Shale premium reservoir prediction method and system based on whale optimization algorithm

Abstract

The application discloses a shale high-quality reservoir prediction method and system based on a whale optimization algorithm, solves technical problems such as strong super parameter dependence, easy falling into local optimization, insufficient noise resistance 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, and specifically comprises the following steps: establishing a nonlinear objective function driven by seismic data, optimizing initial population distribution by using the chaotic mapping, dynamically updating wave impedance parameters by using an adaptive search strategy simulating a whale predation behavior, and realizing quantitative prediction of reservoir porosity in combination with rock physical analysis. The application avoids uncertainty of artificial adjustment of super parameters, significantly improves lateral boundary identification precision and inversion efficiency of a complex reservoir, is especially suitable for high-resolution exploration of low-porosity and low-permeability shale and tight sandstone reservoirs, and provides reliable technical support for fine description of oil and gas reservoirs and well site deployment.

Description

Technical Field

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

[0002] Currently, in the fields of oil and gas exploration and fine characterization of shale reservoirs, acoustic impedance retrieval technology is a core method for revealing reservoir lithology, porosity, and fluid saturation. However, existing methods face the following technical bottlenecks in practical applications:

[0003] 1. Hyperparameter dependence and local optima problem in traditional inversion methods

[0004] Traditional methods, such as sparse pulse inversion and model-constrained inversion, require manual adjustment of multiple regularization hyperparameters (e.g., weighting coefficients) to balance the stability and resolution of the inversion results. In complex geological conditions with large lateral reservoir variations and scarce well logging data, prior estimation errors of hyperparameters can cause the inversion results to deviate from the true geological model. Furthermore, the heterogeneity of the subsurface medium, seismic data noise, and thin interbedded layers impair the ability of traditional methods to distinguish subtle reservoir features (e.g., sandstone / shale pinch-out points), making the inversion results prone to getting trapped in local optima.

[0005] 2. Limitations of Deep Learning Methods

[0006] Although deep learning technology has been introduced to improve inversion accuracy, it relies on massive amounts of well logging data to train the model, while the number of wells in actual work areas is limited, resulting in insufficient model generalization ability. At the same time, the complexity of high-dimensional feature extraction can easily lead to overfitting problems, and the physical meaning of the inversion results is unclear, making it difficult to guide actual exploration decisions.

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

[0008] While existing intelligent optimization algorithms (such as simulated annealing and genetic algorithms) avoid initial model dependencies through global search mechanisms, they suffer from slow convergence speeds, high computational costs, and uneven initial population distribution, which can easily lead to local extremum traps. For example, in low-porosity and low-permeability reservoir applications, traditional intelligent algorithms often result in porosity prediction errors exceeding 15% due to insufficient iterations.

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

[0010] Avoiding hyperparameter dependence: avoiding manual adjustment of weight coefficients and realizing a fully data-driven inversion process;

[0011] Global optimization capability: Breaking through the limitations of local optima, accurately identifying complex reservoir characteristics (such as low-resistivity gas layers and thin sand body boundaries);

[0012] High efficiency and noise resistance: It converges stably within a finite number of iterations and is highly robust to noise and lateral inhomogeneity.

[0013] Low data dependence: High-precision prediction can still be achieved even when well logging samples are scarce.

[0014] In recent years, swarm intelligence optimization algorithms have shown potential in the field of nonlinear inversion due to their simple parameters and strong global optimization capabilities. However, the shortcomings of existing algorithms in initial population distribution strategies and convergence mechanisms limit their accuracy and efficiency in complex reservoir prediction.

[0015] Based on the aforementioned technological gaps, this invention proposes an innovative solution that overcomes the hyperparameter sensitivity, local optimum traps, and computational efficiency bottlenecks of traditional methods by integrating a chaotic mapping initialization strategy with an adaptive parameter update mechanism, thus providing reliable technical support for high-resolution description of high-quality shale reservoirs. Summary of the Invention

[0016] This invention addresses the shortcomings of existing technologies by providing 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 utilizing the whale optimization algorithm to globally optimize the wave impedance model, the hyperparameter dependence and local optima problems in traditional inversion methods are resolved. Simultaneously, Circle chaotic mapping is employed to optimize the initial population distribution, further enhancing the algorithm's global search capability and inversion accuracy. Finally, by combining the inverted wave impedance with well logging data and rock physical analysis results, a statistical regression model of wave impedance and porosity is established, achieving high-resolution prediction of high-quality shale reservoirs.

[0017] To achieve the above-mentioned objectives, the technical solution adopted by the present invention is as follows:

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

[0019] S1: The objective function is established based on the residual norm of post-stack seismic data and synthetic records, expressed as:

[0020] Cost(Z) = ||(df(Z))||

[0021] Where d represents the observed seismic data, and f(Z) represents the synthetic seismic record corresponding to the wave impedance Z, which is calculated through the convolution model;

[0022] S2: The wave impedance parameter population is initialized using the Circle chaotic mapping strategy. The specific formula is as follows:

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

[0024] Where i and j represent the population and dimensional coordinate indices, respectively, Ubi ,j and Lbi ,j This indicates the upper and lower boundaries of the search window.

[0025] Then, the randomly generated population is further initialized using the Circle chaotic map shown in the following equation:

[0026]

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

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

[0029] Prey search phase: When |A|>1, update the position according to a random search strategy.

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

[0031] Surrounding prey phase: 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 web attack phase: Select update strategy based on 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], where 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 shape of the logarithmic spiral, and Z... rand Z(t) represents the randomly selected individual position, and Z(t) represents the current individual position. best(t) represents the position of the best individual in the current population, e bl Represents an 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 rock physics analysis results to achieve quantitative prediction of high-quality shale reservoirs.

[0038] Furthermore, the search window boundary Ub mentioned in step S2 i,j and Lb i,j The determination includes:

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

[0040] The search window boundary is set to a range of wave impedance values ​​within ±20% of the mean, which is used to constrain population initialization.

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

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

[0043] Shale reservoirs and non-reservoirs are classified by wave impedance threshold: gas-bearing and poorly gas-bearing layers correspond to low wave impedance, while dry layers and mudstones correspond to high wave impedance.

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

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

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

[0047] Objective function construction module: Configured to build objective functions based on the residual norms of post-stack seismic data and synthetic records.

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

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

[0050] Chaotic Mapping Optimization Unit: Further optimization of the initial population using Circle chaotic mapping:

[0051] Wave impedance optimization module: configured to update wave impedance parameters through three stages of the whale optimization algorithm, including: prey search submodule: update position according to random search strategy;

[0052] Surrounding Prey Submodule: Updates position according to the optimal individual guidance strategy;

[0053] Bubble web attack submodule: Select update strategy based on probability p;

[0054] 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 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 rock physical analysis results;

[0056] Results display and interaction module: Configured to visualize 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 in the memory and executable by the processor, wherein the processor implements the above-described method for predicting high-quality shale reservoirs when executing the program.

[0058] The present invention also discloses a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for predicting high-quality shale reservoirs.

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

[0060] 1. Traditional wave impedance inversion methods rely on manual adjustment of hyperparameters such as regularization weighting coefficients, whose values ​​significantly affect the inversion results, especially in areas with insufficient well logging data or large lateral reservoir variations, easily leading to model bias. This invention, through the global optimization mechanism of the whale optimization algorithm, completely abandons hyperparameter adjustment, relying solely on seismic data to drive the objective function (Cost(Z)=∥df(Z)∥), effectively solving the hyperparameter sensitivity problem in traditional methods and significantly improving the inversion stability under complex geological conditions.

[0061] 2. This invention optimizes population initialization by introducing a Circle chaotic mapping strategy, ensuring that the initial solution uniformly covers the search space. Combined with the adaptive search, encirclement, and bubble net attack mechanisms of the whale optimization algorithm, it significantly reduces the risk of the algorithm getting trapped in local optima and accelerates convergence. Compared with traditional intelligent optimization algorithms, this invention significantly reduces the number of iterations while maintaining high accuracy, thus significantly improving computational efficiency.

[0062] 3. This invention exhibits strong robustness to low signal-to-noise ratio seismic data, effectively suppressing noise interference and stably acquiring high-precision acoustic impedance inversion results. Furthermore, its ability to characterize stratigraphic boundaries is significantly superior to traditional linear inversion methods, making it particularly suitable for the precise identification of pinch-out points in thin interbedded reservoirs and sandstone-mudstone bodies, providing a reliable technical means for describing complex reservoirs.

[0063] 4. Based on the wave impedance-porosity cross-plot relationship established through rock physical analysis, this invention can convert wave impedance inversion results into reservoir physical parameters (such as porosity) with high accuracy, enabling quantitative prediction of reservoir physical properties. Compared with traditional statistical interpolation methods, this invention demonstrates higher reliability and consistency in predicting the spatial distribution of reservoir parameters.

[0064] 5. This 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 boundaries and rock physical relationship model, it can be quickly adapted to various reservoir types such as tight sandstone and shale gas, and has broad engineering application value. Attached Figure Description

[0065] Figure 1 The logging curve for well A;

[0066] Figure 2 Based on the rock physical analysis results of well A;

[0067] Figure 3 This is a cross-plot of porosity and wave impedance in well A.

[0068] Figure 4 This is a flowchart of a method for predicting high-quality shale reservoirs based on the whale optimization algorithm, as described in an embodiment of the present invention.

[0069] Figure 5 This is a diagram showing the actual wave impedance values ​​of a portion of the Marmousi 2 model in an embodiment of the present invention;

[0070] Figure 6 This is a composite post-stack seismic record image according to an embodiment of the present invention; in the image, (a) contains no noise, and (b) has a signal-to-noise ratio of 1.

[0071] Figure 7The figure shows the impedance inversion results of the stacked wave based on the whale optimization algorithm in an embodiment of the present invention; (a) in the figure is free of noise, and (b) has a signal-to-noise ratio of 2.

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

[0073] Figure 9 This is a post-stack seismic data map of the sandstone pinch-out zone according to an embodiment of the present invention;

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

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

[0076] Figure 12 This is a cross-sectional view of the acoustic impedance inversion results of the shale high-quality reservoir prediction method according to an embodiment of the present invention;

[0077] Figure 13 This is a cross-sectional view of the porosity prediction results of the shale high-quality reservoir prediction method according to an embodiment of the present invention. Detailed Implementation

[0078] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and examples.

[0079] I. Rock Physical Analysis

[0080] Rock physics research studies the physical properties of rocks and their variation patterns, such as the P-wave velocity, S-wave velocity, and density of shale. Different rocks typically possess different physical characteristics. These physical properties are directly related to seismic characteristics; therefore, seismic inversion can extract the physical properties of subsurface rocks and their saturated fluids from seismic data. Thus, rock physics serves as a bridge between seismic characteristics and stratigraphic characteristics. Its specific role is to clarify the physical properties and corresponding seismic characteristics of various rocks through rock physics research, guiding seismic inversion, interpreting inversion results, and evaluating the quality of inversion. Reservoir A, located in the southwestern Sichuan Basin, is a typical low-porosity, low-permeability silty shale and argillaceous sandstone reservoir. Efficient, high-precision, and high-resolution identification of favorable reservoirs, and clarifying their thickness and distribution, are crucial for the detailed exploration of this reservoir. This example first analyzes and explains the necessity of conducting wave impedance inversion on reservoir A from a rock physics perspective. Figure 1The actual logging curve of well A, which runs through reservoir A, shows that the gas layer and poor gas layer mainly exhibit low impedance characteristics, while the dry layer and mudstone layer exhibit high impedance characteristics. Figure 2 Based on the actual logging curves of this well, the rock physics analysis results show that the histogram and cross plot shown in the figure can effectively distinguish between favorable shale reservoirs and non-reservoirs through wave impedance. Figure 3 The cross-plot of acoustic impedance and porosity in reservoir A, obtained from data from this well, shows a very strong correlation between porosity and acoustic impedance. The analysis results demonstrate that acoustic impedance can not only effectively distinguish between favorable and non-reservoirs, but also further characterize reservoir porosity, fully illustrating the necessity of conducting high-precision, high-resolution acoustic impedance inversion in reservoir A.

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

[0082] 2.1 Conventional Post-Stack 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 modeling equation for post-stack seismic data can be expressed as:

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

[0086] Where d is the observed seismic data, f(Z) is the nonlinear function of the seismic data with respect to the 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). It assumes that the reflection coefficient sequence of the strata is sparsely distributed, and the purpose of the inversion is to minimize the following objective function:

[0089] E=∑(r) p +λ q ∑(df(Z)) q +α 2 ∑(tZ) 2 (2)

[0090] Where 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, which are generally taken as 1 and 2 respectively, and λ and α are weighting coefficients, i.e. the hyperparameters mentioned above.

[0091] Based on formula (2), an iterative approach can be used to solve the problem. First, an initial reflection coefficient model is obtained by using a small number of pulses. Then, the model is continuously modified, and the objective function value becomes smaller and smaller. The number of pulses is continuously increased and the iteration is repeated until the inverted wave impedance model can no longer be significantly optimized. The iteration is then 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 objective function can be constructed as follows:

[0094]

[0095] in, These are the hyperparameters mentioned earlier. Represents the variance of the wave impedance parameter. Let represent the noise variance, and μ be the mean vector of the wave impedance parameters, typically taken as the initial model value. The aforementioned nonlinear inversion objective function can be solved using Taylor series expansion.

[0096] It is easy to see that both commonly used methods introduce hyperparameters, and these parameters must be set reasonably during actual inversion; otherwise, satisfactory wave impedance inversion results cannot be obtained. To avoid this problem, intelligent optimization algorithms are a very good alternative.

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

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

[0099] Cost(Z)=‖(df(Z))‖ (4)

[0100] Here, Cost represents the value function.

[0101] The Whale Optimization Algorithm is a swarm intelligence optimization algorithm developed by simulating the hunting behavior of humpback whales. The algorithm mainly consists of three stages: searching for prey, surrounding the prey, and bubble-web attack.

[0102] (1) Searching for prey: In the whale optimization algorithm, each whale represents a feasible solution Z of the objective function. i The wave impedance value of a sampling point in the wave impedance sequence to be inverted in the objective function shown in formula (4) is represented by i, which ranges from 1 to the maximum number of sampling points, i.e., the dimension D of the population. In the initial stage, based on the random walking 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 individual whale 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 time of the whale, and Z(t) represents the position of the whale at the current iteration time. They correspond to the wave impedance values ​​of a sampling point in the wave impedance sequence to be inverted at different times, and t represents the current iteration number. rand (t) represents the current position of the individual whale, a random position selected from the current population. The coefficients A and C are calculated as follows:

[0105] A = 2a· rand [0,1]-a (6)

[0106] C = 2 rand [0,1] (7)

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

[0108] (2) Encircling the prey: When |A|≤1, the whale pod begins to encircle the prey, gradually approaching the position of the whale with the optimal fitness value (corresponding to the minimum value function value). This process of gradually encircling and shrinking the prey, approaching it, and determining the target location can be represented as:

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

[0110] Among them, Z best (t) represents the optimal position of the whale at the current iteration number.

[0111] (3) Bubble Web Attack: This is a unique hunting strategy of whales. While exhaling bubbles, they spiral upwards, surrounding and bringing their prey close to the surface, ultimately capturing it. Considering the asynchronous behavior of different whales, a random probability p is set to balance the two behaviors of encirclement and spiral bubble web hunting, and is expressed as:

[0112]

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

[0114] Based on formulas (5)-(9), global optimization based on the whale optimization algorithm can be achieved. Similar to other optimization algorithms, before implementing the optimization algorithm, it is necessary to define the population size and initialize the population. A good population initialization strategy can further improve the algorithm's convergence progress and speed. Therefore, 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] Where i and j represent the population and dimensional coordinate indices, respectively, and Ub i,j and Lb i,j This indicates the upper and lower boundaries of the search window.

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

[0118]

[0119] in, This indicates the new initialization position, and mod() represents the remainder operation.

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

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

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

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

[0124] 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.

[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 Testing of the Method of the Invention

[0127] 3.1 Synthetic Data Testing

[0128] To test the feasibility and effectiveness of the research method, this embodiment uses a partial Marmousi 2 model to conduct synthetic data testing on the method of the present invention. Figure 5 The sampled value represents the actual longitudinal wave impedance of a portion of the Marmousi 2 model, with a sampling interval of 2ms. The reflection coefficient is calculated using the actual impedance values ​​shown in the figure and convolved with a Ricker wavelet at a dominant frequency of 40Hz, resulting in the following... Figure 6 The synthetic post-stack seismic response profile is shown in figure a. To test the stability and noise resistance of the proposed post-stack wave impedance nonlinear inversion method based on the whale optimization algorithm, in... Figure 6 The noise-free seismic data shown in figure a includes random noise with a root mean square signal-to-noise ratio of 1, as follows: Figure 6 As shown in b. Based on Figure 6 The inversion tests based on the method of this invention were carried out on the noiseless data and the noisy data shown. The inversion results are as follows. Figure 7 As shown. (Through) Figure 5 and Figure 7 The comparison shows that the method of the present invention can obtain wave impedance inversion results with high accuracy, and also has good stability and noise resistance. Figure 8 A comparison of the wave impedance inversion results corresponding to CDP 100 with the actual wave impedance inversion results shows that the inversion results of the method of this invention have a good matching degree with the true values. In summary, the post-stack wave impedance nonlinear inversion method based on the whale optimization algorithm studied in this embodiment can stably and accurately obtain wave impedance inversion results, and performs excellently in terms of stability and noise resistance.

[0129] 3.2 Actual Data Testing

[0130] The target study area, reservoir A, is generally located in a shelf facies sedimentary environment. Deep-water shelf top sedimentary environments exhibit better silty mudstone and black shale reservoirs, with the reservoirs dominated by tight shale and siltstone, characterized by low porosity and low permeability. To demonstrate the advantages of the method of this invention, firstly based on... Figure 9 The post-stack seismic data shown were compared with the sparse pulse inversion method of this invention. Figure 10 and Figure 11The inversion results are shown below, obtained using the sparse pulse inversion method and the method of this invention, respectively. The comparison reveals that the wave impedance inversion result based on the method of this invention more clearly characterizes the pinch-out points of sand bodies marked in the elliptical region, demonstrating better performance and fully validating the effectiveness and advantages of this method in predicting reservoir A. Based on the above conclusions, this embodiment combines rock physical analysis results to predict reservoir A. Figure 12 The well profile obtained based on the wave impedance inversion results of this invention is shown below. Comparison with the inserted logging curves reveals that this invention can obtain wave impedance values ​​with both satisfactory rationality and accuracy. Reservoir A's reservoir space mainly consists of intragranular dissolution pores and micropores, with effective reservoir porosity distributed between 4.0% and 13.9%, an average of 6.6%, and a median of 6.4%. Based on... Figure 3 The porosity vs. acoustic impedance cross plot shown in this embodiment further demonstrates reservoir porosity prediction based on high-precision acoustic impedance inversion results. The prediction results are as follows: Figure 13 As shown, the inserted curves on the profile are porosity curves. Analysis of the prediction results shows that the surface porosity curves are in good agreement with the predicted results. Overall, the porosity of reservoir A is approximately 4.5%–8.2%. Sections one and two have relatively good physical properties, while section three, as seen from the surface curves, has lower porosity. Comparing the physical properties of the upper, middle, and lower sections, sections one and two have good properties, with average porosity around 5%–10%, while section three has relatively poor properties, with a predicted porosity of approximately 4.5%–6.1% in the main drilling platform area. The above prediction results are highly consistent with the actual drilling results, fully verifying the feasibility and effectiveness of the method of this invention.

[0131] IV. Conclusion

[0132] Post-stack impedance retrieval is a crucial tool for fine reservoir exploration. Rock physical analysis results indicate that impedance retrieval can effectively distinguish between favorable and non-reservoirs in the Qiongzhusi Formation and further characterize reservoir porosity, highlighting the necessity of conducting post-stack impedance retrieval for reservoir A. Therefore, to improve the rationality and accuracy of reservoir A prediction results, this embodiment addresses the practical problems faced in target reservoir parameter inversion by developing a high-precision post-stack impedance nonlinear inversion method based on the whale optimization algorithm. A circle chaotic mapping population initialization strategy is introduced to further improve the convergence accuracy and speed of the inversion method. Synthetic data testing results demonstrate that the method can stably and accurately obtain impedance retrieval results. Actual data testing results show that the proposed method effectively overcomes the shortcomings of traditional methods to obtain more reasonable and satisfactory impedance retrieval results. Ultimately, based on the corresponding inversion results, a fine characterization of reservoir A is achieved, fully validating 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. This system can be used to implement the above-described shale high-quality reservoir prediction method. Specifically, it includes the following modules:

[0134] Objective function construction module: Configured to build objective functions based on the residual norms of post-stack seismic data and synthetic records.

[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 optimization of the initial population using Circle chaotic mapping:

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

[0139] Surrounding the prey submodule: When |A|≤1, update the position according to the optimal individual guidance strategy;

[0140] Bubble web attack submodule: Select update strategy based on probability p;

[0141] 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 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 physical analysis results, and achieve quantitative prediction of high-quality shale reservoirs through the following steps:

[0143] Based on the wave impedance threshold, low-impedance reservoirs (gas-bearing layers, poor-quality gas-bearing layers) and high-impedance non-reservoirs (dry layers, mudstone) are classified.

[0144] By utilizing the cross-plot relationship between wave impedance and porosity, the wave impedance inversion results are converted into a porosity distribution profile.

[0145] Results Display and Interaction Module: Configured to visualize the inversion results and provide user interaction functionality, specifically including:

[0146] Visualization submodule: Generates wave impedance inversion profile (corresponding) Figure 7 , Figure 9 Porosity distribution profile ( Figure 12 ) and floor plan ( Figure 13 );

[0147] Supports rendering of 3D geological models and dynamically displays the spatial distribution of reservoirs;

[0148] Interactive control submodule: Provides a dynamic parameter adjustment interface, including a wave impedance threshold slider and a porosity range filter;

[0149] It supports simultaneous linkage of multiple views. When a user clicks on any position in the profile view, it automatically locates the corresponding 3D model coordinates and well logging curve.

[0150] Integrated data export function, outputting inversion results as SEGY format seismic data volume or CSV format porosity table.

[0151] In another embodiment of the present invention, a terminal device is provided, comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or 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 and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to achieve a corresponding method flow or corresponding function. The processor described in this embodiment of the present invention can be used in the operation of a shale high-quality reservoir prediction method, including the following steps:

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

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

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

[0155] 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.

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

[0157] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory). This computer-readable storage medium is a memory device in a terminal device used to store programs and data. It is understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and extended storage media supported by the terminal device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more computer programs (including program code). It should be noted that the computer-readable storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device.

[0158] One or more instructions stored in a 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 embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor in the following steps:

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

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

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

[0162] 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.

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

[0164] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0165] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0166] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0167] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0168] Those skilled in the art will recognize that the embodiments described herein are intended to help the reader understand the implementation methods of the present invention, and should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific modifications and combinations based on the technical teachings disclosed in this invention without departing from the spirit of the invention, and these modifications and combinations are still within the scope of protection of the present invention.

Claims

1. A method for predicting high-quality reservoirs of shale based on a whale optimization algorithm, characterized by, The method comprises the following steps: S1: establishing a target function based on the residual norm of the post-stack seismic data and the synthetic record, expressed as: Cost(Z) = ‖(d-f(Z))‖ Wherein, d is the observed seismic data, f(Z) is the synthetic seismic record corresponding to the wave impedance Z, which is obtained by convolution of the model; S2: initializing the wave impedance parameter population by using the Circle chaotic mapping strategy, and the specific formula is: Z i,j = (Ub i,j - Lb i,j ) · rand[0, 1] + Lb i,j where i and j represent population and dimension coordinate indices, respectively, Ub i,j and Lb i,j denote the upper and lower boundaries of the search window. Then, the randomly generated population is further initialized by using the Circle chaotic mapping shown in the following formula: wherein represents a new initial position, mod() represents a remainder operation; S3: updating the wave impedance parameters through the three stages of the whale optimization algorithm: Search prey stage: when |A|>1, the position is updated according to the random search strategy: Z(t+1) = Z rand (t) - A · |C · Z rand (t) - Z(t) | Surround prey stage: when |A|≤1, the position is updated according to the optimal individual guidance strategy: Z(t+1) = Z best (t) - A · |C · Z best (t) - Z(t) | Bubble net attack stage: the spiral update strategy is selected according to the probability p: Z(t + 1) = { |Z best (t) - Z(t) | · e bl · cos(2πl) + Z best (t) wherein A represents an adaptive coefficient, A = 2a•rand[0,1]-a, C represents a random coefficient, C = 2•rand[0,1], a represents a control parameter attenuation factor, T is the maximum number of iterations, t is the iteration number, l is a random number between 0 and 1, b is a constant for defining a logarithmic spiral shape, Z rand (t) represents a randomly selected individual position, Z(t) represents a current individual position, Z best (t) represents the position of the optimal individual in the current population, e bl represents an exponential function; S4: in the iteration process, the population is dynamically updated according to the target function until the preset convergence criterion is met or the maximum iteration number is reached, and the wave impedance model corresponding to the minimum value of the target function is output; S5: combining the wave impedance model with the logging data and the petrophysical analysis results to realize quantitative prediction of the shale high-quality reservoir.

2. The method of predicting sweet shale reservoirs of claim 1, wherein: The search window boundary Ub in step S2 i,j and Lb i,j The determination includes: Based on the wave impedance logging curve of at least one well in the target work area, the mean value and the standard deviation are counted; The search window boundary is set as the wave impedance value range of the mean value ±20% for constraining the population initialization.

3. The method of predicting sweet shale reservoirs of claim 1, wherein: The parameter setting of the whale optimization algorithm in step S3 includes: the population number is 50-100; the maximum iteration number T=200; the probability p=0.5, used to balance the surrounding contraction and spiral bubble net attack strategy.

4. The method of predicting sweet shale reservoirs of claim 1, wherein: The reservoir prediction in step S5 includes: Dividing the sandstone reservoir and the non-reservoir through the wave impedance threshold value: the gas layer and the poor gas layer correspond to low wave impedance, and the dry layer and the mudstone correspond to high wave impedance; Based on the linear regression relationship between the wave impedance and the porosity, the wave impedance inversion result is converted into the porosity distribution profile.

5. The method of predicting sweet shale reservoirs of claim 1, wherein: The convergence condition in step S3 is that the change rate of the target function value is less than 1% for 10 consecutive iterations, or the maximum iteration number T is reached.

6. A system for shale quality reservoir prediction based on whale optimization algorithm characterized in that: The system can be used to implement the shale high-quality reservoir prediction method in any one of claims 1-5; Specifically includes the following modules: The target function construction module is configured to establish a target function based on the residual norm of the post-stack seismic data and the synthetic record The population initialization module is configured to initialize the wave impedance parameter population by using the Circle chaotic mapping strategy, specifically including: The initial population generation unit generates a random population based on the search window boundary: The chaotic mapping optimization unit further optimizes the initial population by using the Circle chaotic mapping: The wave impedance optimization module is configured to update the wave impedance parameters through the three stages of the whale optimization algorithm, including: The search prey submodule updates the position according to the random search strategy; The surrounding prey submodule updates the position according to the optimal individual guidance strategy; The bubble net attack submodule selects the spiral update strategy according to the probability p; The iteration control and output module is configured to dynamically update the population according to the target function in the iteration process until the preset convergence criterion is met or the maximum iteration number is reached, and output the wave impedance model corresponding to the minimum value of the target function. a reservoir prediction module configured to combine the wave impedance model with well log data and petrophysical analysis results; a result display and interaction module configured to visualize the inversion results and provide user interaction functions.

7. A computer device, characterized by: A computer program product comprising a memory, a processor, and a computer program stored on the memory and executable by the processor, the processor implementing the shale quality reservoir prediction method of one of claims 1 to 5 when executing the program.

8. A computer-readable storage medium, characterized in that: A computer program product comprising a memory, a processor, and a computer program stored on the memory and executable by the processor, the processor implementing the shale quality reservoir prediction method of one of claims 1 to 5 when executing the program.

Images