Reservoir physical property parameter pre-stack seismic inversion method based on resistivity logging constraint

Abstract

The invention provides a reservoir physical property parameter pre-stack seismic inversion method based on resistivity logging constraint. The method comprises the following steps: cutting observation data and depth; normalizing the seismic data, the resistivity data and the travel time to form network input; constructing a neural network framework, substituting the obtained porosity and saturation into an elastic rock physical model, and calculating to obtain an elastic parameter; the elastic parameters are substituted into a zoeppritz equation, and an underground reflection coefficient is obtained through calculation; generating a Ricker wavelet, and performing convolution calculation on the Ricker wavelet and the reflection coefficient to obtain predicted pre-stack seismic data; the porosity and the saturation are substituted into an Archie formula, and the predicted resistivity is obtained through calculation; constructing a target function; automatic differentiation is carried out, network parameters are updated, and then network output porosity and saturation are optimized. Reservoir parameter inversion driven by physical information is realized, and an intelligent scheme is provided for reservoir parameter prediction and evaluation in oil-gas exploration and development.

Description

Technical Field

[0001] The invention provides a pre-stack seismic inversion method for reservoir physical property parameters constrained by resistivity logging, belonging to the technical field of seismic exploration. Background Art

[0002] Seismic and electromagnetic methods play an important role in oil and gas exploration, reservoir identification, carbon dioxide storage monitoring and other fields (Constable, 2010). Among them, seismic exploration has always been the main means of oil and gas exploration and development. However, under complex geological conditions, the severe scattering and shielding effects of seismic waves lead to weak illumination energy of the underlying strata, and there are multiple solutions to infer underground structures using only seismic data (Colombo et al., 2013; Takam et al., 2015). In terms of reservoir description and monitoring, under the condition of limited velocity differences (such as formation water and oil), the change of oil (gas) saturation has no obvious effect on seismic wave velocity (Hu et al., 2009; Lien et al., 2013). Therefore, it is difficult to accurately predict reservoir fluid type and fluid saturation using seismic data alone. Electromagnetic exploration is considered to be an effective auxiliary means of seismic exploration (Colombo et al., 2012; Peng Guomin et al., 2018). Electromagnetic signals are sensitive to high-resistance oil and gas reservoirs, so reservoir resistivity can be linked to porosity and fluid saturation using reservoir rock physics models. The attenuation and scattering of seismic waves and electromagnetic waves are different, that is, the attenuation and scattering of electromagnetic waves are relatively strong, resulting in relatively low resolution of electromagnetic exploration. When seismic data have determined the reservoir structure framework, electromagnetic exploration can be used to predict the fluid type and saturation in the reservoir to reduce exploration risks (Chen et al., 2007; Chen et al., 2012). Therefore, seismic and electromagnetic data can complement each other, overcome the limitations of a single method, and effectively improve the imaging resolution of underground geological structures and the accuracy of reservoir description, so as to better detect underground resources and monitor oil and gas.

[0003] The combination of seismic and electromagnetic data is mainly based on two strategies: structural similarity and the relationship between rock physical parameters. The first inversion method based on structural similarity is achieved by imposing structural similarity constraints on electrical and elastic parameters in the target area. Gallardo and Meju (2003) proposed a joint inversion method of resistivity and seismic travel time based on the cross-gradient method. Wiik et al. (2015) proposed a structural smoothing regularization scheme that introduces seismic stratigraphic attributes into the inversion of marine controlled source electromagnetic method, and confirmed the improvement of inversion results by the structural information contained in seismic data in synthetic data and field measured data. Guo et al. (2017) used the structural tensor extracted from the seismic image to correct the regularized inversion parameters, so that the inversion model allows discontinuous changes in the parameters on both sides of the geological structure during the smoothing process, thereby improving the inversion accuracy. Li et al. (2019) introduced a joint total variation constraint of seismic-electromagnetic data in the inversion objective function and proposed an alternating inversion method for marine controlled source electromagnetic and seismic data. Compared with electromagnetic or seismic inversion alone, the joint inversion can significantly improve the inversion results. Kong Fanxiang et al. (2024) introduced a cross-gradient function to realize the mutual coupling of seismic electromagnetic physical parameter results, and significantly improved the characterization of anomaly morphology, structural structure and physical property numerical recovery. The above application cases prove that joint inversion based on local correlation constraints can effectively improve the convergence and multi-solution problems of single geophysical inversion. However, this method will produce artificially introduced boundaries in some cases, where the boundary of one physical model depends on the boundary of another physical model. However, EM has poor resolution for deep anomalies, and the inaccurate resistivity information conveyed by structural constraints usually misleads the velocity model and introduces errors into the seismic inversion structure.

[0004] Another joint inversion method is based on the joint inversion of rock physical properties, which is achieved through the relationship between reservoir parameters, electrical parameters and elastic parameters. The rock physics model constructs the relationship between reservoir electrical properties, elastic parameters and reservoir physical parameters, and is the theoretical basis for conducting seismic forward inversion and electromagnetic forward inversion studies of reservoir parameters. The empirical parameters in the rock physics model (such as critical porosity, cementation coefficient, saturation coefficient and tortuosity coefficient in Archie's formula) play an important role in the mapping relationship between elastic and electrical parameters and physical parameters. In classical rock physics models, these empirical parameters are usually set to fixed values, which makes it difficult to accurately predict the elastic and electrical parameters and other responses of rocks with real complex pore structures.

[0005] Gao et al. (2012) proposed a Gauss-Newton inversion algorithm for electromagnetic and full-wavefield seismic data to estimate reservoir rock physical parameters, such as porosity and fluid saturation, overcoming the strong non-uniqueness caused by the inversion of single geophysical data. Ren et al. (2017) proposed a joint inversion method for marine controlled source-seismic data using a multi-chain Markov chain Monte Carlo sampler, which improved the accuracy of reservoir saturation evaluation.

[0006] The above numerical analysis based on traditional methods and practical application cases have proved the feasibility of the seismic-electromagnetic joint inversion method and its effectiveness in reducing the multi-solution of a single geophysical inversion result and improving the stability of the solution. However, traditional numerical methods have many limitations, such as easy to fall into local minimum, excessive dependence on initial parameters, model simplification and assumption restrictions, and limited ability to handle nonlinear relationships. In the inversion process, the forward operator of complex porous media reservoirs is often highly nonlinear, so the inversion problem is difficult to solve.

[0007] With the development of artificial intelligence and machine learning methods, deep learning technology has been widely used in the field of geophysics. Deep learning focuses on building the overall framework of the algorithm, rather than accurately designing the algorithm parameters, extracting the hidden internal connections between different data that are not perceived by humans, and "seeing" the effective features hidden in the data, which is difficult to do with traditional numerical methods. Therefore, making full use of deep learning to explore the correlation between different geophysical data to improve inversion and interpretation capabilities has become a research hotspot in the current joint inversion field.

[0008] Guo et al. (2020) first used deep learning constraints to realize the conversion and joint inversion of audio magnetotelluric and seismic travel time data, retaining the structural similarity between different inversion models. This method realizes the nonlinear transformation between resistivity and wave velocity through deep learning. Sun et al. (2020) proposed a supervised deep learning method for predicting salt geometry using seismic and electromagnetic data, and tested the fusion performance of different network stages, which provided good inspiration for the fusion of network architecture to achieve deep learning joint inversion. The above data-driven joint inversion method directly establishes a mapping relationship between input and output, but it has a large demand for geophysical data and does not clearly follow the physical rules in the inversion problem.

[0009] On this basis, embedding physical laws in deep learning networks has become a key challenge to solve the generalization of joint inversion problems. Under the physical-driven strategy, the algorithm incorporates the forward process based on physical laws (such as Maxwell's equations, wave equations, etc.), and does not take the prediction model as the final result in an end-to-end manner, but assists in gradient-based model inversion, such as constraining the direction of gradient descent. Hu et al. (2023) developed a deep learning-enhanced joint inversion framework that combines deep learning with traditional independent inversion work to achieve joint inversion of two-dimensional DC resistivity data and seismic propagation time. Ren qt al. (2024) borrowed Hu's technical theory and conducted a joint inversion study of surface seismic full waveform and high-density electrical method based on deep learning. Liu et al., (2023) proposed a neural network model based on physical information, used Bayesian networks to realize the joint inversion of three-dimensional seismic and electromagnetic, and accurately quantified the undetermined global parameters in the rock physics model, and achieved accurate estimation of reservoir properties in the Sleipner CO2 storage field storage monitoring scenario. Zhou et al. (2023) proposed a seismic image-guided MT inversion strategy based on deep learning to achieve multi-physics coupled inversion embedded with geophysical knowledge. Subsequently, Zhou et al. (2024) embedded the seismic image texture operator extracted by the convolutional neural network VGG19 into the magnetotelluric inversion objective function based on the principle of structural similarity, achieving high-resolution magnetotelluric inversion constrained by seismic images.

[0010] In summary, the deep learning-based seismic-electromagnetic joint inversion method fully utilizes the rich information contained in multi-source data to achieve high-precision inversion imaging of underground reservoirs, which has become a current research hotspot. However, most of the existing studies that introduce deep learning technology for joint inversion are mainly based on the principle of structural similarity, starting from the image, using convolutional neural networks for feature extraction, and then combining with traditional methods. There are few studies on joint inversion based on rock physical properties. In addition, the above studies mostly rely on large amounts of data-driven, or data-physical knowledge-driven, etc. However, in the current situation where the amount of electromagnetic data is small, the data-driven method is difficult to meet the requirements of large amounts of data. Summary of the invention

[0011] Based on the above problems, the present invention obtains porosity and saturation by jointly inverting the acquired one-dimensional pre-stack seismic records and resistivity, establishes a non-mapping relationship between seismic-resistivity data and reservoir physical property parameters, improves the inversion accuracy of underground reservoir physical property parameters, and reduces the inversion multi-solution.

[0012] In view of the above-mentioned problems in the prior art, the present invention provides a resistivity-constrained pre-stack seismic parameter inversion method, which can avoid the traditional method's reliance on the initial model, realize the inversion of reservoir parameters (porosity, saturation) driven by physical information, and provide an intelligent solution for reservoir parameter prediction and evaluation in oil and gas exploration and development.

[0013] The specific technical solution is:

[0014] A pre-stack seismic inversion method for reservoir physical property parameters constrained by resistivity logging comprises the following steps:

[0015] S1, integrated pre-stack seismic records seis And resistivity logging data d res , and the corresponding two-way travel time range t of the seismic record, the observation data and depth are clipped to ensure their dimensions are equal, and the resistivity data are converted to the time domain;

[0016] S2. Perform maximum-minimum normalization on seismic data, resistivity data and time to form network input d input ;

[0017] The maximum-minimum normalization in S2 is as shown in formula (1), and the three are combined to form the network input d input , as shown in formula (2);

[0018]

[0019]

[0020] Among them, d seis ,d res , t are the original seismic data, resistivity logging data and travel time respectively; X.min(), X.max() represent the minimum and maximum values ​​of X respectively; t * They are normalized seismic data, resistivity logging data and travel time respectively.

[0021] S3, build a neural network framework, the network input is d input , the output is m = [Φ, S];

[0022] The network in S3 is a five-layer fully connected network with tanh as the activation function. Normalization layers are added to the hidden layer and the output layer. A sigmoid activation function is added after the output layer to limit the network output range to [0, 1]. This operation adds the prior information of the inverted target attribute value range, that is, the porosity and saturation values ​​are both between [0, 1].

[0023]

[0024] In formula (3) represents the constructed neural network inversion operator; m represents the reservoir physical property parameters, where Φ and S represent porosity and water saturation respectively; θ p Represents the neural network parameters.

[0025] S4, the porosity Φ and saturation S obtained by network output are substituted into the elastic rock physics model to calculate the elastic parameters; the specific formula in S4 is as follows:

[0026]

[0027]

[0028]

[0029]

[0030]

[0031]

[0032]

[0033] μ sat =μ dry (11)

[0034]

[0035]

[0036]

[0037] ρ sat =(1-φ)ρ mat +φρ fl (15)

[0038] ρ mat =(1-v cl )ρ qtz +v cl ρ cl (16)

[0039] ρ cl =(1-S g )ρ w +S g ρ g (17)

[0040] Where P e is the effective pressure, v mat is the Poisson's ratio of the grain; n cis the coordination number, which is 7 in this scheme; φ0 is the critical porosity; φ is the saturated rock porosity; K HM With μ HM are the equivalent bulk modulus and shear modulus respectively; K dry With μ dry are the bulk modulus and shear modulus of dry rock, respectively; K mat With μ mat are the bulk modulus and shear modulus of the rock skeleton respectively; K sat With μ sat are the bulk modulus and shear modulus of fluid-saturated rock, respectively; k fl is the bulk modulus of the fluid phase, which is determined by the pore water saturation S w and supercritical gas saturation S g Calculated; V p 、V s Represent the longitudinal wave velocity and the transverse wave velocity respectively; ρ sat , fl , mat , w , g , cl , qtz They are saturated rock density, fluid density, rock skeleton density, pore water density, supercritical gas density, clay density and quartz density; v cl is the clay volume fraction.

[0041] S5, substituting the elastic parameters calculated in step S4 into the Zoeppritz equation to calculate the underground reflection coefficient r pp ;

[0042] The equations in S5 are as shown in equations (18)-(21), and the underground reflection coefficient r is calculated pp ;

[0043] r pp (t,θ)=R+Gsin 2 θ+F(tan 2 θ-sin 2 θ) (18)

[0044]

[0045]

[0046]

[0047] Where t represents the travel time, θ represents the incident angle of the seismic wave; ΔV P , ΔV s , Δρ represent the P-wave velocity, S-wave velocity and density difference at adjacent travel time positions respectively; V p 、Vs , They represent the average values ​​of longitudinal wave velocity, shear wave velocity and density at adjacent travel time positions respectively.

[0048] S6, generate Ricker wavelet, and perform convolution calculation on the reflection coefficient calculated in step S5 to obtain predicted pre-stack seismic data

[0049] Predicting pre-stack seismic data in S6

[0050]

[0051] In formula (22), w(θ) represents the seismic wavelet.

[0052] S7. Substitute the porosity Φ and saturation S obtained from the network output into Archie's formula to calculate the predicted resistivity

[0053] Predicting resistivity in S7 As shown in formula (23):

[0054]

[0055] Among them, R w is the pore water resistivity; n is the saturation index; m is the cementation coefficient.

[0056] S8, constructing data items in the objective function using real earthquake and resistivity data and the predicted earthquake and resistivity data respectively calculated in step S7 and step S8; at the same time, using the chain rule of the neural network, performing gradient calculation on the porosity φ and saturation S obtained by the network output, thereby constructing a regularization item in the objective function;

[0057] The objective function is constructed in S8 as shown in formula (24);

[0058]

[0059] Among them, the first item is the seismic data fitting item, the second item is the resistivity data fitting item, and the third item is the model fitting item; α1, α2, and α3 are the three weights respectively; σ res and σ seis Represent the standard deviation of real resistivity data and seismic data respectively; m represents the inversion parameter model, Represents the partial derivative of the model.

[0060] S9. Automatic differentiation is performed according to the constructed joint inversion loss function to update the network parameters, thereby optimizing the network output porosity and saturation to fit the real observed data; when the Loss drops to the set minimum threshold, or the number of training times reaches the maximum number, the result is output.

[0061] The solution of the present invention aims at the challenging problems in the exploration of complex oil and gas reservoirs, combines multidisciplinary cross-disciplinary technologies, combines physical models with geophysical observation data, realizes reservoir inversion and fluid identification, and exhibits the following unique advantages and beneficial effects:

[0062] 1) Inversion results and reliability are greatly improved

[0063] In order to solve the problem of low reliability and strong multi-solution inversion of single geophysical data, this patented solution innovatively introduces physical constraint information such as rock physics models and seismic forward operators, combines seismic and resistivity multi-source observation data, and forms a physically constrained neural network joint inversion framework. This method effectively reduces the uncertainty in inversion and realizes high-precision imaging and fluid identification of oil and gas reservoirs.

[0064] 2) Overcoming the difficulty of dealing with nonlinear mapping relationships

[0065] Traditional prestack seismic reservoir inversion methods have limitations when dealing with complex nonlinear mapping relationships and are prone to falling into local extremes, which affects the accuracy of the inversion results. The physical constraint resistivity-prestack seismic joint inversion method based on deep learning proposed in this patent scheme mines the coupling relationship between multi-source data through deep learning technology. It is not only suitable for highly nonlinear inversion problems, but also greatly improves the accuracy and fully demonstrates the joint advantages of resistivity data and prestack seismic data.

[0066] 3) No reliance on initial model

[0067] Compared with traditional methods, this method does not require an initial model, thus getting rid of the traditional method's dependence on the initial model. Under this condition, the stability and consistency of reservoir parameters (such as porosity and saturation) obtained by neural network inversion have been fully verified. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 A flowchart for constructing a neural network framework for the present invention;

[0069] Figure 2 The inversion results of the single method and the combined method for the one-dimensional reservoir parameters of the embodiment;

[0070] Figure 3 1 is the one-dimensional reservoir parameter inversion result of the embodiment. DETAILED DESCRIPTION

[0071] The specific technical solution of the present invention is described in conjunction with the accompanying drawings.

[0072] A pre-stack seismic inversion method for reservoir physical property parameters constrained by resistivity logging, the specific steps are as follows:

[0073] S1, integrated pre-stack seismic records seis And resistivity logging data d res , and the corresponding two-way travel time range t of the seismic record, the observation data and depth are clipped (interpolated and expanded) to ensure that their dimensions are equal, and the resistivity data are converted to the time domain;

[0074] S2. Perform maximum-minimum normalization on seismic data, resistivity data and time (as shown in formula (1)), and combine the three to form the network input d input (As shown in formula (2)).

[0075]

[0076]

[0077] S3, build a neural network framework, such as Figure 1 , the network input is d input , the output is m = [φ, S]. The network is a five-layer fully connected network with tanh as the activation function. Normalization layers are added to the hidden layer and the output layer. A sigmoid activation function is added after the output layer to limit the network output range to [0, 1]. This operation adds the prior information of the inverted target attribute value range, that is, the porosity and saturation values ​​are both between [0, 1].

[0078]

[0079] S4. Substitute the porosity φ and saturation S obtained by the network output into the elastic rock physics model (here, the unconsolidated sandstone Gassmann model is taken as an example) to calculate the elastic parameters (P-wave velocity, S-wave velocity and density). The specific formula is as follows:

[0080]

[0081]

[0082]

[0083]

[0084]

[0085]

[0086]

[0087] μ sat =μ dry (11)

[0088]

[0089]

[0090]

[0091] ρ sat =(1-φ)ρ mat +φρ fl (15)

[0092] ρ mat =(1-v cl )ρ qtz +v cl ρ cl (16)

[0093] ρ cl =(1-S g )ρ w +S g ρ g (17)

[0094] S5. Substitute the elastic parameters calculated in step S4 into the Zoeppritz equation (such as equations (18)-(21)) to calculate the underground reflection coefficient r pp .

[0095] r pp (t, θ) = R + Gsin 2 θ+F(tan 2 θ-sin 2 θ) (18)

[0096]

[0097]

[0098] S6, generate Ricker wavelet, and perform convolution calculation on the reflection coefficient calculated in step S5 to obtain predicted pre-stack seismic data

[0099]

[0100] S7. Substitute the porosity Φ and saturation S obtained from the network output into Archie's formula to calculate the predicted resistivity (As shown in formula (21)). Where Rw is the pore water resistivity, which is 0.17Ω·m; m is the cementation coefficient, which is 1.3; n is the saturation index, which is 2.0.

[0101]

[0102] S8, construct the objective function (such as formula (22)) through the real earthquake, resistivity data and the predicted earthquake, resistivity data calculated in step S7 and step S8 respectively. Among them, the first term is the earthquake data fitting term, the second term is the resistivity data fitting term, and the third term is the model fitting term; α1, α2, α3 are three weights respectively; σ res and σ seis Represent the standard deviation of the real resistivity data and seismic data respectively.

[0103]

[0104] S9. Automatic differentiation is performed according to the constructed joint inversion loss function to update the network parameters, thereby optimizing the network output porosity and saturation to fit the real observation data. When the loss drops to the set minimum threshold or the number of training times reaches the maximum number, the result is output.

[0105] Figure 2 The inversion results of the single method and the combined method for the one-dimensional reservoir parameters of the embodiment; Figure 2 (a) Porosity; (b) Saturation. The black value is the true value; the blue value is the seismic inversion result; the green value is the resistivity inversion result; and the red value is the resistivity-seismic joint inversion result.

[0106] By comparing with the real data (black curve), it can be seen that the joint inversion result is far superior to the single seismic inversion (blue curve) and the single resistivity inversion result (green curve), and can better reflect the distribution and change of underground reservoir parameters. Therefore, this embodiment can fully prove that the resistivity-constrained prestack seismic reservoir parameter inversion effect proposed in this scheme has been greatly improved compared with the single method.

[0107] Figure 3 is the one-dimensional reservoir parameter inversion result of the embodiment; Figure 3 (a) Porosity; (b) Saturation. The black is the true value; the gray is the joint inversion result of different random seeds (a total of 10 different random seeds); the red is the average value of different random seeds.

[0108] The following conclusions can be drawn from the inversion results: The resistivity logging constrained prestack seismic reservoir inversion method proposed in this study can achieve high-precision inversion of underground reservoir parameters. In addition, the method shows good robustness, that is, the network framework constructed using different random seeds can accurately invert reservoir parameters close to the true value.

Claims

1. A prestack seismic inversion method for reservoir physical property parameters constrained by resistivity logging, characterized in that: The following steps are involved: S1, integrated pre-stack seismic records seis And resistivity logging data d res , and the corresponding two-way travel time t of the seismic record, the observation data and depth are clipped to ensure their dimensions are equal, and the resistivity data are converted to the time domain; S2. Perform maximum-minimum normalization on seismic data, resistivity data and time to form network input d input ; S3, build a neural network framework, the network input is d input , the output is m = [Φ, S]; S4, substituting the porosity Φ and saturation S obtained by the network output into the elastic rock physics model to calculate the elastic parameters; S5, substituting the elastic parameters calculated in step S4 into the Zoeppritz equation to calculate the underground reflection coefficient r pp ; S6: Generate Ricker wavelet and perform convolution calculation with the reflection coefficient calculated in step S5 to obtain predicted pre-stack seismic data. S7. Substitute the porosity Φ and saturation S obtained from the network output into Archie's formula to calculate the predicted resistivity S8, constructing data items in the objective function using real earthquake and resistivity data and the predicted earthquake and resistivity data respectively calculated in step S7 and step S8; at the same time, using the chain rule of the neural network, performing gradient calculation on the porosity Φ and saturation S obtained by the network output, thereby constructing a regularization item in the objective function; S9. Automatic differentiation is performed according to the constructed joint inversion loss function to update the network parameters, thereby optimizing the network output porosity and saturation to fit the real observed data and ensure the smoothness of the model; when the Loss drops to the set minimum threshold, or the number of training times reaches the maximum number, the result is output.

2. The method for prestack seismic inversion of reservoir physical property parameters constrained by resistivity logging according to claim 1, characterized in that: The maximum-minimum normalization in S2 is as shown in formula (1), and the three are combined to form the network input d input , as shown in formula (2); Among them, d seis d res , t are the original seismic data, resistivity logging data and travel time respectively; X.min(), X.max() represent the minimum and maximum values ​​of x respectively; t * They are normalized seismic data, resistivity logging data and travel time respectively.

3. The method for prestack seismic inversion of reservoir physical property parameters constrained by resistivity logging according to claim 2, characterized in that: The network in S3 is a five-layer fully connected network with tanh as the activation function. Normalization layers are added to the hidden layer and the output layer. A sigmoid activation function is added after the output layer to limit the network output range to [0, 1]. This operation adds the prior information of the inverted target attribute value range, that is, the porosity and saturation values ​​are both between [0, 1]. In formula (3) represents the constructed neural network inversion operator; m represents the reservoir physical property parameters, where Φ and S represent porosity and water saturation respectively; θ p Represents the neural network parameters.

4. The method for prestack seismic inversion of reservoir physical property parameters constrained by resistivity logging according to claim 3, characterized in that: The specific formula in S4 is as follows: m sat =μ dry (11) r sat =(1-φ)ρ mat +fr fl (15) r mat =(1-v cl )r qtz +v cl r cl (16) r cl =(1-S g )r w +S g r g (17) Where P e is the effective pressure, v mat is the Poisson's ratio of the grain; n c is the coordination number, which is 7 in this scheme; φ0 is the critical porosity; φ is the saturated rock porosity; K HM With μ HM are the equivalent bulk modulus and shear modulus respectively; K dry With μ dry are the bulk modulus and shear modulus of dry rock, respectively; K mat With μ mat are the bulk modulus and shear modulus of the rock skeleton respectively; K sat With μ sat are the bulk modulus and shear modulus of fluid-saturated rock, respectively; K fl is the bulk modulus of the fluid phase, which is determined by the pore water saturation S w and supercritical gas saturation S g Calculated; V p 、V s Represent the longitudinal wave velocity and the transverse wave velocity respectively; ρ sat , fl , mat , w , g , cl , qtz They are saturated rock density, fluid density, rock skeleton density, pore water density, supercritical gas density, clay density and quartz density; v cl is the clay volume fraction.

5. The method for prestack seismic inversion of reservoir physical property parameters constrained by resistivity logging according to claim 4, characterized in that: The equations in S5 are as shown in equations (18)-(21), and the underground reflection coefficient r is calculated pp ; r pp (t,θ)=R+Gsin 2 θ+F(tan 2 θ-sin 2 i) (18) Where t represents the travel time, θ represents the incident angle of the seismic wave; ΔV P , ΔV s , Δρ represent the P-wave velocity, S-wave velocity and density difference at adjacent travel-time positions respectively; They represent the average values ​​of longitudinal wave velocity, shear wave velocity and density at adjacent travel time positions respectively.

6. The method for prestack seismic inversion of reservoir physical property parameters constrained by resistivity logging according to claim 5, characterized in that: Predicting pre-stack seismic data in S6 In formula (22), w(θ) represents the seismic wavelet.

7. The method for prestack seismic inversion of reservoir physical property parameters constrained by resistivity logging according to claim 1, characterized in that: Predicted resistivity in S7 As shown in formula (23): Among them, R w is the pore water resistivity; n is the saturation index; m is the cementation coefficient.

8. The method for prestack seismic inversion of reservoir physical property parameters constrained by resistivity logging according to claim 1, characterized in that: The objective function is constructed in S8 as shown in formula (24); Among them, the first item is the seismic data fitting item, the second item is the resistivity data fitting item, and the third item is the model fitting item; α1, α2, and a3 are the three weights respectively; σ res and σ seis Represent the standard deviation of real resistivity data and seismic data respectively; m represents the inversion parameter model, Represents the partial derivative of the model.

Images