CN-122021225-A - Automatic history fitting method and device for oil and gas reservoir numerical simulation

CN122021225ACN 122021225 ACN122021225 ACN 122021225ACN-122021225-A

Abstract

The invention relates to the technical field of petroleum exploration and development and discloses an automatic history fitting method and device for numerical simulation of a petroleum reservoir, wherein the method comprises the following steps of S1, carrying out principal component analysis on production data of the petroleum reservoir to obtain main control production data which mainly influences the yield of the petroleum reservoir; S2, extracting geological feature parameters which mainly influence seepage dynamics of the oil and gas reservoir from the geological model of the oil and gas reservoir, reducing dimensions of the extracted geological feature parameters based on discrete cosine transform, S4, establishing the oil and gas reservoir model by using main control production data and the geological feature parameters after discrete cosine transform, taking oil and gas reservoir yield as target data, taking errors between model output data and actual production historical data as fitting targets, and using an improved artificial bee colony optimization algorithm to establish target function to adjust model parameters so as to realize automatic historical fitting. The invention can reduce the error between the simulation result and the actual data, and effectively improve the accuracy and the robustness of the automatic history fitting of the oil and gas reservoir numerical simulation.

Inventors

CHEN LIN
LI JUAN
LIU QIANYU
XIE TIANSI
WEN WAN

Assignees

中国石油天然气股份有限公司

Dates

Publication Date: 20260512
Application Date: 20241111

Claims (10)

1. An automatic history fitting method for oil and gas reservoir numerical simulation, which is characterized by comprising the following steps: S1, carrying out principal component analysis on production data of a hydrocarbon reservoir to obtain main control production data affecting the yield of the hydrocarbon reservoir; s2, extracting geological feature parameters which mainly influence seepage dynamics of the oil and gas reservoir from a geological model of the oil and gas reservoir, and reducing dimensions of the extracted geological feature parameters based on discrete cosine transform to obtain main control geological feature parameters; S3, establishing a hydrocarbon reservoir model by using main control production data and main control geological feature parameters, taking the yield of the hydrocarbon reservoir as target data, taking errors between model output data and actual production historical data as a fitting target, and using an improved artificial bee colony optimization algorithm to establish a target function to adjust model parameters so as to realize automatic history fitting of numerical simulation of the hydrocarbon reservoir.
2. The method for automatically fitting a history of hydrocarbon reservoir numerical simulation according to claim 1, wherein in step S1, the main component analysis is performed on the production data of the hydrocarbon reservoir, comprising the steps of: s11, carrying out centering treatment and standardization treatment on production data; S12, calculating a covariance matrix of the standardized data, and decomposing the eigenvalues; s13, selecting main components according to the characteristic values, and then projecting corresponding original data onto the selected main components to obtain main control production data after dimension reduction.
3. The method for automatically fitting a history of hydrocarbon reservoir numerical simulation according to claim 1, wherein in the step S3, the model parameters are adjusted by using an improved artificial bee colony optimization algorithm, comprising the steps of: s31, setting parameters of an improved artificial bee colony optimization algorithm, including a bee colony scale, a maximum iteration number, a trigger threshold, an fitness function, a step length coefficient, a disturbance amplitude, a step length attenuation rate, a disturbance attenuation rate, an initial step length parameter and a maximum iteration number; S32, setting the dimension of an objective function as D, and initializing the positions x i of SN bees of a bee colony; S33, searching a new honey source nearby the honey source for each leading bee, and generating a new solution v i ; s34, optimizing the newly generated solution v i by using a synchronous disturbance random approximation algorithm to obtain an optimized new solution S35, solving the updated solution Calculating the fitness of the bee-following search, and selecting the following bee to continue searching according to the fitness; S36, reinitializing solutions which are not improved for a long time; And S37, repeatedly executing the process to iterate until the preset maximum iteration times are reached or convergence conditions are met, and finally outputting a global optimal solution.
4. A method of simulated automatic history of hydrocarbon reservoir formation as claimed in claim 1, wherein said production data comprises daily gas production, daily water production, daily oil production, wellhead pressure, bottom hole pressure, formation pressure, wellhead temperature and bottom hole temperature.
5. A method of simulated automatic history fitting of a hydrocarbon reservoir as claimed in claim 1, wherein said geological feature parameters comprise permeability, porosity, fluid saturation, fluid viscosity, formation conductivity, initial formation pressure, rock compression factor, capillary pressure and net hair ratio.
6. The method of claim 1, wherein the master production data comprises daily gas production, daily water production, daily oil production, wellhead pressure, bottom hole pressure, and formation pressure.
7. A method of simulated automatic history of hydrocarbon reservoir as claimed in claim 1, wherein said primary control geologic parameters include permeability, porosity, fluid saturation, formation conductivity and initial formation pressure.
8. An automated history fitting apparatus for numerical modeling of a hydrocarbon reservoir, said apparatus for implementing the automated history fitting method according to any one of claims 1-7, comprising: the production data principal component analysis module is used for carrying out principal component analysis on the production data of the oil and gas reservoir to obtain main control production data which mainly influences the yield of the oil and gas reservoir; the geological feature parameter extraction module is used for extracting geological feature parameters which mainly influence the seepage dynamics of the oil and gas reservoir from a geological model of the oil and gas reservoir, and reducing the dimension of the extracted geological feature parameters based on discrete cosine transform; And the improved artificial bee colony optimization algorithm is used for taking the yield of the oil and gas reservoir as target data, taking the error between the model output data and the actual production history data as a fitting target, and adjusting model parameters to realize the automatic history fitting of the numerical simulation of the oil and gas reservoir.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable in the processor, the processor implementing the automatic history matching method according to any one of the preceding claims 1-7 when the computer program is executed.
10. A computer readable storage medium storing a computer program which, when executed in a computer processor, implements the automatic history fitting method according to any one of the preceding claims 1-7.

Description

Automatic history fitting method and device for oil and gas reservoir numerical simulation Technical Field The invention relates to the technical field of petroleum exploration and development, in particular to an automatic history fitting method and device for numerical simulation of a petroleum reservoir. Background The history fitting in the oil and gas reservoir numerical simulation is to reversely calculate the physical property parameters of the oil layer according to the known actual dynamic parameters, thereby correcting and adjusting the parameters in the model. The basic principle is that a numerical simulation method is used for establishing a model capable of describing various aspects such as oil reservoir geological structure, fluid motion, development measures and the like, and the model parameters are automatically adjusted by utilizing an optimization algorithm, so that a simulation result best accords with actual observation data. In the field of automatic oil reservoir history fitting, the objective function is generally the error between the history observation data and the model calculation predicted value, and proper objective function is selected, so that the fitting result is greatly influenced. And the second is an optimization algorithm, namely an algorithm for searching an optimal solution through an iteration method. The optimization algorithm may find the optimal solution that allows the objective function to find the maximum or minimum under the given constraints. The general practice is to convert the history fitting problem into a nonlinear programming to solve a function extremum problem defined in a multidimensional space, wherein the independent variable is required to meet constraint conditions such as a seepage equation expressed by a partial differential equation, a boundary condition, an initial condition and the like. With the increasing demand of petroleum engineering field for oil and gas reservoir numerical simulation, an automatic history fitting method becomes a key tool for better matching actual production data based on an optimization model. In the past, history fitting has relied primarily on manual data processing and modeling, requiring complex calculations and analysis by professionals such as oilfield engineers and geologist. This method has advantages in that the control of data and the interpretability of models are strong, but has disadvantages in that it is time consuming and laborious, and requires expertise and experience. Thus, this approach is prone to incorrect development schemes or technical policy formulation. With the development of computer technology, the automatic history of oil reservoirs is fitted as a new concept, and the method processes and analyzes a large amount of oil field data by utilizing computer programs and algorithms to build a mathematical model and predict the change and distribution of various substances in the oil reservoirs. The method has the advantages of high efficiency, accuracy, good repeatability and the like, does not need manual intervention, and can greatly improve the working efficiency and accuracy, thereby improving the development efficiency of the oil reservoir and reducing the development cost. However, most of the current machine learning algorithms are black box models, and it is difficult to explain the reasons of the decision process and the prediction result, which may affect the reliability and the credibility of the models in practical application. And reservoir data is often highly noisy and complex, the machine learning approach may suffer from over-fitting problems. Meanwhile, the existing automatic history fitting method has the problems that the calculation complexity is high, the local optimal solution is easy to fall into, the geological features and development characteristics of the oil and gas reservoir are difficult to consider, and the efficiency of the method in practical application is limited. Disclosure of Invention In order to solve the problems and the defects in the prior art, the invention specially provides an automatic history fitting method and device for numerical simulation of oil and gas reservoirs, the method effectively improves the adjustment effect of the model parameters, reduces the error of the simulation result and the actual data, effectively reflects the seepage dynamics of the oil and gas reservoir, and can generally effectively improve the accuracy and the robustness of the automatic history fitting of the oil and gas reservoir numerical simulation. In order to achieve the above object, the present invention has the following technical scheme: in one aspect, the invention provides an automatic history fitting method for oil and gas reservoir numerical simulation, which mainly comprises the following steps: S1, carrying out principal component analysis on production data of a hydrocarbon reservoir to obtain main control production data which mainly influences th