A method, apparatus, electronic device, and storage medium for predicting dessert distribution.

Abstract

This invention discloses a method, apparatus, electronic device, and storage medium for predicting sweet spot distribution. The method includes constructing a high-resolution sequence grid based on one-dimensional well data; determining the configuration unit splicing pattern based on two-dimensional well network data and a three-dimensional geological model; constructing sand body spatial distribution characteristics based on the high-resolution sequence grid, reservoir configuration unit splicing pattern, and pre-determined three-dimensional seismic data; constructing reservoir spatial distribution characteristics based on one-dimensional well data and three-dimensional seismic data; and using the sand body spatial distribution characteristics and reservoir spatial distribution characteristics to predict the distribution of sweet spot areas in low-porosity and permeability reservoirs in offshore sparse-well-network oilfields. This technical solution comprehensively considers the use of information from different dimensions in predicting the distribution of sweet spot areas in low-porosity and permeability reservoirs. By comprehensively utilizing multiple data sources to characterize the distribution characteristics of sweet spot areas in low-porosity and permeability reservoirs, it reduces the ambiguity of inter-well predictions for sand bodies and overcomes the problem of strong uncertainty in sweet spot distribution prediction caused by the small number of wells and large well spacing in offshore sparse-well-network oilfields.

Description

Technical Field

[0001] This invention relates to the field of reservoir sweet spot prediction technology, and in particular to a method, apparatus, electronic device and storage medium for predicting sweet spot distribution. Background Technology

[0002] Reservoir sweet spot distribution prediction technology is currently widely used in improving oil and gas production capacity in the mid-to-late stages of oilfield development, studying reservoir heterogeneity, and tapping remaining oil potential. By studying the distribution characteristics of sweet spots in low-porosity and permeability reservoirs, it has been found that they are mainly controlled by the three-dimensional spatial development characteristics such as the development type, combination pattern, and distribution morphology of reservoir configuration units of different levels.

[0003] Currently, the prediction of sweet spot distribution in low-porosity reservoirs in offshore sparse-well network oilfields typically employs inter-well sandbody prediction methods based on geological models. However, due to the strong heterogeneity of Paleogene continental low-porosity reservoirs, existing methods cannot effectively characterize sandbodies with rapid lateral variations and complex vertical stacking relationships. Furthermore, the sweet spot distribution predictions based on geological models have significant discrepancies with actual geological understanding in the later stages of oilfield development, affecting the deployment of development well networks and the implementation of development plans. Simultaneously, the ambiguity of inter-well sandbody predictions often leads to high uncertainty in the predicted sweet spot distribution results, resulting in substantial residual oil during oilfield development. This is a significant reason for the low oil and gas recovery rate in low-porosity reservoirs in offshore sparse-well network oilfields.

[0004] In summary, the existing methods for predicting the distribution of sweet spots in low-porosity and permeable reservoirs in offshore sparse well network oilfields are not well adapted to the highly heterogeneous continental Paleogene strata, and cannot solve the ambiguity of predictions between geological models and sand bodies. As a result, the predicted sweet spot distribution results differ significantly from the actual sweet spot distribution, which has a negative impact on the implementation of oilfield development plans and the degree of oil and gas recovery. Summary of the Invention

[0005] This invention provides a method, device, electronic device, and storage medium for predicting sweet spot distribution. It comprehensively considers the use of information from different dimensions in predicting the distribution of sweet spots in low-porosity reservoirs. By comprehensively utilizing multiple data, it accurately characterizes the distribution features of sweet spots in low-porosity reservoirs, reduces the ambiguity of predictions between sandstone wells, and overcomes the problem of strong uncertainty in sweet spot distribution prediction caused by the small number of wells and large well spacing in offshore sparse well network oilfields.

[0006] According to one aspect of the present invention, a method for predicting dessert distribution is provided, the method comprising:

[0007] A high-resolution sequence lattice is constructed based on one-dimensional well point data; and the splicing pattern of configuration units is determined based on two-dimensional well network data and three-dimensional geological model; wherein, the one-dimensional well point data includes natural gamma curves, density curves and porosity curves; the two-dimensional well network data is composed of one-dimensional well point data in different directions; and the three-dimensional geological model is obtained by stacking pre-determined sand body data.

[0008] The spatial distribution characteristics of sand bodies are constructed based on the high-resolution sequence lattice, reservoir configuration unit splicing pattern, and predetermined three-dimensional seismic data.

[0009] Based on the one-dimensional well point data and three-dimensional seismic data, the spatial distribution characteristics of the reservoir are constructed.

[0010] Using the spatial distribution characteristics of sand bodies and reservoirs, the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields is predicted.

[0011] According to another aspect of the present invention, a dessert distribution prediction device is provided, the device comprising:

[0012] A high-resolution sequence lattice and configuration unit splicing pattern determination module is used to construct a high-resolution sequence lattice based on one-dimensional well point data; and to determine the configuration unit splicing pattern based on two-dimensional well network data and a three-dimensional geological model; wherein, the one-dimensional well point data includes natural gamma curves, density curves, and porosity curves; the two-dimensional well network data is composed of one-dimensional well point data in different directions; and the three-dimensional geological model is obtained by stacking pre-determined sand body data;

[0013] The sand body spatial distribution feature construction module is used to construct the sand body spatial distribution features based on the high-resolution sequence lattice, reservoir configuration unit splicing pattern and predetermined three-dimensional seismic data.

[0014] A reservoir spatial distribution feature construction module is used to construct reservoir spatial distribution features based on the one-dimensional well point data and three-dimensional seismic data.

[0015] The reservoir sweet spot prediction module is used to predict the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields by utilizing the spatial distribution characteristics of the sand bodies and reservoirs.

[0016] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0017] At least one processor; and

[0018] A memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform a dessert distribution prediction method according to any embodiment of the present invention.

[0020] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement a dessert distribution prediction method according to any embodiment of the present invention.

[0021] The technical solution of this invention involves constructing a high-resolution sequence stratigraphic framework based on one-dimensional well point data; determining the configuration unit splicing pattern based on two-dimensional well network data and three-dimensional geological models; constructing sand body spatial distribution characteristics based on the high-resolution sequence stratigraphic framework, reservoir configuration unit splicing pattern, and pre-determined three-dimensional seismic data; constructing reservoir spatial distribution characteristics based on one-dimensional well point data and three-dimensional seismic data; and using the sand body spatial distribution characteristics and reservoir spatial distribution characteristics to predict the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse-well-network oilfields. This technical solution comprehensively considers the use of information from different dimensions in the prediction of sweet spot distribution in low-porosity and permeability reservoirs. By comprehensively utilizing multiple data to characterize the distribution characteristics of sweet spots in low-porosity and permeability reservoirs, it reduces the ambiguity of inter-well prediction of sand bodies and overcomes the problem of strong uncertainty in sweet spot distribution prediction caused by the small number of well points and large well spacing in offshore sparse-well-network oilfields.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a dessert distribution prediction method provided in Embodiment 1 of the present invention;

[0025] Figure 2 This is a flowchart illustrating a method for predicting the distribution of sweet spots in low-porosity and permeable reservoirs in offshore sparse well network oilfields based on multi-dimensional information fusion, as provided in Embodiment 1 of the present invention.

[0026] Figure 3 This is the principle of GR-INPEFA curve division of quasi-sequence groups provided in Embodiment 1 of the present invention;

[0027] Figure 4 This is a schematic diagram of determining the initial phase of original seismic data based on the large variance mode criterion provided in Embodiment 1 of the present invention;

[0028] Figure 5 This is the new image fusion based on frequency division provided in Embodiment 1 of the present invention;

[0029] Figure 6 These are the original sand body properties and edge detection cascade properties provided in Embodiment 1 of the present invention;

[0030] Figure 7 This is a schematic diagram of the seismic multi-attribute fusion principle based on extreme learning machine provided in Embodiment 1 of the present invention;

[0031] Figure 8 This is the three-dimensional lithofacies model provided in Embodiment 1 of the present invention;

[0032] Figure 9 This is a planar distribution feature diagram of a certain reservoir segment provided in Embodiment 1 of the present invention;

[0033] Figure 10 This is the three-dimensional reservoir model provided in Embodiment 1 of the present invention;

[0034] Figure 11 This is a schematic diagram of the structure of a dessert distribution prediction device provided in Embodiment 2 of the present invention;

[0035] Figure 12 This is a schematic diagram of the structure of an electronic device that implements a dessert distribution prediction method according to an embodiment of the present invention. Detailed Implementation

[0036] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0037] It should be noted that the terms "candidate," "target," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0038] Example 1

[0039] Figure 1 This is a flowchart of a sweet spot distribution prediction method according to Embodiment 1 of the present invention. This embodiment is applicable to predicting the distribution of sweet spot zones in low-porosity and permeability reservoirs in offshore sparse-well-network oilfields. The method can be executed by a sweet spot distribution prediction device, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes:

[0040] S110. Constructing a high-resolution sequence lattice based on one-dimensional well point data; and determining the splicing pattern of configuration units based on two-dimensional well network data and three-dimensional geological model; wherein, the one-dimensional well point data includes natural gamma curves, density curves and porosity curves; the two-dimensional well network data is composed of one-dimensional well point data in different directions; and the three-dimensional geological model is obtained by stacking pre-determined sand body data.

[0041] In this embodiment, one-dimensional well point data, two-dimensional well network data, and three-dimensional geological models can be obtained in advance through seismic exploration.

[0042] The sequence framework includes a longitudinal framework and a transverse framework. The longitudinal framework is established by analyzing the evolution and cycles of sedimentary strata over time; the transverse framework is composed of the spatial distribution patterns of sequence strata.

[0043] In this scheme, reservoir configuration unit analysis involves dividing outcrop cross sections into lithofacies, interfaces, and configuration units based on paleocurrent data to reveal the three-dimensional distribution of the sedimentary system and reconstruct its evolutionary history. Reservoir configuration refers to the morphology, scale, orientation, and superposition relationships of reservoir constituent units of different orders.

[0044] Specifically, Figure 2This is a flowchart illustrating a method for predicting the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields based on multi-dimensional information fusion, as provided in Embodiment 1 of the present invention. Figure 2 As shown, a high-resolution sequence grid can be constructed by analyzing one-dimensional well point data; and the splicing pattern of configuration units can be determined by analyzing two-dimensional well network data and three-dimensional geological models.

[0045] Optionally, a high-resolution sequence lattice is constructed based on one-dimensional well point data, including steps A1-A4:

[0046] Step A1: Perform median filtering on the natural gamma curves in the one-dimensional well point data to obtain candidate natural gamma curves;

[0047] Step A2: Perform maximum entropy spectrum analysis on the candidate natural gamma curves to obtain the target natural gamma curve;

[0048] Step A3: Perform error analysis and filtering analysis on the target natural gamma curve to obtain the processed target natural gamma curve;

[0049] Step A4: Construct a high-resolution sequence lattice based on the processed target natural gamma curve.

[0050] In this scheme, the construction of a high-resolution sequence stratigraphic framework is fundamental to the fine characterization of reservoirs. According to Milankovitch theory, sedimentary cycles in continuously deposited strata are gradual, with high periodic peaks corresponding to Milankovitch cycles on the spectral map. In the sedimentary background of continental low-porosity reservoirs, due to the instability of sediment supply and the strong uncertainty of base level changes, sedimentary discontinuities often exist, disrupting the cyclicity of sediments. Error filtering analysis (INPEFA) can be used to extract cyclic information at different levels, achieving the purpose of stratigraphic correlation at different levels.

[0051] Specifically, taking the sandstone and mudstone strata in the study area as an example, the natural gamma curves showed a high correlation with lithological variation trends. First, natural gamma curves with good lithological response were selected, and median filtering was applied to remove unreasonable values. Then, maximum entropy spectrum analysis was performed on the natural gamma curves to further improve their resolution. Finally, error analysis and filtering analysis were conducted on the natural gamma curves, and a high-resolution sequence lattice was constructed based on the processed natural gamma curves.

[0052] In this embodiment, Figure 3 This is the principle of GR-INPEFA curve division of quasi-sequence groups provided in Embodiment 1 of the present invention, such as... Figure 3 As shown, the positive trend of the processed natural gamma curve represents water transgression, the negative trend represents water regression, the negative inflection point represents the maximum ocean flooding, and the positive inflection point represents the sequence boundary. Based on this, a high-resolution sequence lattice is constructed.

[0053] By constructing a high-resolution sequence lattice, the spatial distribution characteristics of sand bodies in reservoirs can be constructed based on the high-resolution sequence lattice.

[0054] Optionally, the splicing pattern of the configuration units is determined based on two-dimensional well network data and three-dimensional geological models, including steps B1-B2:

[0055] Step B1: Determine the reservoir sedimentary development characteristics based on two-dimensional well network data and three-dimensional geological models;

[0056] Step B2: Analyze the sedimentary development characteristics of the reservoir, determine the different hierarchical configuration units, and analyze the different hierarchical configuration units to determine the splicing pattern of the configuration units.

[0057] In this scheme, in order to clarify the development characteristics of low-porosity and permeability reservoirs, it is necessary to address the issue that the well spacing in offshore oilfields is greater than the development scale of a single reservoir configuration unit, and to finely characterize the reservoir configuration unit.

[0058] Specifically, taking a low-porosity, low-permeability reservoir in a certain section of the Paleogene continental facies of an offshore oilfield as an example, the key to identifying different reservoir configuration units using one-dimensional well point data lies in rock electrical calibration. By analyzing the development characteristics of sedimentary facies in a single well, it is considered that the fifth-order configuration unit is a composite channel zone, characterized by a combination of box-shaped, bell-shaped, and funnel-shaped curves, with thick mudstone contact at both the top and bottom interfaces; the fourth-order configuration unit can be divided into distributary channels, distributary bars, and braided channels. Among them, the single-layer thickness of distributary channels is generally 2-8m, with an average thickness of 5m. The electrical logging curves are mostly bell-shaped, box-shaped, and box-bell-shaped combinations. The natural gamma curve of a single-stage channel is bell-shaped, while the vertical superposition of multiple-stage channels is box-shaped; the single-layer thickness of a distributary bar is generally 2-6m, with an average thickness of about 4m. The natural gamma curve is funnel-shaped and has a reverse-cycle characteristic; the natural gamma curve of a braided channel is bell-shaped, and the thickness of a single channel is generally 5-8m. Based on the identification of reservoir configuration unit types and characteristics in different strata, configuration units are divided for all single wells in the study area, which yields the development characteristics of reservoir configuration units in each stratum.

[0059] Furthermore, based on the source-sink system research method, the main source direction and sediment transport characteristics are first analyzed. Methods such as sandstone percentage content and special mineral content methods can be used to reconstruct the source direction, suggesting that the study area mainly receives sediment from the Lufeng East Low Uplift. Secondly, the sedimentary paleogeography and tectonic setting should be re-examined. It is believed that sandbars along the source direction in the study area are mostly superimposed, exhibiting characteristics of composite sand bodies where rivers flow over the sandbars. Based on the source and paleogeographic studies, and on the division of different structural units at different levels in single wells, through comparison of two-dimensional well networks and three-dimensional geological models, it is believed that the fourth-level structural units have two splicing patterns: distributary sandbar-channel overlap and distributary sandbar-distributary sandbar overlap.

[0060] By determining the splicing pattern of the configuration units, the spatial distribution characteristics of the sand bodies in the reservoir can be constructed based on the splicing pattern of the configuration units.

[0061] S120. Construct the spatial distribution characteristics of sand bodies based on the high-resolution sequence lattice, reservoir configuration unit splicing pattern, and predetermined three-dimensional seismic data.

[0062] In this scheme, to clarify the longitudinal stacking relationship and planar distribution characteristics of sand bodies, phase conversion and frequency fusion techniques are used to analyze 3D seismic data and conduct sand body profile and planar studies respectively. For example... Figure 2 As shown, the spatial distribution characteristics of sand bodies can be constructed based on sand body profiles, sand body planes, high-resolution sequence lattice frameworks, and reservoir configuration unit splicing patterns.

[0063] Optionally, the spatial distribution characteristics of sand bodies are constructed based on the high-resolution sequence framework, reservoir configuration unit splicing pattern, and predetermined three-dimensional seismic data, including steps C1-C3:

[0064] Step C1: Process the three-dimensional seismic data using phase conversion technology, establish the correspondence between seismic reflection phase axes and lithology, and construct the sand body profile distribution based on the correspondence between seismic reflection phase axes and lithology.

[0065] Step C2: Process the 3D seismic data using frequency division fusion technology, edge detection technology, and machine learning to establish the planar features of the configuration units, and construct the planar distribution of sand bodies based on the planar features of the configuration units;

[0066] Step C3: Based on the high-resolution sequence lattice, reservoir configuration unit splicing pattern, sand body profile distribution, and sand body planar distribution, construct the spatial distribution characteristics of the sand body.

[0067] In this embodiment, the seismic data used in actual work is usually zero-phase or near-zero-phase data. However, the correspondence between the seismic phase axis and the geological body in zero-phase data is not intuitive. Furthermore, when the vertical resolution of the geological body thickness is less than 1 / 4 wavelength, the interference effect is strong, making it impossible to distinguish the top and bottom interfaces of the geological body. In this case, -90° phase seismic data has a better interpretable advantage, as its effective correspondence range for the sand body center is 0-3 / 4 wavelength. Therefore, phase conversion of the seismic data can better characterize the distribution characteristics of the sand body on the seismic profile and its vertical stacking relationship.

[0068] Furthermore, zero-phase seismic data is usually the result of averaging the entire dataset, which deviates from the phase of a specific segment. Therefore, it is necessary to determine the initial phase of the target segment. The method is to perform constant phase correction on the target segment and continuously adjust the phase shift. The optimal phase shift is determined by the large variance modulus criterion, and the initial phase of the target segment is thus determined. Figure 4 This is a schematic diagram of determining the initial phase of original seismic data based on the large variance modulus criterion provided in Embodiment 1 of the present invention, as shown below. Figure 4 As shown, the initial phase a° of the seismic data of the target layer is determined. Based on this, the seismic data volume with a phase of -90° can be obtained by shifting (-90-a)°, thus establishing the correspondence between the seismic reflection phase axis and the lithology.

[0069] Furthermore, such as Figure 2 As shown, after establishing the correspondence between seismic reflection phase axes and lithology, the distribution of sand body profiles can be constructed by analyzing the correspondence between seismic reflection phase axes and lithology.

[0070] Optionally, frequency division fusion technology, edge detection technology, and machine learning are used to process the 3D seismic data to establish the planar features of the configuration units, and the planar distribution of the sand body is constructed based on the planar features of the configuration units, including steps D1-D3:

[0071] Step D1: Process the 3D seismic data using frequency division fusion technology to establish the planar features of the configuration units to be processed;

[0072] Step D2: Process the planar features of the configuration unit to be processed based on edge detection technology to construct the planar features of the configuration unit;

[0073] Step D3: Process the three-dimensional seismic data and the planar features of the configuration units using machine learning to construct the planar distribution of the sand body.

[0074] In this scheme, since the underground geological bodies have objective invariance, the frequency tuning effect can be used for seismic interpretation. By changing the frequency of the seismic data, the sand bodies of a specific thickness range can be highlighted. That is, the spectral decomposition technique is used to characterize thin sand bodies with high-frequency data and thick sand bodies with low-frequency data. Figure 5 This is a new image created through frequency division and fusion, as provided in Embodiment 1 of the present invention. Frequency division and fusion technology is a commonly used technique in the field of image processing and analysis, such as... Figure 5As shown, by assigning the three primary colors of red, green, and blue to three different images, a new image is generated. The information in the new image is a comprehensive reflection of the three original images. The new image obtained using frequency division fusion technology can clearly display sand bodies of different thicknesses. The color values ​​and brightness correspond to the thickness variations of the sand bodies: white, yellow, and orange represent thicker sand bodies, black represents thinner sand layers, and green, blue, and brown represent medium-thickness sand bodies. Combining the planar distribution characteristics of sand bodies of different thicknesses obtained by frequency division fusion technology with the sand body profile distribution obtained by phase conversion technology can clearly depict the distribution range of sand bodies on a plane.

[0075] Furthermore, when using the attribute values ​​of planar maps to delineate the pinch-out line of a sand body, adjustments to the color scale can affect the boundary position when boundary features are not obvious, leading to a strong degree of subjectivity. Edge detection cascaded attributes can sharpen the boundary and mitigate this issue's impact on interpretation. An edge detection algorithm is used to process the original attributes, identifying the locations of the most significant local changes in the original attributes as edges. The result is a new attribute that inherits and highlights the fundamental information of the planar variations in the original seismic attributes. Figure 6 These are the original sand body properties and edge detection cascaded properties provided in Embodiment 1 of the present invention. Figure 6 (a) represents the original properties of the sand body. Figure 6 (b) represents edge detection cascade attributes. For example... Figure 6 (a) and Figure 6 As shown in (b), taking a quasi-sequence single sand body as an example, the Sobel algorithm is used to perform edge detection on the average wave valley amplitude. Compared with the original attributes, the new attributes obtained are messy and cannot display effective information, but their edges are significantly sharpened and highlighted, providing a more objective basis for the characterization of the sand body boundary. That is, the planar features of the structural unit to be processed can be processed by the edge detection technology to construct the planar features of the structural unit.

[0076] In this embodiment, the integration of machine learning and seismic attribute technology can go from qualitatively explaining the relationship between seismic attributes and geological parameters to further quantitatively and semi-quantitatively characterizing the relationship between seismic attributes and geological parameters. The final result map is a planar map of geological parameters, which helps to more intuitively understand the distribution and changes of geological bodies and more effectively classify sedimentary facies.

[0077] Furthermore, Figure 7 This is a schematic diagram of the seismic multi-attribute fusion principle based on extreme learning machine provided in Embodiment 1 of the present invention, as follows: Figure 7As shown, the specific steps are as follows: Step 1, extract as many seismic attributes as possible from the target layer of the seismic data, covering different categories of seismic attributes; Step 2, calculate the mean of each seismic attribute near the well point as the feature variable for subsequent machine learning; Step 3, calculate the geological information of the target layer at the well point (such as sand body thickness, sand-mud ratio, etc.) as the target variable for machine learning, and use the extreme learning machine method, with the well point seismic attribute feature variable from Step 2 and the well point geological information target variable obtained in this step as the training set, to establish the relationship between the multiple seismic attributes of the well point and the well point geological information; Step 4, input the complete seismic attributes from Step 2 into the relationship established in Step 3, calculate the planar distribution of geological information based on the established relationship, and output the final geological map, i.e., construct the planar distribution of sand bodies.

[0078] S130. Construct reservoir spatial distribution characteristics based on the one-dimensional well point data and three-dimensional seismic data.

[0079] In this plan, such as Figure 2 As shown, spatial distribution characteristics can be constructed by analyzing one-dimensional well point data and three-dimensional seismic data. For example, a pre-trained neural network model can be used to analyze one-dimensional well point data and three-dimensional seismic data to construct spatial distribution characteristics.

[0080] Optionally, reservoir spatial distribution characteristics are constructed based on the one-dimensional well point data and three-dimensional seismic data, including steps E1-E4:

[0081] Step E1: Take the natural gamma curve, density curve and porosity curve in the one-dimensional well point data as input, process the natural gamma curve, density curve and porosity curve based on the predetermined neural network model, and output the lithofacies model.

[0082] Step E2: Based on the lithofacies model and the seismic attribute inversion volume in the three-dimensional seismic data, construct an attribute model;

[0083] Step E3: Process the natural gamma curve and density curve to determine the reservoir classification;

[0084] Step E4: Based on the attribute model and reservoir classification, construct a three-dimensional reservoir model, and construct the reservoir spatial distribution characteristics based on the three-dimensional reservoir model.

[0085] In this scheme, reservoir modeling constrained by one-dimensional well point data alone has great uncertainty in sparse well network oilfields, making it difficult to obtain an accurate three-dimensional geological model. Therefore, it is necessary to involve three-dimensional seismic data in the modeling process. Through collaborative simulation, the vertical resolution of lithofacies is improved, and the horizontal resolution is further improved, so as to achieve the effect of high-precision modeling.

[0086] Specifically, neural network technology can effectively represent and predict uncertain and unstructured data. The neural network method also provides a means to comprehensively utilize various well logging information for lithological identification. Its structure includes an input layer, a hidden layer, and an output layer. First, the well logging curves are standardized, selecting conventional well logging curves that are more sensitive to lithofacies. The natural gamma curve, density curve, and porosity curve are combined as inputs, with lithofacies calibrated at the wall center serving as supervision for the learning process. The lithofacies model is the output. Through prediction, mudstone facies, siltstone facies, fine sandstone facies, coarse sandstone facies, and gravelly coarse sandstone facies can be effectively identified, establishing identification standards for different lithofacies. Then, its neural network structure is applied to a three-dimensional geological model. The inputs become a natural gamma 3D model, a density 3D model, and a porosity 3D model. Through the constructed neural network structure, the traditional discrete lithofacies model is made continuous, and the output becomes a three-dimensional lithofacies model. Figure 8 The three-dimensional lithofacies model provided in Embodiment 1 of this invention outputs the following results: Figure 8 As shown.

[0087] Furthermore, based on the principle that lithofacies control reservoir properties, a property model is constructed using lithofacies control and well logging interpretation data of porosity and permeability from a single well, along with a synergistic Gaussian simulation method and a seismic attribute inversion model.

[0088] Furthermore, to clarify the spatial distribution of high-quality reservoirs, based on the existing porosity and permeability data interpreted from well logging, natural gamma and density logging curves were combined to create natural gamma-porosity cross plots, density-porosity cross plots, and porosity-permeability cross plots. Figure 9 This is a planar distribution feature map of a reservoir segment provided in Embodiment 1 of the present invention, such as... Figure 9 As shown, low-porosity and permeability reservoirs can be roughly divided into two major categories (Class I and Class II) and four subcategories (Class Ia, Class Ib, Class IIa, and Class IIb) based on their physical properties. Class Ia reservoirs are considered to be high-quality sweet spots, Class Ib reservoirs are high-quality sweet spot potential areas, Class IIa reservoirs are poor reservoirs, and Class IIb reservoirs are non-reservoirs.

[0089] In this embodiment, a neural network structure is applied to a three-dimensional geological model. A lithofacies model constructed using a multi-parameter neural network and an attribute model constructed using a facies-controlled co-seismic attribute inversion system are used as inputs, and the output is the three-dimensional reservoir model. Figure 10 The three-dimensional reservoir model provided in Embodiment 1 of this invention outputs the following results: Figure 10 As shown in the figure. Based on the development characteristics of various reservoirs in this model, combined with lithofacies and property models, the classification criteria for the physical property parameters of the four types of reservoirs can be quantitatively determined.

[0090] S140. Using the spatial distribution characteristics of the sand body and the spatial distribution characteristics of the reservoir, the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields is predicted.

[0091] In this plan, such as Figure 2 As shown, the spatial distribution characteristics of sand bodies and reservoirs can be compared and analyzed to predict the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields.

[0092] Optionally, the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse-well-network oilfields can be predicted using the aforementioned sand body spatial distribution characteristics and reservoir spatial distribution characteristics, including step F1:

[0093] Step F1: Compare the spatial distribution characteristics of the sand body and the spatial distribution characteristics of the reservoir to determine the correspondence, and predict the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields based on the correspondence.

[0094] Specifically, the following steps were taken to predict the distribution of sweet spots in a certain low-porosity reservoir section: (1) Using one-dimensional well point data, a high-resolution sequence grid was constructed using the Inference of Errors in Logging Curves (INPEFA); (2) Using two-dimensional well networks and three-dimensional geological models, the splicing pattern of configuration units was determined as a reference for the study of the spatial distribution characteristics of sand bodies; (3) The initial phase of the seismic data of the target section was determined to be a°, and a phase conversion of (-90-a)° was performed to establish the correspondence between the seismic reflection phase axis and the sand body, and the distribution characteristics of the sand body in the seismic profile of this section were analyzed; (4) The seismic data of this section was frequency-divided and fused to generate a new image, and the planar distribution range of the sand body was characterized by combining the sand body profile distribution range; (5) In the process of fine characterizing the sand body distribution range, edge detection technology and seismic multi-attribute fusion based on extreme learning machine were used to finely characterize the planar distribution range of the sand body, and the distribution characteristics of the sand body were analyzed. (6) Using the multi-parameter neural network method, select well logging curves with good lithological response characteristics, use the well logging curves as input, use the lithofacies calibrated by the wall core as the supervision of the learning process, use the lithofacies as the output, identify the lithofacies type, and use it in the geological model to establish a three-dimensional lithofacies model; (7) Use the lithofacies model in conjunction with the seismic attribute inversion body to construct an attribute model; (8) Make a well logging curve intersection diagram, divide the reservoir type, use the neural network, use the lithofacies model and attribute model as input to construct a three-dimensional reservoir model, and obtain the reservoir spatial distribution characteristics; (9) Check the correspondence between the reservoir spatial distribution characteristics and the sand body spatial distribution characteristics. If the correspondence is good, the high-quality reservoir is the sweet spot distribution area. Otherwise, repeat steps (3) and (4), re-describe the sand body spatial distribution characteristics using three-dimensional seismic data, and then conduct a comparison test until the correspondence is good.

[0095] The technical solution of this invention involves constructing a high-resolution sequence stratigraphic framework based on one-dimensional well point data; determining the configuration unit splicing pattern based on two-dimensional well network data and three-dimensional geological models; constructing sand body spatial distribution characteristics based on the high-resolution sequence stratigraphic framework, reservoir configuration unit splicing pattern, and pre-determined three-dimensional seismic data; constructing reservoir spatial distribution characteristics based on one-dimensional well point data and three-dimensional seismic data; and using the sand body spatial distribution characteristics and reservoir spatial distribution characteristics to predict the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse-well-network oilfields. This technical solution comprehensively considers the use of information from different dimensions in the prediction of sweet spot distribution in low-porosity and permeability reservoirs. By comprehensively utilizing multiple data to characterize the distribution characteristics of sweet spots in low-porosity and permeability reservoirs, it reduces the ambiguity of inter-well prediction of sand bodies and overcomes the problem of strong uncertainty in sweet spot distribution prediction caused by the small number of well points and large well spacing in offshore sparse-well-network oilfields. The sweet spot prediction is more consistent with the situation of strong reservoir heterogeneity and well spacing greater than the development scale of configuration units, resulting in higher accuracy in sweet spot distribution prediction.

[0096] Example 2

[0097] Figure 11 This is a schematic diagram of a dessert distribution prediction device provided in Embodiment 2 of the present invention. Figure 11 As shown, the device includes:

[0098] The high-resolution sequence lattice and configuration unit splicing pattern determination module 1110 is used to construct a high-resolution sequence lattice based on one-dimensional well point data; and to determine the configuration unit splicing pattern based on two-dimensional well network data and a three-dimensional geological model; wherein, the one-dimensional well point data includes natural gamma curves, density curves, and porosity curves; the two-dimensional well network data is composed of one-dimensional well point data in different directions; and the three-dimensional geological model is obtained by stacking pre-determined sand body data;

[0099] The sand body spatial distribution feature construction module 1120 is used to construct the sand body spatial distribution features based on the high-resolution sequence lattice, reservoir configuration unit splicing pattern and predetermined three-dimensional seismic data.

[0100] The reservoir spatial distribution feature construction module 1130 is used to construct reservoir spatial distribution features based on the one-dimensional well point data and three-dimensional seismic data.

[0101] The reservoir sweet spot prediction module 1140 is used to predict the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields by utilizing the spatial distribution characteristics of the sand bodies and the spatial distribution characteristics of the reservoirs.

[0102] Optional, the high-resolution sequence lattice and configuration unit splicing pattern determination module 1110 is specifically used for:

[0103] Median filtering is performed on the natural gamma curves in one-dimensional well point data to obtain candidate natural gamma curves.

[0104] Maximum entropy spectral analysis is performed on the candidate natural gamma curves to obtain the target natural gamma curve;

[0105] Error analysis and filtering analysis are performed on the target natural gamma curve to obtain the processed target natural gamma curve;

[0106] A high-resolution hierarchical lattice is constructed based on the processed target natural gamma curve.

[0107] Optionally, the high-resolution sequence lattice and configuration unit splicing pattern determination module 1110 is also used for:

[0108] Based on two-dimensional well network data and three-dimensional geological models, the reservoir sedimentary development characteristics were determined;

[0109] The sedimentary development characteristics of the reservoir are analyzed to determine the different hierarchical configuration units, and the splicing pattern of the configuration units is determined by analyzing the different hierarchical configuration units.

[0110] Optionally, the sand body spatial distribution feature construction module 1120 includes:

[0111] The sand body profile distribution construction submodule is used to process three-dimensional seismic data using phase conversion technology, establish the correspondence between seismic reflection phase axes and lithology, and construct the sand body profile distribution based on the correspondence between seismic reflection phase axes and lithology.

[0112] The sand body planar distribution construction submodule is used to process three-dimensional seismic data using frequency division fusion technology, edge detection technology, and machine learning, establish the planar features of configuration units, and construct the sand body planar distribution based on the planar features of the configuration units.

[0113] The sand body spatial distribution feature construction submodule is used to construct the sand body spatial distribution features based on the high-resolution sequence lattice, reservoir configuration unit splicing pattern, sand body profile distribution, and sand body planar distribution.

[0114] Optional, the sand body planar distribution construction submodule is specifically used for:

[0115] Frequency-division fusion technology is used to process 3D seismic data and establish the planar features of the configuration units to be processed;

[0116] The planar features of the configuration unit to be processed are processed based on edge detection technology to construct the planar features of the configuration unit;

[0117] The planar distribution of sand bodies is constructed by processing the three-dimensional seismic data and the planar features of the configuration units using machine learning.

[0118] The reservoir spatial distribution feature construction module 1130 is specifically used for:

[0119] The natural gamma curve, density curve, and porosity curve in the one-dimensional well point data are used as input. Based on a pre-determined neural network model, the natural gamma curve, density curve, and porosity curve are processed to output a lithofacies model.

[0120] Based on the lithofacies model and the seismic attribute inversion volume in the three-dimensional seismic data, an attribute model is constructed.

[0121] The natural gamma curve and density curve are processed to determine the reservoir classification;

[0122] Based on the attribute model and reservoir classification, a three-dimensional reservoir model is constructed, and the spatial distribution characteristics of the reservoir are constructed based on the three-dimensional reservoir model.

[0123] Optionally, the reservoir sweet spot prediction module 1140 is specifically used for:

[0124] The spatial distribution characteristics of the sand bodies and the spatial distribution characteristics of the reservoirs are compared to determine the correspondence, and the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields is predicted based on the correspondence.

[0125] The dessert distribution prediction device provided in this embodiment of the invention can execute a dessert distribution prediction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0126] Example 3

[0127] Figure 12 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0128] like Figure 12As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0129] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0130] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as a dessert distribution prediction method.

[0131] In some embodiments, a dessert distribution prediction method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the dessert distribution prediction method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform a dessert distribution prediction method by any other suitable means (e.g., by means of firmware).

[0132] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0133] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0134] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0135] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0136] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0137] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0138] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0139] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method of dessert distribution prediction, characterized by, include: A high-resolution sequence lattice is constructed based on one-dimensional well point data; and the splicing pattern of configuration units is determined based on two-dimensional well network data and three-dimensional geological model; wherein, the one-dimensional well point data includes natural gamma curves, density curves and porosity curves; the two-dimensional well network data is composed of one-dimensional well point data in different directions; and the three-dimensional geological model is obtained by stacking pre-determined sand body data. Phase conversion technology was used to process three-dimensional seismic data to establish the correspondence between seismic reflection phase axes and lithology, and sand body profile distribution was constructed based on the correspondence between seismic reflection phase axes and lithology. Frequency-division fusion technology is used to process 3D seismic data and establish the planar features of the configuration units to be processed; The planar features of the configuration unit to be processed are processed based on edge detection technology to construct the planar features of the configuration unit; The planar distribution of sand bodies is constructed by processing the three-dimensional seismic data and the planar features of the structural units using machine learning. Based on the high-resolution sequence lattice, reservoir configuration unit splicing pattern, sand body profile distribution, and sand body planar distribution, the spatial distribution characteristics of sand bodies are constructed. Based on the one-dimensional well point data and three-dimensional seismic data, the spatial distribution characteristics of the reservoir are constructed. Using the spatial distribution characteristics of sand bodies and reservoirs, the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields is predicted.

2. The method of claim 1, wherein, A high-resolution sequence lattice is constructed based on one-dimensional well point data, including: Median filtering is performed on the natural gamma curves in one-dimensional well point data to obtain candidate natural gamma curves. Maximum entropy spectral analysis is performed on the candidate natural gamma curves to obtain the target natural gamma curve; Error analysis and filtering analysis are performed on the target natural gamma curve to obtain the processed target natural gamma curve; A high-resolution hierarchical lattice is constructed based on the processed target natural gamma curve.

3. The method of claim 1, wherein, Based on two-dimensional well network data and three-dimensional geological models, the splicing style of the structural units was determined, including: Based on two-dimensional well network data and three-dimensional geological models, the reservoir sedimentary development characteristics were determined; The sedimentary development characteristics of the reservoir are analyzed to determine the different hierarchical configuration units, and the splicing pattern of the configuration units is determined by analyzing the different hierarchical configuration units.

4. The method of claim 1, wherein, Based on the aforementioned one-dimensional well point data and three-dimensional seismic data, the spatial distribution characteristics of the reservoir are constructed, including: The natural gamma curve, density curve, and porosity curve in the one-dimensional well point data are used as input. Based on a pre-determined neural network model, the natural gamma curve, density curve, and porosity curve are processed to output a lithofacies model. Based on the lithofacies model and the seismic attribute inversion volume in the three-dimensional seismic data, an attribute model is constructed. The natural gamma curve and density curve are processed to determine the reservoir classification; Based on the attribute model and reservoir classification, a three-dimensional reservoir model is constructed, and the spatial distribution characteristics of the reservoir are constructed based on the three-dimensional reservoir model.

5. The method of claim 1, wherein, Using the aforementioned spatial distribution characteristics of sand bodies and reservoirs, the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse-well-network oilfields is predicted, including: The spatial distribution characteristics of the sand bodies and the spatial distribution characteristics of the reservoirs are compared to determine the correspondence, and the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields is predicted based on the correspondence.

6. A dessert distribution prediction apparatus characterized by comprising: include: A high-resolution sequence lattice and configuration unit splicing pattern determination module is used to construct a high-resolution sequence lattice based on one-dimensional well point data; and to determine the configuration unit splicing pattern based on two-dimensional well network data and a three-dimensional geological model; wherein, the one-dimensional well point data includes natural gamma curves, density curves, and porosity curves; the two-dimensional well network data is composed of one-dimensional well point data in different directions; and the three-dimensional geological model is obtained by stacking pre-determined sand body data; The sand body spatial distribution feature construction module is used to construct the sand body spatial distribution features based on the high-resolution sequence lattice, reservoir configuration unit splicing pattern and predetermined three-dimensional seismic data. A reservoir spatial distribution feature construction module is used to construct reservoir spatial distribution features based on the one-dimensional well point data and three-dimensional seismic data. The reservoir sweet spot prediction module is used to predict the distribution of sweet spots in low-porosity and permeability reservoirs in offshore sparse well network oilfields by utilizing the spatial distribution characteristics of the sand bodies and the spatial distribution characteristics of the reservoirs. The sand body spatial distribution feature construction module includes: The sand body profile distribution construction submodule is used to process three-dimensional seismic data using phase conversion technology, establish the correspondence between seismic reflection phase axes and lithology, and construct the sand body profile distribution based on the correspondence between seismic reflection phase axes and lithology. The sand body planar distribution construction submodule is used to process three-dimensional seismic data using frequency division fusion technology and establish the planar features of the configuration unit to be processed; The planar features of the configuration unit to be processed are processed based on edge detection technology to construct the planar features of the configuration unit; The planar distribution of sand bodies is constructed by processing the three-dimensional seismic data and the planar features of the structural units using machine learning. The sand body spatial distribution feature construction submodule is used to construct the sand body spatial distribution features based on the high-resolution sequence lattice, reservoir configuration unit splicing pattern, sand body profile distribution, and sand body planar distribution.

7. An electronic device, comprising: The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform a dessert distribution prediction method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute a dessert distribution prediction method according to any one of claims 1-5.

Images