Intelligent oil reservoir simulation and safety prediction system

Abstract

The invention provides an intelligent oil reservoir simulation and safety prediction system. The system comprises a parameter setting module, a numerical simulation software generation module, an oil reservoir simulation and safety prediction module, an information interaction module, an application management module and a program modularization module, the parameter setting module is used for acquiring numerical simulation parameters set by a user from a user side and importing the numerical simulation parameters into the numerical simulation software generation module; configuring numerical simulation parameters in a human-computer interface in the parameter setting module, wherein the numerical simulation parameters comprise well, reservoir, fluid and well network parameters; the well network parameters comprise a well network type, a well group size, a well name and a well number, grid mapping of a well, a position of the well, a control radius of the well, a grid influence range of the well and yield data of the well; and the numerical simulation software generation module combines the obtained numerical simulation parameters with a preset parameterized model and an intelligent optimization algorithm through a neural network model after receiving the numerical simulation parameters.

Description

Technical Field

[0001] The invention belongs to the field of safety prediction, and in particular relates to an intelligent reservoir simulation and safety prediction system. Background Art

[0002] The intelligent reservoir simulation and safety prediction system is a comprehensive technology that integrates advanced information technology and geological reservoir engineering knowledge, aiming to improve the efficiency and safety of oilfield development. The system achieves accurate prediction and management of reservoir dynamics through high-precision numerical simulation, real-time monitoring and data analysis. However, although the existing technology has made significant progress in reservoir simulation and safety prediction, there are still many shortcomings.

[0003] Current reservoir simulation technology mainly relies on numerical simulation methods, which use reservoir flow simulators to predict reservoir and single well production and pressure dynamics by establishing detailed reservoir geological models, porosity and permeability models, and fluid property parameters. Although numerical simulation methods have been widely used in remaining oil research and reservoir dynamic prediction, their results are highly dependent on reliable reservoir and fluid property parameters and reservoir simulators suitable for reservoir conditions. In actual operation, obtaining these parameters is often challenging, and unreasonable sampling and testing methods may lead to inaccurate model prediction results. In addition, the accuracy and applicability of numerical simulation methods are also limited when dealing with heterogeneous reservoirs and complex geological structures.

[0004] In terms of security prediction, although the construction of smart oilfields has made significant progress in data collection, transmission and analysis, data security and privacy protection issues are becoming increasingly prominent. In the process of data collection, transmission and storage, smart oilfields need to face multiple challenges such as data leakage, malicious tampering and personal privacy information protection. At the same time, the differences in standards and protocols in different technical fields make it difficult for equipment and systems to be interconnected, and data transmission efficiency is low, affecting the operating efficiency and performance of smart oilfield systems. In addition, the construction of smart oilfields requires a lot of capital and resource investment, and how to balance construction costs and expected benefits has become a major problem. The shortage of technical talents and the rapidity of technological updates also pose challenges to the construction of smart oilfields.

[0005] Although the intelligent reservoir simulation and safety prediction system has broad application prospects, the existing technology still has many shortcomings. In order to improve the accuracy of reservoir simulation and safety prediction, it is necessary to continuously develop new technical means, optimize numerical simulation methods, strengthen data security and privacy protection, and promote technical standardization and interconnection to achieve the sustainable development of intelligent oil fields. Summary of the invention

[0006] The present invention proposes an intelligent reservoir simulation and safety prediction system, which solves the technical problems of cumbersome parameter settings, inflexible numerical simulation software generation, low reservoir simulation and safety prediction accuracy, unintuitive information interaction, and inconvenient application management in traditional reservoir simulation. By integrating multiple modules, intelligent management of the entire chain from parameter setting to result display is realized.

[0007] The technical solution of the present invention is implemented as follows: the intelligent reservoir simulation and safety prediction system includes a parameter setting module, a numerical simulation software generation module, a reservoir simulation and safety prediction module, an information interaction module, an application management module and a program modularization module for data interaction;

[0008] The parameter setting module is used to obtain the numerical simulation parameters set by the user from the user end, and import the numerical simulation parameters into the numerical simulation software generation module; the numerical simulation parameters are configured in the human-machine interface of the parameter setting module, and the numerical simulation parameters include: well, reservoir, fluid and well network parameters; the well network parameters include well network type, well group size, well name and well number, grid mapping of the well, well location, well control radius, grid influence range of the well and well production data;

[0009] After receiving the numerical simulation parameters, the numerical simulation software generation module combines the acquired numerical simulation parameters with the preset parameterized model and the intelligent optimization algorithm through the neural network model to generate the associated numerical simulation software, and stores the associated numerical simulation software in the storage location set by the user;

[0010] The reservoir simulation and safety prediction module calls the associated numerical simulation software, and then simulates the reservoir parameters by importing the associated parameters and reservoir parameters in the numerical simulation software, and performs dynamic tracking monitoring and early warning; wherein the reservoir parameters are generated by combining the numerical simulation parameters set by the user with the reservoir parameters generated based on the machine learning algorithm to simulate the parameters;

[0011] The information interaction module calls the dynamic data and early warning data in the reservoir simulation and safety prediction module, and visualizes the data simulation results to the user;

[0012] The application management module is used to schedule the associated parameters and reservoir parameters imported from the reservoir simulation and safety prediction module, and simultaneously store and record the dynamic tracking monitoring and early warning information;

[0013] The program modularization module is used to integrate the parameter setting module, the numerical simulation software generation module, the reservoir simulation and safety prediction module, the information interaction module and the application management module and package them into executable program software.

[0014] In terms of parameter setting, the existing technology usually relies on manual input, which is not only inefficient but also prone to errors. However, this system, through the parameter setting module, directly obtains and imports numerical simulation parameters from the user end, greatly improving the accuracy and efficiency of parameter setting. In terms of numerical simulation software generation, the existing technology often uses fixed numerical simulation software, which is difficult to adapt to changes in different reservoir conditions and user needs. However, this system, through the numerical simulation software generation module, combines neural network models, parameterized models and intelligent optimization algorithms, can dynamically generate related numerical simulation software and store it in the user-specified location, thereby improving the flexibility and adaptability of numerical simulation software. In terms of reservoir simulation and safety prediction, the existing technology can usually only perform static simulation and lack dynamic tracking, monitoring and early warning functions. However, this system, through the reservoir simulation and safety prediction module, realizes dynamic simulation and tracking monitoring of reservoir parameters, and can issue timely warnings based on early warning data, thereby improving the safety and reliability of reservoir management. In terms of information interaction, the existing technology often uses traditional data display methods, such as tables, charts, etc., which are difficult to intuitively display reservoir simulation results. Through the information interaction module, this system realizes the display of dynamic data and warning data to users in a visual manner, thereby improving the intuitiveness and convenience of information interaction.

[0015] In terms of application management, existing technologies usually lack a unified management platform, making it difficult to achieve parameter scheduling and information backup and recording. However, this system uses the application management module to achieve the scheduling of associated parameters and reservoir parameters, and simultaneously backup and record dynamic tracking monitoring and early warning information, thereby improving the convenience and traceability of application management.

[0016] As a preferred embodiment, the numerical simulation software generation module is also provided with a data analysis and inspection module, a data cleaning module, a feature engineering module and a feature selection module, wherein the data analysis and inspection module is used to check whether the input numerical simulation parameters meet the parameter requirements of the preset neural network model; the data cleaning module is used to delete redundant information in the input numerical simulation parameters and divide the input numerical simulation parameters into a training set and a verification set; the feature engineering module is used to perform feature engineering processing on the training set data; the feature engineering processing includes data binning, data encoding, data normalization, data splitting and data crossover; the feature selection module is used to perform feature selection on the training set and verification set after data splitting based on machine learning algorithm and feature correlation algorithm to obtain optimal numerical simulation parameters.

[0017] As a preferred implementation, the feature selection module performs feature selection on the training set validation set by first analyzing the data type and constructing a feature dictionary based on the variable characteristics and the classification data, dividing the continuous variables into numerical and categorical types; then, based on the linear regression algorithm, continuous variables are screened to obtain the sensitivity factors of the continuous variables, and based on the feature dictionary, a multivariate regression model is established for each continuous variable, and the sensitive factors are screened based on the R-square indicator; finally, based on the random forest algorithm, categorical variables are screened to obtain the sensitive factors of the categorical variables, a decision tree is established and its corresponding decision weights are assigned.

[0018] As a preferred embodiment, the reservoir simulation and safety prediction module also includes a dynamic analysis module and a dynamic monitoring module. The dynamic analysis module is used to dynamically track the reservoir, and regularly update the associated numerical simulation software parameters, and import the dynamic tracking data into the application management module for backup; the dynamic monitoring module is used to monitor the reservoir and obtain reservoir monitoring data; the associated numerical simulation software, reservoir monitoring data and machine learning algorithm are coupled to obtain real-time dynamic reservoir safety warnings.

[0019] As a preferred implementation, the dynamic analysis module decomposes the calculation process of the numerical simulation parameters set by the user to obtain decomposition parameters, performs calculations based on the decomposition parameters, and the reservoir simulation and safety prediction module transmits the calculation results of the decomposition parameters to the parameter setting module and the information interaction module for the user to view.

[0020] As a preferred implementation, the dynamic monitoring module screens the numerical simulation parameters set by the user to obtain the screened preferred numerical simulation parameters, performs data preprocessing on the preferred numerical simulation parameters to obtain preprocessed data; wherein the data preprocessing includes data cleaning, data reduction, feature extraction, feature binning and feature selection; performs intelligent decision-making operation on the preprocessed data according to a preset machine learning algorithm to obtain the intelligent decision-making operation result, monitors the reservoir according to the intelligent decision-making operation result, and obtains reservoir monitoring data.

[0021] After adopting the above technical scheme, the beneficial effects of the present invention are as follows: through the intelligent processing of the parameter setting module, the possibility of manual intervention and erroneous input is greatly reduced, and the accuracy and efficiency of parameter setting are improved. Through the dynamic generation function of the numerical simulation software generation module, suitable numerical simulation software can be flexibly generated according to different reservoir conditions and user needs, thereby improving the accuracy and adaptability of numerical simulation. This can not only reduce the cost of numerical simulation, but also improve the efficiency and accuracy of numerical simulation. Through the dynamic tracking monitoring and early warning functions, potential safety hazards can be discovered and solved in time, thereby avoiding accidents. This can not only ensure the safety of reservoir management, but also improve the exploitation efficiency and economic benefits of reservoirs. Through the visual data display method, users can intuitively understand the results and early warning information of reservoir simulation, so as to make decisions and management more conveniently. This can not only improve the efficiency and accuracy of information interaction, but also reduce information misunderstanding and communication costs. Through a unified management platform and backup record function, the scheduling of associated parameters and reservoir parameters and the backup record of information can be realized. This can not only improve the convenience and traceability of application management, but also provide strong data support for subsequent reservoir management. Compared with existing technologies, this intelligent reservoir simulation and safety prediction system has significant beneficial effects, which can not only improve the efficiency and accuracy of reservoir management, but also ensure the safety and reliability of reservoir management. This is of great significance for promoting the development of the oil industry and improving the utilization efficiency of oil resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0023] Figure 1 It is a flow chart of the system of the present invention. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0025] Example:

[0026] like Figure 1As shown in the figure, the intelligent reservoir simulation and safety prediction system is a comprehensive reservoir management tool that integrates advanced information technology, artificial intelligence algorithms and numerical simulation technology. Through data interaction between multiple modules, the system realizes a complete workflow from parameter setting to result display, providing an efficient, accurate and reliable solution for reservoir management. The working principle of the system will be explained in detail below, and its workflow will be demonstrated through a specific work scenario.

[0027] The working principle of the intelligent reservoir simulation and safety prediction system is mainly based on the following core modules and their data interaction:

[0028] Parameter setting module: This module is the entrance to the system and is responsible for obtaining the numerical simulation parameters set by the user from the user end. These parameters include detailed information on wells, reservoirs, fluids, and well networks, such as well type, location, control radius, production data, etc. Users configure these parameters through the human-computer interface to ensure the accuracy and pertinence of the simulation.

[0029] Numerical simulation software generation module: After receiving the numerical simulation parameters transmitted by the parameter setting module, this module uses the neural network model to combine these parameters with the preset parameterized model and intelligent optimization algorithm to dynamically generate the associated numerical simulation software. This process makes full use of the self-learning and optimization capabilities of artificial intelligence, so that the generated numerical simulation software can better adapt to different reservoir conditions and user needs. The generated software is then stored in the location specified by the user for subsequent calls and simulations.

[0030] Reservoir simulation and safety prediction module: This module is the core part of the system and is responsible for calling the associated numerical simulation software to simulate reservoir parameters. During the simulation process, the module will import the associated parameters in the numerical simulation software and the reservoir parameters generated by the machine learning algorithm to perform dynamic simulation of the reservoir. At the same time, the module also has dynamic tracking, monitoring and early warning functions, which can monitor the status of the reservoir in real time and promptly discover and warn of potential safety hazards.

[0031] Information interaction module: This module is responsible for calling the dynamic data and early warning data in the reservoir simulation and safety prediction modules, and presenting these data to users in a visual way. Through intuitive data display, users can easily understand the results of reservoir simulation and early warning information, so as to make more accurate decisions.

[0032] Application management module: This module is mainly used to schedule and manage the associated parameters and reservoir parameters imported into the reservoir simulation and safety prediction modules. At the same time, it will also synchronously record dynamic tracking monitoring and early warning information to provide strong data support for subsequent reservoir management.

[0033] Program modularization module: This module integrates and encapsulates the above modules to form an executable program software. Through modular design, the system has good scalability and maintainability, and can easily adapt to future technology upgrades and changes in user needs.

[0034] The following is a specific working scenario to demonstrate the workflow of the intelligent reservoir simulation and safety prediction system:

[0035] Suppose an oil field is planning a new mining project and needs to simulate and predict the target reservoir in detail to ensure the safety and economy of the mining plan. The project team decides to use an intelligent reservoir simulation and safety prediction system to complete this task.

[0036] First, the project team input the relevant numerical simulation parameters of the target reservoir through the human-machine interface of the parameter setting module. These parameters include well network parameters such as well type, location, and control radius, as well as reservoir parameters such as reservoir lithology and permeability. At the same time, the team also set the type of fluid and production data according to the needs of the mining plan.

[0037] After completing the parameter setting, the system automatically passes these parameters to the numerical simulation software generation module. This module uses the neural network model and intelligent optimization algorithm, combined with the preset parameterized model, to dynamically generate numerical simulation software that matches the target reservoir. The generated software is stored in the location specified by the project team for subsequent calls.

[0038] Next, the reservoir simulation and safety prediction module called the generated numerical simulation software and began to simulate the target reservoir. During the simulation process, the module imported the associated parameters in the numerical simulation software and the reservoir parameters generated by the machine learning algorithm, and simulated the dynamic changes of the reservoir in detail. At the same time, the module also monitored the status of the reservoir in real time and discovered potential safety hazards in a timely manner based on the preset early warning rules.

[0039] The information interaction module presents these simulation results and early warning information to the project team in a visual way. Through intuitive data display, the team can clearly understand the dynamic changes and potential risks of the target reservoir, and thus formulate a more reasonable exploitation plan. In the application management module, the project team dispatched and managed the associated parameters and reservoir parameters imported during the simulation process. At the same time, they also synchronously recorded dynamic tracking monitoring and early warning information, providing strong data support for subsequent reservoir management. With the help of the intelligent reservoir simulation and safety prediction system, the project team successfully completed the simulation and prediction tasks of the target reservoir. They not only obtained detailed reservoir dynamic data, but also discovered potential safety hazards and formulated corresponding countermeasures in a timely manner. This not only improves the safety and economy of the exploitation plan, but also provides strong technical support for subsequent reservoir management.

[0040] The intelligent reservoir simulation and safety prediction system provides efficient, accurate and reliable solutions for reservoir management through its advanced working principle and modular design. In specific work scenarios, the system can help the project team quickly complete reservoir simulation and prediction tasks, and provide strong technical support for the formulation and implementation of mining plans.

[0041] The numerical simulation software generation module is also provided with a data analysis and inspection module, a data cleaning module, a feature engineering module and a feature selection module, wherein the data analysis and inspection module is used to check whether the input numerical simulation parameters meet the parameter requirements of the preset neural network model; the data cleaning module is used to delete redundant information in the input numerical simulation parameters and divide the input numerical simulation parameters into a training set and a verification set; the feature engineering module is used to perform feature engineering processing on the training set data; the feature engineering processing includes data binning, data encoding, data normalization, data splitting and data crossing; the feature selection module is used to perform feature selection on the training set and the verification set after the data split based on the machine learning algorithm and the feature correlation algorithm to obtain the optimal numerical simulation parameters.

[0042] In the reservoir simulation project, the project team used the intelligent reservoir simulation and safety prediction system to generate numerical simulation software suitable for specific reservoirs. After entering the numerical simulation parameters, the data analysis and inspection module first verifies whether these parameters meet the parameter requirements of the preset neural network model to ensure the accuracy and effectiveness of subsequent processing. Then, the data cleaning module deletes redundant information of the input parameters and divides the data into training sets and validation sets, providing a basis for subsequent feature engineering and feature selection.

[0043] The feature engineering module performs in-depth processing on the training set data, including data binning, data encoding, data normalization, data splitting and data crossover, to improve the availability of data and the predictive ability of the model. The feature selection module uses machine learning algorithms and feature correlation algorithms to select features from the processed data to obtain the optimal numerical simulation parameters that best reflect the reservoir characteristics, thereby improving the accuracy and adaptability of the numerical simulation software.

[0044] This implementation adds multiple steps such as data analysis and inspection, data cleaning, feature engineering and feature selection to the process of generating numerical simulation software, forming a complete data preprocessing and feature optimization process. The introduction of these steps not only improves the accuracy and availability of data, but also optimizes the parameter settings through feature selection, making the generated numerical simulation software more consistent with the actual characteristics of the reservoir. In addition, this implementation also makes full use of machine learning algorithms and feature correlation algorithms to improve the intelligence level of feature selection.

[0045] The steps of the feature selection module for feature selection of the training set and validation set are as follows: firstly, the data type is analyzed, and a feature dictionary is constructed according to the variable characteristics and the classification data, and the continuous variables are divided into numerical and categorical types; then, the continuous variables are screened based on the linear regression algorithm to obtain the sensitive factors of the continuous variables, and a multivariate regression model is established for each continuous variable according to the feature dictionary, and the sensitive factors are screened based on the R-square index; finally, the categorical variables are screened based on the random forest algorithm to obtain the sensitive factors of the categorical variables, a decision tree is established and its corresponding decision weights are assigned.

[0046] In the feature selection process, the project team first built a feature dictionary based on the data type, dividing continuous variables into numerical and categorical types. Then, the continuous variables were screened based on the linear regression algorithm to obtain sensitive factors, and a multivariate regression model was established for each continuous variable, and the sensitive factors were further screened based on the R-squared indicator. Finally, the categorical variables were screened using the random forest algorithm to obtain the sensitive factors of the categorical variables, and a decision tree model was established to assign corresponding decision weights.

[0047] This implementation method uses a more sophisticated and intelligent method in the feature selection process. By constructing a feature dictionary and classifying and processing continuous variables and categorical variables, this implementation method can more accurately identify feature variables that have an important impact on reservoir simulation results. At the same time, the linear regression algorithm and random forest algorithm are used for feature screening and decision tree model construction, which improves the accuracy and intelligence level of feature selection. This feature selection method that comprehensively applies multiple machine learning algorithms has significant innovation and practicality in the field of reservoir simulation.

[0048] The reservoir simulation and safety prediction module also includes a dynamic analysis module and a dynamic monitoring module. The dynamic analysis module is used to dynamically track the reservoir, regularly update the associated numerical simulation software parameters, and import the dynamic tracking data into the application management module for backup; the dynamic monitoring module is used to monitor the reservoir and obtain reservoir monitoring data; the associated numerical simulation software, reservoir monitoring data and machine learning algorithm are coupled to obtain real-time dynamic reservoir safety warnings.

[0049] During the reservoir simulation process, the dynamic analysis module dynamically tracks the reservoir and regularly updates the parameters of the associated numerical simulation software. These dynamic tracking data are imported into the application management module for backup and subsequent analysis and use. At the same time, the dynamic monitoring module monitors the reservoir in real time and obtains reservoir monitoring data. The associated numerical simulation software, reservoir monitoring data and machine learning algorithms are combined to generate real-time dynamic reservoir safety warnings, providing timely and effective decision support for reservoir management.

[0050] Compared with the existing technology, this implementation method adds two modules, dynamic analysis and dynamic monitoring, to the process of reservoir simulation and safety prediction, and realizes real-time tracking and monitoring of reservoirs and reservoirs. This dynamic management method not only improves the accuracy and timeliness of reservoir simulation, but also provides more comprehensive and timely information support for reservoir management. In addition, this implementation method also makes full use of the advantages of machine learning algorithms and associated numerical simulation software to improve the intelligence level of reservoir safety warning.

[0051] The dynamic analysis module decomposes the operation process of the numerical simulation parameters set by the user to obtain decomposition parameters, performs calculations based on the decomposition parameters, and the reservoir simulation and safety prediction module transmits the calculation results of the decomposition parameters to the parameter setting module and the information interaction module for the user to view. During the dynamic analysis process, the project team decomposes the operation process of the numerical simulation parameters set by the user to obtain decomposition parameters. Then, calculations are performed based on these decomposition parameters, and the calculation results are transmitted to the parameter setting module and the information interaction module for the user to view and analyze. This decomposition and calculation method enables the project team to have a deeper understanding of the calculation process and results of the numerical simulation parameters, thereby providing a more accurate and reliable basis for reservoir simulation.

[0052] Compared with the prior art, this implementation adopts the method of parameter decomposition and calculation in the dynamic analysis process, so that the project team can more accurately grasp the calculation process and results of the numerical simulation parameters. This refined analysis method not only improves the accuracy of reservoir simulation, but also provides users with more detailed and intuitive information display. In addition, this implementation also feeds back the calculation results to users in real time through the information interaction module, enhancing user participation and interactivity. In the dynamic monitoring process, the numerical simulation parameters set by the user are screened and optimized to obtain the preferred numerical simulation parameters. Then, these preferred parameters are preprocessed, including data cleaning, data reduction, feature extraction, feature binning and feature selection. Then, the preprocessed data is intelligently decided by using the preset machine learning algorithm to obtain the intelligent decision-making calculation results. Finally, the reservoir is monitored in real time according to these calculation results to obtain reservoir monitoring data.

[0053] This implementation method uses more sophisticated and intelligent data processing and decision-making methods in the dynamic monitoring process. By screening and optimizing numerical simulation parameters, performing data preprocessing and intelligent decision-making operations, this implementation method can more accurately identify the changing trends and potential risks of the reservoir. At the same time, the use of machine learning algorithms for intelligent decision-making operations improves the accuracy and timeliness of monitoring. This dynamic monitoring method that integrates multiple technologies and methods has significant innovation and practicality in the field of reservoir management.

[0054] The dynamic monitoring module screens the numerical simulation parameters set by the user to obtain the screened preferred numerical simulation parameters, performs data preprocessing on the preferred numerical simulation parameters to obtain preprocessed data; wherein the data preprocessing includes data cleaning, data reduction, feature extraction, feature binning and feature selection; performs intelligent decision-making operation on the preprocessed data according to a preset machine learning algorithm to obtain the intelligent decision-making operation result, monitors the reservoir according to the intelligent decision-making operation result, and obtains reservoir monitoring data.

[0055] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. Intelligent reservoir simulation and safety prediction system, characterized by: It includes data interaction among parameter setting module, numerical simulation software generation module, reservoir simulation and safety prediction module, information interaction module, application management module and program modularization module; The parameter setting module is used to obtain the numerical simulation parameters set by the user from the user terminal, and import the numerical simulation parameters into the numerical simulation software generation module; The numerical simulation parameters are configured in the human-machine interface of the parameter setting module, and the numerical simulation parameters include: well, reservoir, fluid and well network parameters; the well network parameters include well pattern, well group size, well name and well number, well grid mapping, well location, well control radius, well grid influence range and well production data; After receiving the numerical simulation parameters, the numerical simulation software generation module combines the acquired numerical simulation parameters with the preset parameterized model and the intelligent optimization algorithm through the neural network model to generate the associated numerical simulation software, and stores the associated numerical simulation software in the storage location set by the user; The reservoir simulation and safety prediction module calls the associated numerical simulation software, and then simulates the reservoir parameters by importing the associated parameters and reservoir parameters in the numerical simulation software, and performs dynamic tracking monitoring and early warning; wherein the reservoir parameters are generated by combining the numerical simulation parameters set by the user with the reservoir parameters generated based on the machine learning algorithm to simulate the parameters; The information interaction module calls the dynamic data and early warning data in the reservoir simulation and safety prediction module, and visualizes the data simulation results to the user; The application management module is used to schedule the associated parameters and reservoir parameters imported from the reservoir simulation and safety prediction module, and simultaneously store and record the dynamic tracking monitoring and early warning information; The program modularization module is used to integrate the parameter setting module, the numerical simulation software generation module, the reservoir simulation and safety prediction module, the information interaction module and the application management module and package them into executable program software.

2. The intelligent reservoir simulation and safety prediction system according to claim 1, characterized in that: The numerical simulation software generation module is also provided with a data analysis and inspection module, a data cleaning module, a feature engineering module and a feature selection module, wherein the data analysis and inspection module is used to check whether the input numerical simulation parameters meet the parameter requirements of the preset neural network model; the data cleaning module is used to delete redundant information in the input numerical simulation parameters and divide the input numerical simulation parameters into a training set and a verification set; the feature engineering module is used to perform feature engineering processing on the training set data; the feature engineering processing includes data binning, data encoding, data normalization, data splitting and data crossing; the feature selection module is used to perform feature selection on the training set and the verification set after the data split based on the machine learning algorithm and the feature correlation algorithm to obtain the optimal numerical simulation parameters.

3. The intelligent reservoir simulation and safety prediction system according to claim 2, characterized in that: The steps of the feature selection module for feature selection of the training set and validation set are as follows: firstly, the data type is analyzed, and a feature dictionary is constructed according to the variable characteristics and the classification data, and the continuous variables are divided into numerical and categorical types; then, the continuous variables are screened based on the linear regression algorithm to obtain the sensitive factors of the continuous variables, and a multivariate regression model is established for each continuous variable according to the feature dictionary, and the sensitive factors are screened based on the R-square index; finally, the categorical variables are screened based on the random forest algorithm to obtain the sensitive factors of the categorical variables, a decision tree is established and its corresponding decision weights are assigned.

4. The intelligent reservoir simulation and safety prediction system according to claim 1, characterized in that: The reservoir simulation and safety prediction module also includes a dynamic analysis module and a dynamic monitoring module. The dynamic analysis module is used to dynamically track the reservoir, regularly update the associated numerical simulation software parameters, and import the dynamic tracking data into the application management module for backup; The dynamic monitoring module is used to monitor the reservoir and obtain reservoir monitoring data; the associated numerical simulation software, reservoir monitoring data and machine learning algorithm are coupled to obtain real-time dynamic reservoir safety warning.

5. The intelligent reservoir simulation and safety prediction system according to claim 4, characterized in that: The dynamic analysis module decomposes the calculation process of the numerical simulation parameters set by the user to obtain decomposition parameters, and performs calculations based on the decomposition parameters. The reservoir simulation and safety prediction module transmits the calculation results of the decomposition parameters to the parameter setting module and the information interaction module for the user to view.

6. The intelligent reservoir simulation and safety prediction system according to claim 4, characterized in that: The dynamic monitoring module screens the numerical simulation parameters set by the user to obtain the screened optimal numerical simulation parameters, and performs data preprocessing on the optimal numerical simulation parameters to obtain preprocessed data; The data preprocessing includes data cleaning, data reduction, feature extraction, feature binning and feature selection; the preprocessed data is subjected to intelligent decision-making operation according to a preset machine learning algorithm to obtain the intelligent decision-making operation result; the reservoir is monitored according to the intelligent decision-making operation result to obtain reservoir monitoring data.

Images