profile diagrams( Figure 12 ) and planar distribution diagrams( Figure 13 )

[0147] Support 3D geological model rendering and dynamically display the spatial distribution of reservoirs;

[0148] Interactive control sub-module: Provide an interface for dynamic parameter adjustment, including a wave impedance threshold slider and a porosity range filter;

[0149] Support multi-view synchronous linkage. When the user clicks on any position of the cross-section diagram, it automatically locates to the corresponding 3D model coordinates and logging curves;

[0150] Integrate the data export function and output the inversion results as a SEGY format seismic data volume or a CSV format porosity table.

[0151] In another embodiment of the present invention, a terminal device is provided. The terminal device includes a processor and a memory. The memory is used to store a computer program. The computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the shale high-quality reservoir prediction method, including the following steps:

[0152] S1: Establish an objective function based on the residual norm of post-stack seismic data and synthetic seismograms;

[0153] S2: Initialize the wave impedance parameter population using the Circle chaotic mapping strategy;

[0154] S3: Update the wave impedance parameters through the three stages of the whale optimization algorithm;

[0155] S4: During the iteration process, dynamically update the population according to the objective function until the preset convergence criterion is met or the maximum number of iterations is reached, and output the wave impedance model corresponding to the minimum value of the objective function;

[0156] S5: Combine the wave impedance model with logging data and rock physics analysis results to achieve quantitative prediction of shale high-quality reservoirs.

[0157] In another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a terminal device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and, of course, the extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space, and the operating system of the terminal is stored in this storage space. Moreover, one or more instructions suitable for being loaded and executed by a processor are stored in this storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory.

[0158] One or more instructions stored in the computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the shale high-quality reservoir prediction method in the above embodiment; one or more instructions in the computer-readable storage medium are loaded and executed by a processor to perform the following steps:

[0159] S1: Establish an objective function based on the residual norm of post-stack seismic data and synthetic seismograms;

[0160] S2: Initialize the wave impedance parameter population using the Circle chaotic mapping strategy;

[0161] S3: Update the wave impedance parameters through three stages of the whale optimization algorithm;

[0162] S4: During the iteration process, dynamically update the population according to the objective function until the preset convergence criterion is met or the maximum number of iterations is reached, and output the wave impedance model corresponding to the minimum value of the objective function;

[0163] S5: Combine the wave impedance model with well logging data and the results of rock physics analysis to achieve quantitative prediction of shale high-quality reservoirs.

[0164] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0165] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and combinations of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general-purpose computers, special-purpose computers, embedded processors, or other programmable data processing devices to produce a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices generate means for implementing the functions specified in the flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks or blocks.

[0166] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory produce a manufactured article including instruction means that implement the functions specified in the flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks or blocks.

[0167] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in the flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or means for implementing the functions specified in multiple blocks or blocks.

[0168] Those of ordinary skill in the art will realize that the embodiments described herein are to assist the reader in understanding the implementation methods of the present invention, and it should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not depart from the essence of the present invention based on the technical revelations disclosed in the present invention, and these deformations and combinations are still within the protection scope of the present invention.

Claims

1. A method for predicting high-quality shale reservoirs based on the whale optimization algorithm, characterized in that, It includes the following steps: S1: Establish an objective function based on the residual norm of post-stack seismic data and synthetic seismograms, expressed as: Cost(Z) = ‖(d - f(Z))‖ where d is the observed seismic data, and f(Z) is the synthetic seismogram corresponding to the acoustic impedance Z, which is obtained by convolution model calculation; S2: Initialize the population of acoustic impedance parameters using the Circle chaotic mapping strategy. The specific formula is: Z i,j = (Ub i,j - Lb i,j )·rand[0,1] + Lb i,j where i and j represent the population and dimension coordinate indices respectively, Ub i,j and Lb i,j represent the upper and lower boundaries of the search window; Then, further initialize the randomly generated population using the Circle chaotic mapping shown in the following formula: Among them, represents the new initialization position, and mod() represents the remainder operation; S3: Update the acoustic impedance parameters through three stages of the whale optimization algorithm: Search for prey stage: When |A| > 1, update the position according to the random search strategy: Z(t + 1) = Z rand (t) - A·|C·Z rand (t) - Z(t)| Encircle prey stage: When |A| ≤ 1, update the position according to the optimal individual guidance strategy: Z(t + 1) = Z best (t) - A·|C·Z best (t) - Z(t)| Bubble-net attack stage: Select the spiral update strategy with probability p: Z(t + 1) = {|Z best (t) - Z(t)|·e bl ·cos(2πl) + Z best (t) Among them, A represents the adaptive coefficient, A = 2a·rand[0,1] - a, C represents the random coefficient, C = 2·rand[0,1], and a represents the control parameter attenuation factor. T is the maximum number of iterations, t is the number of iterations, l is a random number between 0 and 1, b is a constant used to define the logarithmic spiral shape, Z rand (t) represents the randomly selected individual position, Z(t) represents the current individual position, Z best (t) represents the position of the optimal individual in the current population, e bl represents the exponential function; S4: During the iteration process, dynamically update the population according to the objective function until the preset convergence criterion is met or the maximum number of iterations is reached, and output the acoustic impedance model corresponding to the minimum value of the objective function; S5: Combine the acoustic impedance model with well logging data and rock physics analysis results to achieve quantitative prediction of high-quality shale reservoirs.

2. The shale high-quality reservoir prediction method according to claim 1, wherein: The search window boundaries Ub i,j and Lb i,j in step S2 are determined as follows: Based on the acoustic impedance logging curves of at least one well in the target work area, statistically calculate its mean value and standard deviation; Set the search window boundary to the range of acoustic impedance values of mean ± 20% to constrain population initialization.

3. The shale high-quality reservoir prediction method according to claim 1, wherein: The parameter settings of the whale optimization algorithm described in step S3 include: the population size is 50 - 100; the maximum number of iterations T = 200; the probability p = 0.5, which is used to balance the encirclement contraction and spiral bubble-net attack strategies.

4. The shale high-quality reservoir prediction method according to claim 1, wherein: The reservoir prediction described in step S5 includes: Divide the sandstone reservoir and non-reservoir through the acoustic impedance threshold: gas layers and poor gas layers correspond to low acoustic impedance, and dry layers and mudstones correspond to high acoustic impedance; Based on the linear regression relationship between acoustic impedance and porosity, convert the acoustic impedance inversion result into a porosity distribution profile.

5. The shale high-quality reservoir prediction method according to claim 1, characterized in that: The convergence condition described in step S3 is: the change rate of the objective function value in 10 consecutive iterations is less than 1%, or the maximum number of iterations T is reached.

6. A shale high-quality reservoir prediction system based on the whale optimization algorithm, characterized in that: This system can be used to implement the high-quality shale reservoir prediction method described in any one of claims 1 to 5; Specifically, it includes the following modules: Objective function construction module: Configured to establish an objective function based on the residual norm of post-stack seismic data and synthetic seismograms Population initialization module: Configured to initialize the population of acoustic impedance parameters using the Circle chaotic mapping strategy, specifically including: Initial population generation unit: Generate a random population based on the search window boundary: Chaotic mapping optimization unit: Further optimize the initial population using the Circle chaotic mapping: Acoustic impedance optimization module: Configured to update the acoustic impedance parameters through three stages of the whale optimization algorithm, including: Search for prey sub-module: Update the position according to the random search strategy; Encircle prey sub-module: Update the position according to the optimal individual guidance strategy; Bubble-net attack sub-module: Select the spiral update strategy with probability p; Iteration control and output module: Configured to dynamically update the population according to the objective function during the iteration process until the preset convergence criterion is met or the maximum number of iterations is reached, and output the acoustic impedance model corresponding to the minimum value of the objective function; Reservoir prediction module: configured to combine the wave impedance model with well logging data and the results of petrophysical analysis; Result display and interaction module: configured to visualize the inversion results and provide user interaction functions.

7. A computer device, characterized in that: It includes a memory, a processor, and a computer program stored on the memory and executable by the processor. When the processor executes the program, it implements the shale high-quality reservoir prediction method according to any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: It stores a computer program, which implements the shale high-quality reservoir prediction method according to any one of claims 1 to 5 when executed by the processor.

Images