Dual-medium reservoir prediction method of hybrid vector autocorrelation function
By combining hybrid vector autocorrelation function and seismic attribute analysis with nonlinear learning and multi-attribute fusion, the problem of insufficient prediction accuracy for buried hill carbonate dual-medium reservoirs was solved, achieving high-precision prediction of underground fractured-vuggy reservoirs and assisting in oil and gas reservoir exploration.
Patent Information
- Application Number
- CN202410548216.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-06
- Publication Date
- 2025-11-07
AI Technical Summary
Existing technologies struggle to accurately predict fracture geometry and pore characteristics in buried hill carbonate dual-medium reservoirs, resulting in insufficient reservoir prediction accuracy. This is especially true in highly heterogeneous underground dual-medium reservoirs, where existing model assumptions do not conform to actual random distribution characteristics.
By employing a hybrid vector autocorrelation function, combined with seismic attribute analysis and nonlinear learning, and through forward modeling and multi-attribute fusion techniques, a dual-pore stochastic medium model capable of characterizing the random distribution of fractures and pores in underground space is established. Sensitive attributes are then optimized to achieve precise prediction of fracture-vuggy reservoirs.
It improves the accuracy of carbonate reservoir prediction and helps high-quality exploration and development of unconventional oil and gas reservoirs. Through fine modeling and multi-attribute fusion technology, it achieves effective prediction of underground dual-medium reservoirs.
Smart Images
Figure CN120908871A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of oil and gas field exploration, and particularly relates to a mixed vector autocorrelation function dual medium reservoir prediction method. BACKGROUND
[0002] The Paleozoic in Jiyang Depression is mainly formed by marine carbonate rock deposition under the influence of multi-stage tectonic movements, and the Paleozoic buried hill reservoir has good overall reservoir forming conditions, rich reservoir types and high single well productivity, and is an important reservoir position for exploration breakthrough. The reservoir is the key to determine the enrichment of such buried hill oil and gas, and the buried hill reservoir has obvious dual medium characteristics, i.e. the reservoir space is mainly composed of pores and fractures. The fracture-vug type reservoir space formed by karstification is a high-quality carbonate rock buried hill reservoir, and is also the current focus of exploration and research. However, due to the deep burial of the buried hill reservoir, strong heterogeneity of the reservoir, poor regularity of the fracture-vug scale of the underground dual medium reservoir, and no systematic and comprehensive understanding of the influence of the matrix-fracture composition on the petrophysical and seismic response characteristics of the buried hill, it is necessary to carry out innovative research on the fine modeling of the dual pore medium reservoir, the analysis of the seismic response characteristics and the multi-attribute reservoir fusion prediction method to essentially improve the prediction accuracy of the carbonate rock dual medium reservoir of the buried hill.
[0003] Hudson model (1981) and Schoenberg et al. (1980) proposed linear sliding model, based on the sliding model derived from the fracture reservoir prediction technology and fracture parameter inversion technology can generally calculate the fracture direction and density, but the model assumes that the background medium embedded in the fracture is a single elastic solid medium without pores and fluid, so that this kind of model lacks sensitivity to fracture geometry information, pore characteristics and the combination of the two reservoir space; Parra (2000) integrated BISQ model, Hudson fracture medium model, Thomsen et al. (2000) proposed a fluid flow equivalent medium geological model containing microcracks and connected pores, and described the influence of permeability difference on seismic wave dispersion and attenuation; Chapman (2009) proposed a dynamic equivalent medium model to describe the elastic properties of fractured porous medium, and extended the model to two groups of reservoir space with different directions, sizes and different connectivity, and studied the influence of fracture on the anisotropy of porous medium; Xuan Yihua, He Qideng et al. (2006) proposed a single layer double phase EDA medium wave field simulation based on BISQ mechanism, and simulated the ground seismic record of three layer double phase medium; Zhang Xianwen, Wang Deli et al. (2010) derived the dispersion equation of three-dimensional double phase anisotropic medium based on BISQ, studied the anisotropy of solid skeleton, and provided a theoretical basis for predicting the distribution of reservoir fluid and pore structure characteristics; Du Qizhen, Kong Laiyun et al. (2009) based on the theory of fracture induced anisotropy and double phase medium, obtained the equivalent porosity and permeability of HTI double porous medium, established the fracture induced HTI double porous medium model, and the double layer model is more conducive to the multi-wave seismic inspection of underground fracture distribution; Liu Cai et al. (2013) combined with the theory of fracture anisotropy, proposed a pseudo-spectrum simulation and analysis method of double phase HTI medium wave propagation, and established an equivalent medium model of local jet flow between microcracks and equal diameter pores.
[0004] The above models all consider the connectivity between directional fractures and pores and the fluid flow induced by seismic wave, but since these theories are based on directional fracture analysis, the occurrence and size of actual fracture type reservoir are randomly distributed, and the prediction of actual underground fracture and the inversion accuracy of fracture type reservoir are not enough; At the same time, considering that the fracture and pore of buried hill type carbonate reservoir generally show mutual combination and connectivity as effective fracture-pore type primary space, it is necessary to establish a random medium and superimposed fracture-pore type reservoir equivalent seismic geological model of different sizes. SUMMARY
[0005] The purpose of the present application is to provide a double medium reservoir prediction method based on a hybrid vector autocorrelation function, adopt a hybrid vector autocorrelation function, establish a continuous double pore random medium capable of effectively depicting the random distribution of underground space cracks and holes; through forward simulation, combined with seismic profiles of actual drilling wells, application of seismic attribute analysis technology, analysis of the influence of factors such as matrix pore type, hole size, fracture occurrence, filling property and pore-fracture heterogeneity on seismic response characteristics, optimization of reservoir sensitivity attributes, development characteristics of double medium reservoirs are learned, mined and extracted through nonlinear mapping, and finally the effective prediction of double medium reservoirs is realized through multi-attribute fusion technology, which essentially improves the prediction accuracy of carbonate reservoirs and helps the high-quality exploration and development of unconventional oil and gas reservoirs.
[0006] The present application is realized by the following measures: a double medium reservoir prediction method of a hybrid vector autocorrelation function, characterized by comprising: Collecting relevant data of double medium reservoirs in actual work areas; According to the collected data, a geological model close to the actual formation structure is established; Simulate random medium according to hybrid vector autocorrelation function; Establish a "stacked" multi-scale double medium model; Embed the "stacked" multi-scale double medium model into the geological model to form a superimposed model; Perform non-uniform medium forward simulation, and obtain a migration stack section through pre-stack depth migration imaging processing; Seismic attribute analysis, optimization of fracture-cave reservoir sensitive parameters; Nonlinear learning, mining and extraction of double medium reservoir characteristics; Post-stack multi-attribute prediction of double medium reservoir development zone.
[0007] Collecting relevant data of double medium reservoirs in actual work areas includes seismic, geophysical, drilling, well logging data, actual drilling conditions, and counting various lithological rock physical parameters related to carbonate fracture-cave reservoirs, and determining the velocity difference of different types of reservoirs.
[0008] According to the actual work area seismic data and actual drilling conditions, a geological interpretation scheme is established, and a geological model close to the actual formation structure of the work area is established according to the rock physical parameters.
[0009] The hybrid vector autocorrelation function is: In the formula, φ is the autocorrelation function, r is the roughness factor, x is the spatial coordinate, and a is the autocorrelation function in the x direction.
[0010] The "stacked" multi-scale double medium model is a model that stacks more scale random perturbations on large-scale average characteristics, which can be represented by the following formula: ; wherein each is a single scale random medium model of different scale, and the physical property parameters of the final multi-scale random medium model are equal to the sum of the physical property parameters of each single scale random medium model.
[0011] The sensitive parameters of the fracture-vug type reservoir are preferably selected as follows: based on the results of the dual-medium forward modeling, on the basis of the characteristics of the seismic response of the fracture-vug type reservoir of different scales, post-stack reservoir sensitive attributes are selected, and the sensitivity of different attributes to the fracture-vug type reservoir is determined.
[0012] Nonlinear learning, mining and extraction of dual-medium reservoir characteristics, including: The selected dual-medium fracture-vug type reservoir sensitive attributes are subjected to nonlinear learning, and the number of single well samples participating in the nonlinear learning cannot be too small and the well distribution cannot be too concentrated.
[0013] Post-stack multi-attribute prediction of dual-medium reservoir development zones, including, based on the selected dual-medium reservoir sensitive attributes, the fusion weight calculation formula is: ; wherein, is the weight value of different attributes in the attribute fusion calculation; is the correlation degree of the unit attribute, and n is the number of all selected attributes to be fused; After multi-attribute fusion, normalization processing is performed, and a post-stack multi-attribute fusion map is obtained according to the weight value; Based on the post-stack multi-attribute fusion map, the dual-medium reservoir development zone is predicted.
[0014] In an embodiment of the present application: a storage medium is provided, characterized in that the storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the mixed vector autocorrelation function dual-medium reservoir prediction method.
[0015] In an embodiment of the present application: an electronic device is provided, characterized in that it includes a processor and a memory, and the processor is used to execute the program of the mixed vector autocorrelation function dual-medium reservoir prediction method stored in the memory to implement the mixed vector autocorrelation function dual-medium reservoir prediction method.
[0016] The technical scheme provided by the embodiment of the present application has the beneficial effects that: the method adopts a mixed vector autocorrelation function, establishes a double-pore random medium capable of effectively describing the random distribution of underground space cracks and pores, through forward simulation, in combination with a real drilling seismic profile, application of a seismic attribute analysis technology, analysis of the influence of factors such as matrix pore types, pore sizes, crack occurrences, filling properties and pore-crack heterogeneity on seismic response characteristics, optimization of reservoir sensitivity attributes, mining and extraction of double-medium reservoir development characteristics through nonlinear mapping learning, and finally effective prediction of the double-medium reservoir by applying a multi-attribute fusion technology, which essentially improves the carbonate reservoir prediction accuracy and helps high-quality exploration and development of unconventional oil and gas reservoirs. BRIEF DESCRIPTION OF DRAWINGS
[0017] In order to more clearly illustrate the technical scheme of the present application, the drawings used in the embodiments will be briefly introduced as follows. Obviously, the drawings listed below are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.
[0018] Figure 1 The flow chart of the double-medium reservoir prediction method of the mixed vector autocorrelation function in the embodiment of the present application; Figure 2 The geological model close to the actual formation structure in the embodiment of the present application; Figure 3 The vector autocorrelation function-based random medium model in the embodiment of the present application; Figure 4 The three multi-scale random medium models simulated based on the mixed vector autocorrelation function in the embodiment of the present application; Figure 5 The “stacked type” double-medium reservoir model in the embodiment of the present application; Figure 6 The different scale crack model and pore model in the embodiment of the present application; Figure 7 The double-medium fracture-cavity type seismic geological model close to the actual reservoir in the embodiment of the present application; Figure 8 The fine reservoir random model in the embodiment of the present application; Figure 9 The fracture-cavity reservoir depth migration imaging profile in the embodiment of the present application; Figure 10 The attribute prediction map of the favorable development area of the fracture-cavity type reservoir in the research area in the embodiment of the present application. DETAILED DESCRIPTION
[0019] In order to make the purpose, technical scheme and advantages of the present application more clear, the present application is further described in detail below with examples. Of course, the specific examples described here are only used to explain the present application, and are not used to limit the present application.
[0020] Example one: Referring to Figures 1-10 A mixed vector autocorrelation function dual medium reservoir prediction method, characterized in that the specific steps include: Step S1, collecting relevant data of the actual work area dual medium reservoir; The collection of relevant data of the actual work area dual medium reservoir includes seismic data, seismic, geophysical, drilling, logging data, actual drilling conditions, and statistics of various lithological rock physical parameters related to the carbonate fracture-cave type reservoir, and the velocity difference of different types of reservoirs is clear. Taking the research area as an example, based on the time-depth relationship of the typical well ZGX473 well in the target work area and the logging curve analysis data, the formation physical parameters of the work area are obtained, as shown in Table 1.
[0021]
[0022] Table 1 Formation physical parameters of the geological model of the research area Step S2, according to the collected data, a geological model close to the actual formation structure is established. According to the actual work area seismic data and actual drilling conditions, a geological interpretation scheme is established, and according to the rock physical parameters in step S1, a geological model close to the actual formation structure of the work area is established.
[0023] Step S3, simulate the random medium according to the mixed vector autocorrelation function.
[0024] Based on the mixed vector autocorrelation function, the random medium filled in the geological model in step 2 is simulated. First, the actual autocorrelation function is obtained by using the covariance formula and the autocorrelation function formula; The actual autocorrelation function is: In the formula, φ is the autocorrelation function, r is the roughness factor, x is the spatial coordinate, and a is the autocorrelation function in the x direction.
[0025] Then use the relationship: Among them, And Can be obtained by = 0.5, , through the solution of the above equation group, the type of the random medium (the roughness factor r) and the autocorrelation length can be obtained from the known
[0026] Since the underground medium is heterogeneous, the vector autocorrelation function can introduce the dominant direction parameter into the autocorrelation function, and with the change of the dominant direction parameter, the dominant direction of the double medium model will also change, which is convenient to describe the randomness characteristics of the underground medium in different directions, such as Figure 3 , where (a) θ = π / 6; (b) θ = π / 4; (c) θ = π / 3; (d) θ = 2π / 3; (e) θ = 3π / 4; (f) θ = 5π / 6; θ is a direction factor, which is set according to the vector autocorrelation function.
[0027] Step S4, establishing a “stacked type” multi-scale double medium model; The multi-scale double medium model is a non-uniform medium model (such as Figure 4 indicated), since the heterogeneity of the actual underground medium is multi-scale, at a large scale, we describe the change of the underground medium through adjacent strata in a regional range; at a small scale, we describe the change of the underground medium through adjacent strata in a small range; at different scales, we can observe the change of the non-uniformity of the actual underground medium. This method restores the reality of the geological medium as much as possible, including the average characteristics of the medium, the large-scale non-uniformity and the small-scale non-uniformity of the random disturbance stacked on the large scale.
[0028] The “stacked type” random medium model is to stack more scale random disturbances on the large-scale average characteristics. The model can be represented by the following formula:
[0029] where each is a single-scale random medium model of different scales, and the physical property parameters of the final multi-scale random medium model are equal to the sum of the physical property parameters of each single-scale random medium model (as Figure 5 indicated).
[0030] By establishing different combinations of the “stacked type” multi-scale double medium model, i.e., the fracture type reservoir model (as Figure 6 above) and the pore type reservoir model (as Figure 6 below), the effective connected pore space can be truly simulated. Through the multi-data analysis of the drilled wells in the work area in the first step, the parameters such as the radius, the pore rate and the fracture scale of the double medium reservoir are determined, so as to construct the double medium fracture-pore type seismic and geological model (as Figure 7 ) close to the actual reservoir in the work area.
[0031] Step S5, embedding the “stacked type” multi-scale double medium model into the geological model to form a superimposed model; By superimposing the established “stacked type” multi-scale double medium model in the actual geological model, a fine reservoir random model can be obtained, as Figure 8 indicated, where (a) l = 20, s = 5, = 20%, = 30°; (a) l = 50, s = 5, = 20%, = 30°; (a) l = 50, s = 5, = 90%, = 30°; (b) l = 50, s = 5, = 20%, = 120°, l represents a long axis; s represents a short axis; Indicates the porosity; Indicates the dip angle, which is set according to the actual work area.
[0032] In order to fill different scale fracture-cavity reservoir geological models, from left to right are small scale, medium scale and large scale fracture-cavity reservoir models, the background velocity is 5200 m / s, and the filling velocity in the fracture-cavity is 4800 m / s.
[0033] Step S6, non-uniform medium forward modeling is performed, prestack depth migration imaging processing is performed to obtain a migration stacking section; in order to reduce errors in the processing process and reduce the influence of errors on the migration result, prestack depth migration is directly performed using a model velocity for processing and is converted to a time domain.
[0034] Step S7, seismic attribute analysis is performed, and sensitive parameters of the fracture-cavity type reservoir are preferably selected; Based on the double medium forward modeling result (Fig. 2), Figure 9 On the basis of the seismic response characteristics of the different scale stacked fracture-cavity type reservoirs being clear, poststack reservoir sensitive attribute optimization is carried out, and the sensitivity of different attributes to the fracture-cavity type reservoir is clear. In the embodiment of the application, through the double medium model of four groups of different porosities, fracture-cavity radii, fracture-cavity filling velocities and skeleton velocities, and different shapes, three types of volume attributes and 13 layer attributes are extracted, and comparative analysis is performed, and it is obtained that the different fracture-cavity reservoir spaces and the sensitive attributes corresponding thereto are as shown in Table 2:
[0035] Table 2: Statistical table of sensitive attribute analysis of fracture-cavity reservoir ①The seismic reflection of the double medium cavity type reservoir mainly presents a chaotic medium-strong in-phase axis reflection characteristic, and the sensitive attributes are amplitude class attributes such as root mean square, amplitude variation rate, reflection intensity, absolute amplitude total amount, average absolute amplitude, average valley amplitude and average peak amplitude; ②The sensitive attributes of the stacked multi-scale double medium fracture-cavity type reservoir are dip scanning, variance volume, edge detection and ant body attribute.
[0036] Based on the double medium reservoir sensitive attributes obtained through the analysis, the next step of multi-attribute fusion attribute prediction analysis is performed.
[0037] Step S8, nonlinear learning, mining, extracting dual medium reservoir characteristics; The multiple dual medium fractured-vug type reservoir sensitive properties selected in step S7 are subjected to nonlinear learning. The number of single well samples participating in the nonlinear learning cannot be too small. The actual drilling well data is at least greater than 3. The well distribution cannot be too concentrated. Otherwise, the nonlinear mapping relationship established cannot accurately depict the lateral variation trend of the sedimentary facies.
[0038] Step S9, predicting the dual medium reservoir development zone based on the post-stack multi-attribute. Based on the dual medium reservoir sensitive properties selected in step 7, the fusion weight calculation formula is: , Among them, is the weight value of different attributes in the attribute fusion calculation; is the correlation degree of the unit attribute, and n is the number of all selected attributes to be fused. After the multi-attribute fusion, normalization processing is performed. The post-stack multi-attribute fusion map is obtained according to the weight value. The dual medium reservoir development zone is predicted based on the post-stack multi-attribute fusion map. As shown in Figure 10 The spatial distribution characteristics of the fractures in the working area are obvious. A plurality of secondary fractures with different orientations and small scales are developed. Around these secondary fractures, some fracture-vug type reservoirs with different orientations are developed. The effective reservoirs are mainly distributed in the middle and north of the working area, and the whole presents a blocky distribution in the north-west direction.
[0039] Example two The embodiment of the application also provides an electronic device, characterized in that comprising: a processor and a memory, the processor is used for executing the program of the dual medium reservoir prediction method of the hybrid vector autocorrelation function to realize the dual medium reservoir prediction method of the hybrid vector autocorrelation function.
[0040] An electronic device includes at least one processor, memory, at least one network interface, and other user interfaces. The various components in the electronic device are coupled together through a bus system. It can be understood that the bus system is used to realize the connection communication between the components. In addition to including a data bus, the bus system also includes a power bus, a control bus, and a status signal bus.
[0041] Among them, the user interface can include a display, a keyboard or a clicking device (for example, a mouse, a trackball, a touchpad or a touch screen, etc.). It can be understood that the memory in the embodiment of the application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories.
[0042] In the embodiments of the present application, the processor is configured to execute the method steps provided by the embodiments of the method by invoking the programs or instructions stored in the memory, and the programs or instructions stored in the memory can be specifically programs or instructions stored in an application.
[0043] In some embodiments, the memory stores the following elements, executable units or data structures, or a subset thereof, or an extended set thereof: an operating system and an application.
[0044] The operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., and is configured to implement various basic services and process hardware-based tasks. The application includes various application programs, such as a media player (Media Player), a browser (Browser), etc., and is configured to implement various application services. The programs for implementing the method embodiments of the present application can be included in the application.
[0045] Embodiment three: The embodiments of the present application also provide a storage medium, characterized by storing one or more programs, and the one or more programs are executable by one or more processors to implement the method for predicting a dual-media reservoir by using a hybrid vector autocorrelation function.
[0046] The method steps described in combination with the embodiments disclosed herein can be implemented by hardware, a software module executed by a processor, or a combination of both. The software module can be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the technical field.
[0047] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims
1. A hybrid vector self-correlation function method for dual media reservoir prediction, characterized in that, The method comprises the following steps: collecting relevant information of a dual-medium reservoir in an actual work area; establishing a geological model close to an actual formation structure according to the collected information; simulating a random medium according to a mixed vector autocorrelation function; establishing a "stacked" multi-scale dual-medium model; embedding the "stacked" multi-scale dual-medium model into the geological model to form a superimposed model; performing forward modeling of a non-uniform medium, and obtaining a migration stack section through pre-stack depth migration imaging processing; analyzing seismic attributes and optimizing sensitive parameters of a fracture-vug reservoir; nonlinearly learning, mining and extracting features of the dual-medium reservoir; post-stack multi-attribute prediction of a dual-medium reservoir development zone.
2. The mixed-type vector self-correlation function-based dual-media reservoir prediction method according to claim 1, wherein, The relevant information of the dual-medium reservoir in the actual work area includes seismic data, seismic, geophysical, drilling and logging data, and actual drilling conditions, and various lithologic rock physical parameters related to the carbonate fracture-vug reservoir are counted, and the velocity differences of different types of reservoirs are determined.
3. The mixed-type vector self-correlation function-based dual-media reservoir prediction method according to claim 2, wherein, According to the seismic data and the actual drilling conditions in the actual work area, a geological interpretation scheme is established, and a geological model close to the actual formation structure of the work area is established according to the rock physical parameters.
4. The mixed vector self-correlation function based reservoir prediction method of claim 1, wherein, The mixed vector autocorrelation function is: wherein is the autocorrelation function, r is the roughness factor, x is the spatial coordinate, and a is the autocorrelation function in the x direction.
5. The mixed vector self-correlation function based dual-media reservoir prediction method of claim 1, wherein, The "superimposed" multiscale dual medium model is a model in which more scale random perturbations are superimposed on large scale average characteristics, and the model can be expressed by the following formula: wherein each is a single scale random medium model of a different scale, and the physical property parameters of the final multi-scale random medium model are equal to the sum of the physical property parameters of each single scale random medium model.
6. The mixed vector self-correlation function based dual-media reservoir prediction method of claim 1, wherein, The specific steps of optimizing the sensitive parameters of the fracture-vug reservoir are as follows: based on the results of the dual-medium forward modeling, the seismic response characteristics of the fracture-vug reservoir at different scales are determined, and then post-stack reservoir sensitive attribute optimization is carried out to determine the sensitivity of different attributes to the fracture-vug reservoir.
7. The mixed vector self-correlation function based dual-media reservoir prediction method of claim 1, wherein, The nonlinear learning, mining and extracting features of the dual-medium reservoir include: The nonlinear learning is performed on the optimized sensitive attributes of the dual-medium fracture-vug reservoir, and the number of single well samples participating in the nonlinear learning cannot be too small, and the well distribution cannot be too concentrated.
8. The mixed vector self-correlation function based dual-media reservoir prediction method of claim 1, wherein, The post-stack multi-attribute prediction of the dual-medium reservoir development zone includes: based on the optimized sensitive attributes of the dual-medium reservoir, the fusion weight calculation formula is: wherein X i is the weight value of different attributes in the calculated attribute fusion; Y i is the correlation degree of the unit attribute, and n is the number of all preferred attributes that need to be fused. After the multi-attribute fusion, normalization processing is performed, and then a post-stack multi-attribute fusion map is obtained according to the weight value; The dual-medium reservoir development zone is predicted based on the post-stack multi-attribute fusion map.
9. A storage medium, characterized by The storage medium stores one or more programs, which can be executed by one or more processors to implement the dual-medium reservoir prediction method of the mixed vector autocorrelation function according to any one of claims 1-8.
10. An electronic device, comprising: The method comprises the following steps: The processor is used to execute the program of the dual-medium reservoir prediction method of the mixed vector autocorrelation function stored in the memory to implement the dual-medium reservoir prediction method of the mixed vector autocorrelation function according to any one of claims 1-8.