A reservoir production optimization method and system based on expensive optimization

Abstract

The present invention relates to a reservoir production optimization method and system based on expensive optimization. The method includes: obtaining a plurality of reservoir production parameters of a target oilfield, and constructing an expensive optimization problem model according to the plurality of reservoir production parameters; the reservoir production parameters include water injection cost, oil price, and water removal cost; determining a test function and test parameters of the expensive optimization problem model; based on the plurality of reservoir production parameters, performing multiple iterations on the expensive optimization problem model through a surrogate model and a grey prediction evolutionary algorithm until the number of iterations reaches a preset value, to obtain the optimal solution of the expensive optimization problem model; simulating the expensive optimization problem model based on the test function and test parameters to verify the effectiveness of the expensive optimization problem model. The present invention applies the surrogate model and the grey prediction evolutionary algorithm to the expensive optimization problem, further improving the optimization effect and convergence speed of the surrogate-assisted evolutionary algorithm in solving the expensive optimization problem.

Description

Technical Field

[0001] The present invention belongs to the technical field of reservoir production and intelligent algorithms, and particularly relates to a reservoir production optimization method and system based on expensive optimization. Background Art

[0002] Reservoir production optimization regards the reservoir production system as a complex system, and through means such as mathematical methods, numerical simulation techniques, and optimization algorithms, various parameters in the production process (such as water injection volume, gas injection volume, well location layout, production rate, etc.) are regulated to seek the optimal production plan.

[0003] In the field of engineering applications, many optimization problems require expensive costs to calculate the values of objective functions, metrics, and constraints, which are usually referred to as expensive optimization problems. With the continuous progress of society and the emerging challenges brought about by the era of smart cities, the Internet of Things, and big data, more efficiently solving expensive optimization problems is becoming a necessary condition for promoting the prosperous development of various fields.

[0004] Currently, the surrogate-assisted evolutionary algorithm has become the mainstream technology for solving expensive optimization problems. Its basic idea is to use a surrogate model to replace the expensive function for evaluation, define a pre-screening strategy, and select the individuals that need to be evaluated by the expensive function from the search population of the evolutionary algorithm, thereby reducing the computational cost. Commonly used surrogate models include the Kriging model, polynomial regression surface (PRS), radial basis function (RBF), and support vector regression (SVR), etc.

[0005] Although the existing surrogate-assisted evolutionary algorithms show superior performance in solving expensive optimization problems, for high-dimensional expensive optimization problems, the balance between the convergence speed and optimization effect of the surrogate-assisted evolutionary algorithm needs to be further enhanced.

[0006] Disadvantages of the prior art:

[0007] (1) Traditional evolutionary algorithms perform population search by mutating and crossing between current individuals and historical individuals. The surrogate model affects the update of the next-generation population by providing predicted fitness values for the offspring. When the surrogate model cannot provide accurate prediction values, it may mislead the population search.

[0008] (2) The existing surrogate-assisted evolutionary algorithms are overly dependent on the surrogate model. Most algorithms still require a large amount of computational resources to obtain sufficient sample data to construct a high-precision surrogate model before obtaining high-quality solutions. Summary of the Invention

[0009] To improve the accuracy and convergence speed of expensive optimization model for reservoir production optimization, in the first aspect of the present invention, a reservoir production optimization method based on expensive optimization is provided, including: obtaining multiple reservoir production parameters of a target oilfield, and constructing an expensive optimization problem model according to the multiple reservoir production parameters; the reservoir production parameters include water injection cost, oil price, and water removal cost; determining a test function and test parameters of the expensive optimization problem model; based on the multiple reservoir production parameters, iterating the expensive optimization problem model multiple times through a surrogate model and a grey prediction evolutionary algorithm until the number of iterations reaches a preset value to obtain the optimal solution of the expensive optimization problem model; simulating the expensive optimization problem model based on the test function and test parameters to verify the effectiveness of the expensive optimization problem model.

[0010] In some embodiments of the present invention, the determining the test function and test parameters of the expensive optimization problem model includes: using multiple benchmark functions as test functions; determining the population parameters, sample parameters, problem dimension, and maximum number of evaluations of each test function.

[0011] In some embodiments of the present invention, the iterating the expensive optimization problem model multiple times through a surrogate model and a grey prediction evolutionary algorithm until the number of iterations reaches a preset value to obtain the optimal solution of the expensive optimization problem model based on the multiple reservoir production parameters includes: randomly generating the first three generations of populations within the feasible region of the expensive optimization problem model based on the multiple reservoir production parameters; performing global search within the search space through a global surrogate model and a grey prediction evolutionary algorithm based on the first three generations of populations; adding the individual with the optimal fitness value to the candidate sample set to participate in the next iteration; selecting multiple samples with the optimal fitness value from the candidate sample set through a surrogate model to generate multiple offspring individuals; performing local search among the multiple offspring individuals to screen out the individual with the optimal fitness value; based on the optimal function value, alternately performing global search and local search to screen out the individual with the optimal fitness value until the number of iterations reaches a preset value.

[0012] Further, the randomly generating the first three generations of populations within the feasible region of the expensive optimization problem model based on the multiple reservoir production parameters includes: generating the first generation of population using a random number distribution in the search space; generating the second and third generations of populations based on the first generation of population through a heuristic algorithm.

[0013] Further, the performing global search within the search space through a global surrogate model and a grey prediction evolutionary algorithm based on the first three generations of populations includes: generating an initial sample through Latin hypercube sampling; constructing a global radial basis function surrogate model to predict the fitness value of the population; generating offspring through a grey prediction evolutionary algorithm, comparing the predicted fitness values of the generated offspring with those of the parents, and retaining the individual with the optimal predicted fitness value among the offspring.

[0014] In the above embodiments, constructing the expensive optimization problem model according to the multiple reservoir production parameters includes: constructing an objective function based on the multiple reservoir production parameters; determining multiple constraint parameters of the expensive optimization problem model based on the synthetic reservoir model and the multiple reservoir production parameters; and constructing the expensive optimization problem model according to the objective function and the multiple constraint parameters.

[0015] In a second aspect of the present invention, there is provided an oil reservoir production optimization system based on expensive optimization, including: an acquisition module, configured to acquire multiple reservoir production parameters of a target oilfield and construct an expensive optimization problem model according to the multiple reservoir production parameters; the reservoir production parameters include water injection cost, oil price, and water removal cost; a determination module, configured to determine a test function and test parameters of the expensive optimization problem model; an iteration module, configured to perform multiple iterations on the expensive optimization problem model through a surrogate model and a grey prediction evolutionary algorithm based on the multiple reservoir production parameters until the number of iterations reaches a preset value, to obtain an optimal solution of the expensive optimization problem model; and a verification module, configured to simulate the expensive optimization problem model based on the test function and the test parameters to verify the effectiveness of the expensive optimization problem model.

[0016] In a third aspect of the present invention, there is provided an electronic device, including: one or more processors; a storage device, configured to store one or more programs, and when the one or more programs are executed by the one or more processors, enable the one or more processors to implement the expensive optimization-based oil reservoir production optimization method provided by the present invention in the first aspect.

[0017] In a fourth aspect of the present invention, there is provided a computer-readable medium, on which a computer program is stored, wherein when the computer program is executed by a processor, it implements the expensive optimization-based oil reservoir production optimization method provided by the present invention in the first aspect.

[0018] The beneficial effects of the present invention are as follows:

[0019] (1) The present invention uses a grey prediction evolutionary algorithm to perform global search with the assistance of a surrogate model to find the optimal value, inheriting the advantages of the grey prediction evolutionary algorithm of relatively fast convergence speed and fewer required parameters, and reducing its technical complexity;

[0020] (2) In the iterative process of the present invention, by alternately performing global search and local search, the method of finding the optimal value is optimized, the calculation cost is reduced, and the calculation efficiency is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic flowchart of the basic process of the expensive optimization-based oil reservoir production optimization method in some embodiments of the present invention;

[0022] Figure 2 Schematic diagram of the principle of the reservoir production optimization method based on expensive optimization in some embodiments of the present invention;

[0023] Figure 3 Schematic diagram of the structure of the reservoir production optimization system based on expensive optimization in some embodiments of the present invention;

[0024] Figure 4 Schematic diagram of the structure of an electronic device in some embodiments of the present invention. Detailed implementation manners

[0025] The principles and features of the present invention will be described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0026] Reference Figure 1 And Figure 2 , in the first aspect of the present invention, there is provided a reservoir production optimization method based on expensive optimization, including: S100. Obtain multiple reservoir production parameters of a target oilfield, and construct an expensive optimization problem model according to the multiple reservoir production parameters; the reservoir production parameters include water injection cost, oil price, and water removal cost; S200. Determine a test function and test parameters of the expensive optimization problem model; S300. Based on the multiple reservoir production parameters, perform multiple iterations on the expensive optimization problem model through a surrogate model and a grey prediction evolutionary algorithm until the number of iterations reaches a preset value to obtain an optimal solution of the expensive optimization problem model; S400. Simulate the expensive optimization problem model based on the test function and test parameters to verify the effectiveness of the expensive optimization problem model.

[0027] In step S100 of some embodiments of the present invention, the constructing an expensive optimization problem model according to the multiple reservoir production parameters includes:

[0028] S101. Construct an objective function based on multiple reservoir production parameters;

[0029] Specifically, the calculation formula for economic benefit during waterflooding reservoir production is as follows:

[0030] ,

[0031] In the formula, J is the objective function of decision variable x and state variable and can be calculated by a numerical simulator . is the time step, , and are the oil price, water removal cost, and water injection cost, respectively, in US dollars per barrel. , and are the oil production rates of all production wells at time step , the water production rates of all production wells at time step , and the injection flow rates of all injection wells at time step , all in barrels per day. represents the annual discount rate, represents the elapsed time at the th time step, in years.

[0032] A synthetic reservoir model - Egg Model was used for simulation experiments. Random sampling was to randomly select 200 samples within the designed water injection interval. The population size of GPE was 200, and the initial value of the maximum number of iterations t max was set to 1000. Other parameters used in the present invention were set according to the parameters usually given in existing literature. To reduce the error caused by randomness, each group of experiments was independently run 30 times. The simulation experiment set the water injection cost at $5 per barrel, the oil price at $80 per barrel, and the water removal cost at $19 per barrel.

[0033] S102. Based on the synthetic reservoir model and multiple reservoir production parameters, determine multiple constraint parameters of the expensive optimization problem model;

[0034] S103. According to the objective function and multiple constraint parameters, construct the expensive optimization problem model.

[0035] Based on the economic benefits generated during the water - flooding reservoir production as the objective function and the synthetic reservoir model as the constraint parameter, construct the expensive optimization problem model.

[0036] In step S200 of some embodiments of the present invention, the determination of the test function and test parameters of the expensive optimization problem model includes:

[0037] S201. Use multiple benchmark functions as test functions;

[0038] Specifically, 12 benchmark functions were used to test the proposed method. Among them, Ellipsoid, Rosenbrock, Ackley, Griewank, and Rastrigin are common benchmark functions. Three very complex multimodal functions were selected from the CEC2005 test set, and four complex benchmark functions were selected from the CEC2017 test set. These test functions are widely used for the performance test of surrogate - assisted evolutionary algorithms.

[0039] S202. Determine the population parameters, sample parameters, problem dimension, and maximum number of evaluations for each test function. Specifically, the parameters used are the size of the population, the size of the initial sample, and the optimal sample size used to construct the local surrogate model, all of which are set to twice the problem dimension, and the maximum number of evaluations is set to 1000. Other parameters used in the present invention are set according to the parameters usually given in the existing literature. To reduce the error caused by randomness, each group of experiments is independently run 30 times. Specifically, in the initialization process, three generations of populations must be randomly generated within the feasible region. The

[0040] In step S300 of some embodiments of the present invention, the method of obtaining the optimal solution of the expensive optimization problem model by performing multiple iterations on the expensive optimization problem model through the surrogate model and the grey prediction evolutionary algorithm until the number of iterations reaches a preset value based on the multiple reservoir production parameters includes:

[0041] S301. Randomly generate the first three generations of populations within the feasible region of the expensive optimization problem model based on the multiple reservoir production parameters;

[0042] Specifically, in the initialization process, three generations of populations must be randomly generated within the feasible region. The t th i individual of the j th generation of the population is denoted as

[0043] ,

[0044] where D is the dimension of the problem; and are the upper and lower bounds of the j th dimension respectively; represents a random number generated uniformly within the range [0, 1]. The first three generations of populations , , ( t≥ 2) are subjected to boundary processing.

[0045] S302. Based on the first three generations of populations, perform a global search within the search space through the global surrogate model and the grey prediction evolutionary algorithm; add the individual with the optimal fitness value to the candidate sample set to participate in the next iteration;

[0046] Specifically, the global surrogate model focuses on exploring the optimal solution in the entire search space. Using the samples in the database, train the global RBF surrogate model to evaluate the offspring to obtain the predicted fitness value. The functional expression of RBF is:

[0047] ,

[0048] wherein and represent n input points in an n-dimensional space and their corresponding output values in the output space. represents the function value calculated by the RBF interpolation function, represents the weight coefficient, is the kernel function. The present method uses a cubic function as the kernel function. is the Euclidean distance between c and . represents a polynomial.

[0049] In the reproduction operation of the grey prediction evolutionary algorithm, the population sequence is regarded as a time series, three unordered original data are converted into a sequence with the property of an approximate exponential function, and the offspring are predicted by constructing an exponential function. Three consecutive populations , , ( t ≥ 2) are regarded as a population sequence. Three individuals , , ( t≥ 2) are randomly selected from the population sequence , and . Let , , , , . Assume that the absolute value of the difference between the maximum value and the minimum value among the three individuals is less than the difference threshold set by the grey prediction evolutionary algorithm, then random perturbation is used to generate new individuals; if the absolute value of the difference between any two of the three selected individuals is less than the difference threshold, linear fitting is used to generate new individuals, otherwise, an even grey model is used to generate new individuals.

[0050] The test individual can be generated by the following formula:

[0051] ,

[0052] wherein is a parameter for controlling the difference threshold of prediction, and is set to in the present invention. , and are the grey development coefficient, grey control parameter and perturbation coefficient respectively. The formulas for calculating , and are as follows:

[0053] ,

[0054] Each individual in the generated test population is to be subjected to constraint processing. And for the test individuals that meet the constraints predictive fitness evaluation is performed to obtain the predictive fitness value .

[0055] Next, by comparing the predictive fitness values of the target individuals and the test individuals , the better individuals are retained to enter the next generation. The selection operation can be carried out through the following formula:

[0056] ;

[0057] Then, the expensive function f is used to perform a true evaluation on the individual with the optimal predictive fitness value in the test population to obtain the objective function value , and the evaluated individual is added to the database.

[0058] S303. From the candidate sample set, multiple samples with the optimal fitness value are selected through the surrogate model to generate multiple offspring individuals; local search is performed among the multiple offspring individuals to screen out the individual with the optimal fitness value;

[0059] Specifically, the local surrogate model can improve the search speed in the promising regions. Based on the optimal samples in the database, a local RBF surrogate model is trained to predict the true function values of the offspring. Similarly, the local surrogate model uses the cubic function as the kernel function for constructing the RBF model. The even grey model in the grey pre-evolution algorithm is used to generate offspring, and the optimal individual among the offspring is selected for true evaluation and added to the database in the same way as the global surrogate model. Until the maximum number of iterations is reached, the algorithm ends and the optimal solution is output.

[0060] S304. Based on the optimal function value, global search and local search are alternately performed to screen out the individual with the optimal fitness value until the number of iterations reaches the preset value.

[0061] In step S400 of some embodiments of the present invention, based on the test function and test parameters, the expensive optimization problem model is simulated to verify the effectiveness of the expensive optimization problem model. Specifically, the performance of the model on 12 benchmark functions is tested. The results show that the algorithm of the present invention is generally superior to other algorithms, verifying that the solution obtained by the present invention is better, demonstrating the feasibility, effectiveness and superiority of the present invention.

[0062] Embodiment 2

[0063] Reference Figure 3 Figure 3 , the second aspect of the present invention provides an expensive optimization-based reservoir production optimization system 1, including: an acquisition module 11 for acquiring a plurality of reservoir production parameters of a target oilfield and constructing an expensive optimization problem model according to the plurality of reservoir production parameters; the reservoir production parameters include water injection cost, oil price, and water removal cost; a determination module 12 for determining a test function and test parameters of the expensive optimization problem model; an iteration module 13 for iterating the expensive optimization problem model multiple times based on the plurality of reservoir production parameters through a surrogate model and a grey prediction evolutionary algorithm until the number of iterations reaches a preset value to obtain the optimal solution of the expensive optimization problem model; a verification module 14 for simulating the expensive optimization problem model based on the test function and test parameters to verify the effectiveness of the expensive optimization problem model.

[0064] Further, the iteration module 13 includes: a generation unit for randomly generating the first three generations of populations within the feasible region of the expensive optimization problem model based on the plurality of reservoir production parameters; a global unit for performing global search within the search space through a global surrogate model and a grey prediction evolutionary algorithm based on the first three generations of populations; adding the individual with the optimal fitness value to the candidate sample set to participate in the next iteration; a local unit for selecting a plurality of samples with the optimal fitness value from the candidate sample set through a surrogate model to generate a plurality of offspring individuals; performing local search among the plurality of offspring individuals to screen out the individual with the optimal fitness value; an iteration unit for alternately performing global search and local search based on the optimal function value to screen out the individual with the optimal fitness value until the number of iterations reaches a preset value.

[0065] Embodiment 3

[0066] Reference Figure 4 Figure 4 , the third aspect of the present invention provides an electronic device, including: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, enabling the one or more processors to implement the expensive optimization-based reservoir production optimization method of the first aspect of the present invention.

[0067] The electronic device 500 may include a processing device (such as a central processing unit, a graphics processing unit, etc.) 501, which may perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 508 into a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the electronic device 500 are also stored. The processing device 501, the ROM 502, and the RAM 503 are connected to each other through a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0068] Typically, the following devices may be connected to the I / O interface 505: an input device 506 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 508 including, for example, a hard disk, etc.; and a communication device 509. The communication device 509 may allow the electronic device 500 to communicate with other devices wirelessly or wireline to exchange data. Although Figure 4 the electronic device 500 with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. Instead, more or fewer devices may be implemented or had. Figure 4 Each block shown in

[0069] In particular, according to an embodiment of the present disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, an embodiment of the present disclosure includes a computer program product that includes a computer program carried on a computer-readable medium, and the computer program includes program code for performing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network via a communication device 509, or installed from a storage device 508, or installed from a ROM 502. When the computer program is executed by a processing device 501, the above-described functions defined in the methods of the embodiments of the present disclosure are performed. It should be noted that the computer-readable medium described in the embodiments of the present disclosure can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the embodiments of the present disclosure, the computer-readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the embodiments of the present disclosure, the computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and the computer-readable signal medium can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0070] The above computer-readable medium can be included in the above electronic device; or it can exist separately and not be assembled into the electronic device. The above computer-readable medium carries one or more computer programs, and when the above one or more programs are executed by the electronic device, the electronic device is caused to:

[0071] Computer program code for performing the operations of the embodiments of the present disclosure may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, Python, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may execute entirely on the user's computer, partially on the user's computer, execute as a stand-alone software package, execute partially on the user's computer and partially on a remote computer, or execute entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0072] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions noted in the blocks may occur in a different order than noted in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0073] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for optimizing oil reservoir production based on costly optimization, characterized in that: include: Acquire multiple reservoir production parameters of the target oil field, and construct an expensive optimization problem model according to the multiple reservoir production parameters: construct an objective function based on the multiple reservoir production parameters; Based on the synthetic reservoir model and multiple reservoir production parameters, multiple constraint parameters of the expensive optimization problem model are determined; the expensive optimization problem model is constructed according to the objective function and the multiple constraint parameters; the reservoir production parameters include water injection cost, oil price and water removal cost; wherein the objective function is expressed as: Where J is the objective function of the decision variable x and the state variable s; n is the time step, r0, r w and r i are the oil price, water removal cost and water injection cost, all in US dollars per barrel; Q 0,t , Q w，t and Q i,t are the oil production rate of all production wells at time step t, the water production rate of all production wells at time step t, and the injection flow rate of all injection wells at time step t, all in barrels per day; b represents the annual discount rate, p t represents the elapsed time of the t-th time step, in years; Determining a test function and test parameters of the expensive optimization problem model; Based on the multiple reservoir production parameters, the expensive optimization problem model is iterated multiple times by using a proxy model and a grey prediction evolutionary algorithm until the number of iterations reaches a preset value, thereby obtaining an optimal solution to the expensive optimization problem model; The expensive optimization problem model is simulated based on the test function and the test parameters to verify the validity of the expensive optimization problem model.

2. The method for optimizing oil reservoir production based on costly optimization according to claim 1, characterized in that: The test function and test parameters for determining the expensive optimization problem model include: Use multiple benchmark functions as test functions; Determine the population parameter, sample parameter, problem dimension, and maximum number of evaluations for each test function.

3. The method for optimizing oil reservoir production based on costly optimization according to claim 1, characterized in that: The expensive optimization problem model is iterated multiple times based on the multiple reservoir production parameters by using the proxy model and the grey prediction evolutionary algorithm until the number of iterations reaches a preset value, and obtaining the optimal solution of the expensive optimization problem model includes: Based on the plurality of reservoir production parameters, randomly generating first three generations of populations within a feasible region of the expensive optimization problem model; Based on the first three generations of populations, a global search is performed in the search space through a global proxy model and a grey prediction evolutionary algorithm; individuals with the best fitness value are added to the sample set to be selected to participate in the next iteration; From the candidate sample set, multiple samples with the best fitness values ​​are selected through the proxy model to generate multiple offspring individuals; local search is performed among the multiple offspring individuals to screen out the individual with the best fitness value; Based on the optimal function value, global search and local search are performed alternately to screen the individuals with the best fitness value until the number of iterations reaches the preset value.

4. The method for optimizing oil reservoir production based on costly optimization according to claim 3, characterized in that: The randomly generating the first three generations of populations in the feasible region of the expensive optimization problem model based on the multiple reservoir production parameters comprises: Generate the first generation population using random number distribution in the search space; Based on the first generation population, the second generation population and the third generation population are generated through a heuristic algorithm.

5. The method for optimizing oil reservoir production based on costly optimization according to claim 3, characterized in that: The global search in the search space based on the first three generations of populations by using a global proxy model and a grey prediction evolutionary algorithm includes: Generate the initial sample through Latin hypercube sampling; Construct a global radial basis function surrogate model to predict the fitness value of the population; The offspring are generated by the grey prediction evolutionary algorithm, the predicted fitness values ​​of the generated offspring are compared with those of the parent generation, and the individuals with the best predicted fitness values ​​among the offspring are retained.

6. A reservoir production optimization system based on costly optimization, characterized in that: include: An acquisition module is used to acquire a plurality of reservoir production parameters of a target oil field, and construct an expensive optimization problem model according to the plurality of reservoir production parameters: construct an objective function based on the plurality of reservoir production parameters; Based on the synthetic reservoir model and multiple reservoir production parameters, multiple constraint parameters of the expensive optimization problem model are determined; the expensive optimization problem model is constructed according to the objective function and the multiple constraint parameters; the reservoir production parameters include water injection cost, oil price and water removal cost; wherein the objective function is expressed as: Where J is the objective function of the decision variable x and the state variable s; n is the time step, r0, r w and r i are the oil price, water removal cost and water injection cost, all in US dollars per barrel; Q 0,t , Q w，t and Q i,t are the oil production rate of all production wells at time step t, the water production rate of all production wells at time step t, and the injection flow rate of all injection wells at time step t, all in barrels per day; b represents the annual discount rate, p t represents the elapsed time of the t-th time step, in years; A determination module, used to determine a test function and test parameters of the expensive optimization problem model; An iteration module, used to perform multiple iterations on the expensive optimization problem model based on the multiple reservoir production parameters through a proxy model and a grey prediction evolutionary algorithm until the number of iterations reaches a preset value, thereby obtaining an optimal solution to the expensive optimization problem model; The verification module is used to simulate the expensive optimization problem model based on the test function and the test parameters to verify the validity of the expensive optimization problem model.

7. The reservoir production optimization system based on costly optimization according to claim 6, characterized in that: The iteration module includes: A generation unit, configured to randomly generate first three generations of populations within a feasible region of the expensive optimization problem model based on the plurality of reservoir production parameters; A global unit is used to perform a global search in the search space based on the first three generations of populations through a global agent model and a grey prediction evolutionary algorithm; and add individuals with the best fitness value to a sample set to be selected so as to participate in the next iteration; The local unit is used to select multiple samples with the best fitness value from the candidate sample set through the proxy model to generate multiple offspring individuals; perform local search among the multiple offspring individuals to screen out the individual with the best fitness value; The iteration unit is used to alternately perform global search and local search based on the optimal function value to screen the individual with the best fitness value until the number of iterations reaches a preset value.

8. An electronic device, comprising: one or more processors; A storage device for storing one or more programs, which, when executed by the one or more processors, enables the one or more processors to implement the reservoir production optimization method based on costly optimization as described in any one of claims 1 to 5.

9. A computer readable medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the method for optimizing oil reservoir production based on costly optimization according to any one of claims 1 to 5 is implemented.

Images