Natural gas production scheduling optimization method, device and equipment and storage medium

Abstract

The invention relates to the technical field of natural gas production scheduling, and provides a natural gas production scheduling optimization method, device and equipment and a storage medium, and the method comprises the steps: obtaining natural gas production process data, constructing a natural gas production scheduling optimization model, carrying out the aggregation of a plurality of sub-objective functions in a multi-objective function, and obtaining a scalar cost value function; performing space division on the decision variables according to preset priori knowledge; constructing a Gaussian model according to the scalar cost value function and the plurality of decision variable space groups; constructing an additive structure dual-target acquisition function and an upper boundary confidence function according to the additive Gaussian model; solving the additive structure function to obtain a candidate solution of each decision variable space group; aggregating the candidate solutions of each decision variable space group to obtain a candidate solution set; and optimizing the candidate solution set to obtain an optimal natural gas production scheduling scheme. According to the embodiment of the invention, the optimization efficiency and precision of natural gas production scheduling can be improved.

Description

Technical Field

[0001] The embodiments in this specification relate to the field of natural gas production scheduling technology, and in particular to a natural gas production scheduling optimization method, apparatus, equipment, and storage medium. Background Technology

[0002] Production scheduling optimization in the natural gas industry chain is a complex and crucial process, encompassing multiple stages including production, transportation, storage, sales, and trade. Optimization allows for the rational arrangement of production plans, ensuring efficient equipment operation, thereby improving resource supply, reducing unnecessary energy waste, and ultimately enhancing the overall efficiency of the industry chain. In the production scheduling optimization problem of the natural gas industry chain, the numerous and interrelated parameters involved, the production process often involves multiple optimization objectives, and the constructed optimization models are quite complex. This results in high evaluation costs for each objective function in the multi-objective optimization problem, thus modeling the natural gas industry chain production scheduling optimization problem as a high-dimensional and expensive multi-objective optimization problem. As the dimensionality of the decision variable space of the optimization objectives increases, the computational complexity of the optimization algorithm typically increases exponentially, making it extremely difficult to find an accurate optimization solution in high-dimensional cases.

[0003] Most existing expensive multi-objective Bayesian optimization methods based on the expectation-enhancing criterion are only applicable to low-dimensional decision spaces and are extremely inefficient when solving high-dimensional problems. In multi-objective optimization problems with high-dimensional decision variable spaces, there may be interactions between decision variables, which jointly influence the objective function. Therefore, based on whether there are interactions between decision variables, multi-objective optimization problems can be divided into separable and inseparable problems. In separable problems, the interactions between decision variables are weak, and the decision space can be easily divided into multiple disjoint subspaces. This partitioning helps reduce the dimensionality of the decision space, thereby reducing sampling complexity and improving the optimization efficiency of the acquisition function. However, in inseparable problems, the interactions between decision variables are strong, making space partitioning difficult, and traditional dimensionality reduction methods are no longer applicable. Related multi-objective optimization methods do not fully consider the dependencies between decision variables when partitioning the space, resulting in some irrationality in the partitioning of the decision variable space and low optimization accuracy. Therefore, there is an urgent need for a natural gas production scheduling optimization method that can improve the optimization efficiency and accuracy of the natural gas production scheduling problem. Summary of the Invention

[0004] In view of the above-mentioned problems in the prior art, the purpose of the embodiments of this specification is to provide a natural gas production scheduling optimization method, apparatus, equipment and storage medium to solve the problems of low optimization efficiency and low optimization accuracy in the prior art for natural gas production scheduling.

[0005] To solve the above-mentioned technical problems, the specific technical solutions of the embodiments in this specification are as follows:

[0006] On the one hand, embodiments of this specification provide a natural gas production scheduling optimization method, the method comprising:

[0007] Acquire natural gas production process data, which includes raw material data, production operation data, product quality data, and storage and transportation data;

[0008] A natural gas production scheduling optimization model is constructed based on the natural gas production process data and production scheduling optimization objectives. The natural gas production scheduling optimization model includes constraints, multi-objective functions, and decision variables.

[0009] Several sub-objective functions are obtained from the multi-objective function, and the several sub-objective functions are aggregated using the augmented Chebyshev function to obtain the scalar cost function;

[0010] The decision variables are spatially partitioned based on pre-defined prior knowledge using Bayesian variational inference methods, resulting in several spatial groupings of decision variables.

[0011] Based on the scalar cost function and the spatial grouping of several decision variables, an additive Gaussian model of the scalar cost function is constructed.

[0012] Based on the additive Gaussian model, construct the additive structure biobjective acquisition function and the additive structure upper boundary confidence function for each decision variable spatial grouping;

[0013] The upper boundary confidence function of the additive structure and the bi-objective acquisition function of the additive structure are solved to obtain candidate solutions for each decision variable spatial grouping;

[0014] The candidate solutions of each decision variable space group are aggregated to obtain the candidate solution set of the natural gas production scheduling optimization model;

[0015] The candidate solution set is optimized to obtain the optimal natural gas production scheduling scheme.

[0016] Furthermore, the scalar cost function includes:

[0017] The scalar cost function can be expressed using the following formula:

[0018]

[0019] Among them, f λ (x) represents the scalar cost function, λ i f represents the weight of the sub-objective function of the i-th sub-objective function. i(x) represents the sub-objective function corresponding to the i-th optimization objective, M represents the number of sub-objective functions, and x is the decision variable.

[0020] Furthermore, the step of using Bayesian variational inference to spatially partition the decision variables based on pre-defined prior knowledge includes:

[0021] Based on a pre-set number of spatial groups, a latent variable array is constructed, wherein the latent variable parameters in the latent variable array are used to represent the spatial group to which each decision variable belongs;

[0022] Using Bayesian variational inference methods, the latent variable array is inferred based on pre-defined prior knowledge to obtain the posterior distribution function of the latent variable array;

[0023] Gibbs sampling is used to sample the latent variable parameters in the latent variable array of the posterior distribution function, and the decision variable spatial grouping of each decision variable is determined based on the sampled values.

[0024] Further, the step of constructing an additive Gaussian model of the scalar value function based on the scalar value function and the spatial grouping of several decision variables includes:

[0025] By taking the decision variables in each decision variable space grouping as variables in the scalar value function, several sub-scalar value functions are constructed.

[0026] Construct a Gaussian model of the cost function of each subscalar;

[0027] The additive Gaussian model of the scalar cost function is obtained based on the Gaussian model of each subscalar cost function.

[0028] Further, the step of constructing the additive structure biobjective acquisition function and the additive structure upper boundary confidence function for each decision variable spatial grouping based on the additive Gaussian model includes:

[0029] Construct a bi-objective acquisition function and an upper boundary confidence function for spatial grouping of each decision variable;

[0030] The additive Gaussian model is used to transform the bi-objective acquisition function and the upper boundary confidence function to obtain the additive bi-objective acquisition function and the additive upper boundary confidence function for each spatial group.

[0031] Furthermore, the addable biobjective acquisition function includes:

[0032] The additable structure bi-objective acquisition function is expressed by the following formula:

[0033]

[0034] in, This represents an additable biobjective acquisition function. This represents the additable biobjective acquisition subfunction corresponding to the nth spatial group. This represents the mean function corresponding to the nth spatial group. Let t represent the decision variables included in the nth spatial group, t represent the iteration number, and N represent the number of spatial groups. Let represent the covariance function corresponding to the nth spatial group during the t-th iteration. This represents the decision variables included in another spatial grouping in the t-th iteration.

[0035] Furthermore, the confidence function of the upper boundary of the additable structure includes:

[0036] The confidence function of the upper boundary of the additable structure is expressed by the following formula:

[0037]

[0038] in, Let β represent the upper boundary confidence function of the additive structure. t For the pre-set coefficients, This represents the mean function corresponding to the nth spatial group. Let represent the square root of the covariance function corresponding to the nth spatial grouping during the t-th iteration. Let t represent the decision variables included in the nth spatial group, t represent the iteration number, and N represent the number of spatial groups. This represents the decision variables included in another spatial grouping in the t-th iteration.

[0039] Furthermore, the optimization of the candidate solution set includes:

[0040] Substitute the candidate solution set into the multi-objective function to perform function evaluation, and obtain the objective value corresponding to each candidate solution;

[0041] Update the set of known observation points of the addable Gaussian model according to the target value;

[0042] A new Gaussian model is constructed based on the updated set of observations, and the multi-objective function is re-solved based on the new Gaussian model.

[0043] Repeat the above steps until the preset termination condition is met to obtain the approximate Pareto optimal solution set and Pareto front of the multi-objective function.

[0044] On the other hand, embodiments of this specification provide a natural gas production scheduling optimization device, the device comprising:

[0045] The acquisition module is used to acquire natural gas production process data, which includes raw material data, production operation data, product quality data, and storage and transportation data.

[0046] The optimization model construction module is used to construct a natural gas production scheduling optimization model based on the natural gas production process data and production scheduling optimization objectives. The natural gas production scheduling optimization model includes constraints, multi-objective functions, and decision variables.

[0047] The aggregation module is used to obtain several sub-objective functions from the multi-objective function, and to aggregate the several sub-objective functions using the augmented Chebyshev function to obtain the scalar cost function.

[0048] The partitioning module is used to spatially partition the decision variables based on pre-defined prior knowledge using the Bayesian variational inference method, resulting in several spatial groups of decision variables.

[0049] The Gaussian model construction module is used to construct an additive Gaussian model of the scalar value function based on the scalar value function and several decision variable space groupings.

[0050] The function construction module is used to construct an additive structure biobjective acquisition function and an additive structure upper boundary confidence function for each decision variable spatial grouping based on the additive Gaussian model.

[0051] The solution module is used to solve the upper boundary confidence function of the additive structure and the bi-objective acquisition function of the additive structure to obtain candidate solutions for each decision variable space grouping;

[0052] The candidate solution acquisition module is used to aggregate the candidate solutions of each decision variable space group to obtain the candidate solution set of the natural gas production scheduling optimization model;

[0053] The optimization scheme acquisition module is used to optimize the candidate solution set to obtain the optimal natural gas production scheduling scheme.

[0054] In another aspect, embodiments of this specification also provide a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the computer program, when executed by the processor, performs instructions of any of the methods described above.

[0055] In another aspect, embodiments of this specification also provide a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor of a computer device to perform instructions for any of the methods described above.

[0056] In another aspect, embodiments of this specification also provide a computer program product, which, when run by the processor of a computer device, executes instructions for any of the methods described above.

[0057] Using the above technical solution, the natural gas production scheduling optimization method provided in this specification firstly aggregates several sub-objective functions in a multi-objective function using an augmented Chebyshev function to obtain a scalar cost function, thereby avoiding the challenge of learning an additive Gaussian model for each sub-objective. Secondly, it uses Bayesian variational inference to spatially partition decision variables based on pre-defined prior knowledge, reducing the dimensionality of the decision space and alleviating the curse of dimensionality in high-dimensional decision spaces. Then, to achieve parallel function evaluation in high-dimensional decision spaces, it employs additive structure upper boundary confidence and additive structure dual-objective acquisition functions. By optimizing the two additive structure functions, multiple candidate solutions are obtained, thereby achieving parallel function evaluation, improving optimization efficiency, and accelerating convergence speed. Finally, by optimizing the candidate solution set, an approximate Pareto optimal solution set and Pareto front for the natural gas production scheduling optimization problem are obtained, improving optimization accuracy.

[0058] The above description is merely an overview of some embodiments of the technical solutions in this specification. In order to better understand the technical means of some embodiments of this specification and to implement them in accordance with the content of the specification, and to make the above and other objects, features and advantages of the embodiments of this specification more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0059] To more clearly illustrate the technical solutions in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0060] Figure 1 A flowchart illustrating a natural gas production scheduling optimization method in some embodiments of this specification is shown.

[0061] Figure 2 A schematic diagram of the process for spatial partitioning of the decision variables in some embodiments of this specification is shown;

[0062] Figure 3 The flowchart illustrating the construction of the additive Gaussian model of the scalar cost function in some embodiments of this specification is shown.

[0063] Figure 4The flowcharts illustrating the additive structure biobjective acquisition function and additive structure upper boundary confidence function for constructing each decision variable spatial grouping in some embodiments of this specification are shown.

[0064] Figure 5 The flowchart illustrating the optimization of the candidate solution set in some embodiments of this specification is shown;

[0065] Figure 6 This specification shows a schematic diagram of the module structure of a natural gas production scheduling optimization device in some embodiments;

[0066] Figure 7 A schematic diagram of the structure of a computer device is shown in this specification.

[0067] Explanation of symbols in the attached drawings:

[0068] 601. Acquisition Module;

[0069] 602. Optimize the model building module;

[0070] 603. Aggregation Module;

[0071] 604. Divide into modules;

[0072] 605. Gaussian model construction module;

[0073] 606. Function Construction Module;

[0074] 607. Solver Module;

[0075] 608. Candidate Solution Acquisition Module;

[0076] 609. Optimization scheme acquisition module;

[0077] 702. Computer equipment;

[0078] 704, Processor;

[0079] 706. Memory;

[0080] 708. Drive mechanism;

[0081] 710. Input / Output Module;

[0082] 712. Input devices;

[0083] 714. Output devices;

[0084] 716. Presentation equipment;

[0085] 718. Graphical User Interface;

[0086] 720. Network interface;

[0087] 722. Communication link;

[0088] 724. Communication bus. Detailed Implementation

[0089] The technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0090] Natural gas production scheduling mainly involves multiple stages, including natural gas extraction, processing, storage, transportation, and distribution to end users. Faced with the impact of new energy sources and the ensuing fierce competition, natural gas companies must further improve their production efficiency and comprehensively utilize various optimization methods to simultaneously reduce costs and increase efficiency. This is crucial to enhancing their competitiveness in a globalized market characterized by intense competition, declining aggregate demand, accelerated product quality upgrades, and increasingly stringent energy conservation and environmental protection requirements. Therefore, optimizing natural gas scheduling schemes without modifying production equipment has become a key research direction for natural gas companies to improve economic efficiency. Natural gas industry chain production scheduling considers numerous factors, including gas supply, product demand, pipeline equipment status, energy consumption, and product quality. These parameters are interrelated and collectively influence the scheduling results. Furthermore, natural gas production scheduling often has multiple optimization objectives, such as maximizing output, minimizing energy consumption, and ensuring product quality. These objectives may conflict, requiring trade-offs and coordination during the scheduling process. As the dimensionality of the decision variable space for optimization objectives increases, the computational complexity of optimization algorithms typically grows exponentially, making it extremely difficult to find accurate optimization solutions in high-dimensional situations.

[0091] To address the aforementioned issues, this specification provides a method for optimizing natural gas production scheduling. Figure 1This is a flowchart illustrating a natural gas production scheduling optimization method provided in the embodiments of this specification. This specification provides the operational steps of the method described in the embodiments or flowcharts, but based on conventional or non-creative labor, it may include more or fewer operational steps. The order of steps listed in the embodiments is merely one possible execution order among many, and does not represent the only execution order. In actual system or device product execution, the methods shown in the embodiments or accompanying drawings can be executed sequentially or in parallel. Specifically, a high-dimensional multi-objective Bayesian optimization method based on additive Gaussian structures is proposed. By introducing additive Gaussian structures, high-dimensional multi-objective problems are solved. Specifically, the additive Gaussian structure from single-objective Bayesian optimization methods is introduced into multi-objective optimization. Given prior knowledge of decision variable blocks, Bayesian inference is used to derive the final possible subspaces of the decision variable space, thereby reducing the dimensionality of the decision variable space and alleviating the curse of dimensionality problem in high-dimensional decision variable spaces. Furthermore, to achieve parallelized function evaluation in a high-dimensional decision variable space, this embodiment employs an additive structured upper boundary confidence function and an additive structured bi-objective acquisition function. Multiple candidate solutions are obtained by optimizing these two acquisition functions, thereby achieving parallelized function evaluation. Moreover, unlike the bi-objective acquisition function in a low-dimensional decision variable space, the bi-objective acquisition function in this embodiment is an additive Gaussian structure bi-objective acquisition function, which can be derived from multiple acquisition functions in the subspace, thus improving optimization efficiency.

[0092] It should be noted that the terms "first," "second," etc., used in this specification, claims, and the foregoing drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0093] Reference Figure 1 As shown in the embodiments of this specification, a natural gas production scheduling optimization method is provided, the method comprising:

[0094] S101: Acquire natural gas production process data, which includes raw material data, production operation data, product quality data, and storage and transportation data;

[0095] S102: Construct a natural gas production scheduling optimization model based on the natural gas production process data and production scheduling optimization objectives. The natural gas production scheduling optimization model includes constraints, multi-objective functions, and decision variables.

[0096] S103: Obtain several sub-objective functions from the multi-objective function, and aggregate the several sub-objective functions using the augmented Chebyshev function to obtain the scalar cost function;

[0097] S104: Using the Bayesian variational inference method, the decision variables are spatially partitioned based on pre-defined prior knowledge to obtain several spatial groupings of decision variables;

[0098] S105: Based on the scalar cost function and the spatial grouping of several decision variables, construct an additive Gaussian model of the scalar cost function;

[0099] S106: Construct an additive structure biobjective acquisition function and an additive structure upper boundary confidence function for each decision variable spatial grouping based on the additive Gaussian model;

[0100] S107: Solve the upper boundary confidence function of the additive structure and the bi-objective acquisition function of the additive structure to obtain candidate solutions for each decision variable spatial grouping;

[0101] S108: Aggregate the candidate solutions of each decision variable spatial group to obtain the candidate solution set of the natural gas production scheduling optimization model;

[0102] S109: Optimize the candidate solution set to obtain the optimal natural gas production scheduling scheme.

[0103] Using the above technical solution, the natural gas production scheduling optimization method provided in this specification firstly aggregates several sub-objective functions in a multi-objective function using an augmented Chebyshev function to obtain a scalar cost function, thereby avoiding the challenge of learning an additive Gaussian model for each sub-objective. Secondly, it uses Bayesian variational inference to spatially partition decision variables based on pre-defined prior knowledge, reducing the dimensionality of the decision space and alleviating the curse of dimensionality in high-dimensional decision spaces. Then, to achieve parallel function evaluation in high-dimensional decision spaces, it employs additive structure upper boundary confidence and additive structure dual-objective acquisition functions. By optimizing the two additive structure functions, multiple candidate solutions are obtained, thereby achieving parallel function evaluation, improving optimization efficiency, and accelerating convergence speed. Finally, by optimizing the candidate solution set, an approximate Pareto optimal solution set and Pareto front for the natural gas production scheduling optimization problem are obtained, improving optimization accuracy.

[0104] In this embodiment of the specification, step S101 mainly includes raw natural gas information obtained during the natural gas extraction process, such as the geographical location of the gas well, extraction depth, natural gas reserves, and extraction method. This data is typically collected and organized by the natural gas extraction company or its relevant departments. Production operation data refers to data monitored and recorded in real time at each production stage during the natural gas production process, including the operating status, processing capacity, operating parameters (such as temperature, pressure, and flow rate) of each production equipment, energy consumption data (such as electricity, steam, and fuel gas) during production, and equipment failure and maintenance records. Product quality data refers to data used for quality testing and evaluation of natural gas products during natural gas processing and treatment, such as the purity, calorific value, and water content of natural gas. Storage and transportation data includes the transportation methods, volume, and time of raw materials and products, the capacity and inventory of storage tanks, temperature, pressure, pipeline pressure, flow rate, and any abnormalities such as losses and leaks during storage and transportation. When acquiring natural gas production process data, it is necessary to ensure the accuracy and reliability of the data to provide an accurate data foundation for subsequent optimization model construction.

[0105] In the embodiments of this specification, in step S102, considering the physical limitations of the actual production equipment in the natural gas plant, such as the number of transportation pipelines, the safety margin in storage tanks, and the power limitations of pumps, when establishing the natural gas production scheduling optimization model, it is necessary to establish constraints based on the acquired natural gas production process data so that the natural gas production scheduling optimization model can solve for the optimal solution under the set constraints. Specifically, the constraints may include raw material supply constraints, such as the type, quantity, and quality of raw materials needing to meet production requirements; production equipment constraints, such as the processing capacity and operating parameters of each piece of equipment needing to be within safe limits; product quality constraints, such as products needing to meet quality standards and non-conforming products needing to be handled promptly; storage and transportation constraints, such as the transportation and storage of raw materials and products needing to meet safety and environmental protection requirements; and time constraints, such as production cycles and delivery dates.

[0106] In the embodiments of this specification, the production scheduling optimization objectives include maximizing output, minimizing operating costs, minimizing energy consumption, maximizing economic benefits, minimizing environmental pollution, and maximizing production efficiency. Decision variables mainly include raw material selection and proportioning, production equipment operating parameters, running parameters, equipment utilization rate, the quantity of products to be produced by each production equipment in different time periods, the processing sequence of different products or raw materials in the production unit, and production schedule planning. After setting the decision variables and constraints, the set production scheduling optimization objectives are used as the optimization objectives of the natural gas production scheduling optimization model. A multi-objective function is established, including the objective function for maximizing output, the objective function for minimizing operating costs, the objective function for minimizing energy consumption, the objective function for maximizing economic benefits, the objective function for minimizing environmental pollution, and the objective function for maximizing production efficiency. Then, based on the multi-objective function, the values ​​of each decision variable in different time periods are optimized using the multi-objective function to obtain the optimal natural gas production scheduling scheme with the maximum output, minimum operating costs, minimum energy consumption, maximum economic benefits, minimum environmental pollution, and maximum production efficiency as the optimal solution.

[0107] In this embodiment, for an expensive multi-objective optimization problem containing M optimization objectives and D decision variables, the multi-objective function can be expressed as follows:

[0108] minF t (x)={f1(x),f2(x),Λ,f M (x)} T

[0109] subjecttox∈R D

[0110] Where Ft(x) is the mapping from the decision variable space to the objective function space, x is the decision variable, and R0 is the target function space. D Let D be the decision variable space, and f be the dimension of the decision space. M (x) represents the objective function corresponding to the Mth optimization objective, where M is the number of optimization objectives, is the mapping from the decision space to the objective space, and t is the current iteration number. The embodiments in this specification utilize an augmented Chebyshev function to aggregate the values ​​of the M sub-objective functions, thereby transforming the multi-objective optimization problem into a single-objective problem containing multiple different single-objective functions. This effectively avoids the challenge of learning an additive Gaussian model for each single-objective function in the subsequent optimization process.

[0111] In some embodiments of this specification, the scalar cost function of the multi-objective function is constructed based on the sub-objective function and the sub-objective weights pre-set for the sub-objective function, as shown below:

[0112]

[0113] Among them, f λ(x) represents the scalar cost function, λ i f represents the weight of the sub-objective function of the i-th sub-objective function. i (x) represents the sub-objective function corresponding to the i-th optimization objective, M represents the number of sub-objective functions, x is the decision variable, and λ i ·f i (x) represents the nonlinear part. The linear part is the scalar cost function derived from the augmented Chebyshev function. The nonlinear part can make points on the nonconvex region of the Pareto front the minimum of the function, while the linear part ensures that the Pareto optimal set can achieve more improvement than its weakly dominated points.

[0114] In some embodiments of this specification, in step S104, refer to Figure 2 The step of using Bayesian variational inference to spatially partition the decision variables based on pre-defined prior knowledge includes:

[0115] S201: Construct a latent variable array based on the pre-set number of spatial groups, wherein the latent variable parameters in the latent variable array are used to represent the spatial group to which each decision variable belongs.

[0116] In this embodiment, for example, in a multi-objective optimization problem with 100 decision variables, the number of spatial partitioning groups can be set to 10 or 15, etc. To obtain the spatial grouping of the decision variables, a latent variable array is introduced from the perspective of variational inference. As an optional embodiment, the latent variable array is shown below:

[0117] z = (z1, z2, Λ, z D )

[0118] Where z represents the array of hidden variables, z D The latent variable parameters represent the spatial groupings to which the decision variables belong. For example, for 100 decision variables, where the first decision variable belongs to group 8, the second decision variable belongs to group 2, and the 100th decision variable belongs to group 4, the latent variable array can be represented as follows:

[0119] z = (8, 2, Λ, 4)

[0120] In this embodiment, from the perspective of dividing decision variable dimensions, the division of decision variable dimensions is called spatial grouping, which is another representation of decision variables. Among them, the latent variable parameter z in the latent variable array... D Follows a multinomial distribution z D ~Mult(θ), where the parameter θ follows a Dirichlet distribution, i.e., θ ~ Dir(α), where Dir(α) represents the Dirichlet distribution and α is a hyperparameter.

[0121] S202: Using the Bayesian variational inference method, infer the latent variable array based on pre-set prior knowledge to obtain the posterior distribution function of the latent variable array.

[0122] In this embodiment, each subspace in the decision variable space D is A. n ={j:z j =n}, n∈[1,ΛN], the corresponding target subspace is f n (x). By adding prior knowledge to the partitioning of the decision space, the final subspace partitioning of each decision variable in the decision space and its corresponding objective component can be obtained. In this embodiment, as an optional embodiment, the posterior distribution function of the latent variable array obtained through Bayesian variational inference is:

[0123]

[0124] Among them, D t The dataset used to build the Gaussian model is the prior knowledge, and α n Let A be the hyperparameter corresponding to the nth hidden variable array. n Grouping for the nth space, p(D) t |z) represents the likelihood function.

[0125] S203: Use Gibbs sampling to sample the latent variable parameters in the latent variable array of the posterior distribution function, and determine the decision variable spatial grouping of each decision variable based on the sampled values.

[0126] In this embodiment, Gibbs sampling is used to sample the latent variable array z = (z1, z2, Λ, z D Sampling is performed based on the likelihood function p(D) t The posterior distribution value of z with the largest |z) value is used to partition the final spatial grouping of the decision variables. As an optional embodiment, the Gibbs sampling loop samples the latent variable parameters z in the latent variable array using the following formula. j We obtain the posterior distribution value of the decision variable corresponding to the latent variable parameter:

[0127]

[0128] Where, φ n D is the approximate Pareto front corresponding to the nth spatial grouping. t Let α be prior knowledge, and A be... n Grouping for the nth space, p(D) t |z) represents the likelihood function. Then, based on the posterior distribution value of the latent variable array that maximizes the likelihood function value in the posterior distribution function, the decision variables corresponding to the latent variable parameters are spatially grouped.

[0129] In some embodiments of this specification, in step S105, refer to Figure 3 The step of constructing an additive Gaussian model of the scalar cost function based on the scalar cost function and the spatial grouping of several decision variables includes:

[0130] S301: Take the decision variables in each decision variable space grouping as variables in the scalar value function, and construct several sub-scalar value functions;

[0131] S302: Construct a Gaussian model of the cost function for each subscalar;

[0132] S303: Obtain the additive Gaussian model of the scalar arithmetic value function based on the Gaussian model of each sub-scalar arithmetic value function.

[0133] In this embodiment, the Gaussian surrogate model has the additive property, meaning that the mean and covariance functions of the Gaussian model are obtained by a weighted sum of the mean and covariance functions of the surrogate models in the decision subspace. As an optional embodiment, each scalar cost component in the scalar cost function corresponding to the objective function... All can be approximated by the Gaussian model. Because each spatial group A... n Since the variables in n∈[1,Λ,N] are independent, according to the properties of the Gaussian distribution, the covariance between spatial groups is 0. Therefore, the scalar cost function... The mean function μ(x) and covariance function κ(x) can be derived from the mean function μ of the component Gaussian model corresponding to each scalar cost component. n (x) and covariance function κ n (x) is obtained by weighted summation. Then, based on the spatial grouping obtained from the partitioning of the decision variable space, the posterior distribution p(z) of the variable grouping can be derived. Thus, based on the posterior distribution value, the spatial partitioning of the decision variable in the current algorithm iteration can be performed, and the scalar cost function f can be obtained. λ The prior Gaussian distribution of (x) can be used to deduce that the predicted value of the next scalar cost also follows a Gaussian distribution, i.e.:

[0134] P(y t+1 |D 1:t ,x t+1 )=N(μ t (x t+1 ),κ t (x t+1 ))

[0135] In this embodiment, based on Bayesian variational inference, the potential additive Gaussian structure of the decision variables is learned. As an optional embodiment, the additive Gaussian model is represented as follows:

[0136]

[0137] Among them, f λ ′(x) is an additive Gaussian model, N is the number of spatial partitions, and A i The decision variables included in the i-th spatial group. Let it be the subscalar value function.

[0138] In some embodiments of this specification, in step S106, refer to Figure 4 The step of constructing an additive structure biobjective acquisition function and an additive structure upper boundary confidence function for each decision variable spatial grouping based on the additive Gaussian model includes:

[0139] S401: Construct a bi-objective acquisition function and an upper boundary confidence function for spatial grouping of each decision variable;

[0140] S402: Transform the bi-objective acquisition function and the upper boundary confidence function according to the additive Gaussian model to obtain the additive bi-objective acquisition function and the additive upper boundary confidence function for each spatial group.

[0141] As an optional embodiment, the dual-objective acquisition function is as follows:

[0142] mop t (x)=(μ t (x)-σ t (x)) T

[0143] Among them, mop t (x) is the dual-objective acquisition function, σ t (x) is the covariance function κ t The square root of (x).

[0144] In high-dimensional decision variable spaces, the optimization efficiency of bi-objective acquisition functions is low due to the curse of dimensionality, resulting in a significant reduction in the quality of the final solution obtained through these functions. To improve the optimization efficiency of bi-objective acquisition functions while achieving parallel function evaluation, this embodiment uses latent variables z = (z1, z2, Λ, z...) D The partitioning of the learned decision variable space transforms the bi-objective acquisition function into a bi-objective acquisition function with an additive structure, i.e., a multi-point sampling function. The additive bi-objective acquisition function is shown below:

[0145]

[0146] in, This represents an additable biobjective acquisition function. This represents the additable biobjective acquisition subfunction corresponding to the nth spatial group. This represents the mean function corresponding to the nth spatial group. Let t represent the decision variables included in the nth spatial group, t represent the iteration number, and N represent the number of spatial groups. Let represent the covariance function corresponding to the nth spatial group during the t-th iteration. This represents the decision variables included in another spatial grouping in the t-th iteration.

[0147] In this embodiment, the upper boundary confidence function (UCB function) is as follows:

[0148]

[0149] Based on the established additive Gaussian model, the UCB function is transformed to obtain the upper boundary confidence function of the additive structure as follows:

[0150]

[0151] in, Let β represent the upper boundary confidence function of the additive structure. t For the pre-set coefficients, This represents the mean function corresponding to the nth spatial group. Let represent the square root of the covariance function corresponding to the nth spatial grouping during the t-th iteration. Let t represent the decision variables included in the nth spatial group, t represent the iteration number, and N represent the number of spatial groups. This represents the decision variables included in another spatial grouping in the t-th iteration.

[0152] In some embodiments of this specification, in step S107, a non-dominated sorting genetic algorithm is used to solve the upper boundary confidence function and the additive bi-objective acquisition function of the additive structure to obtain candidate solutions for each decision variable space group. The candidate solutions for each decision variable space group are then aggregated to obtain a candidate solution set for the natural gas production scheduling optimization model. Then, based on a selection strategy, B-1 candidate solutions are selected from the candidate solution set for evaluation to obtain the final candidate solution set, where B is the batch size. Thus, this specification transforms the optimization of the acquisition function in the high-dimensional decision space into the simultaneous optimization of multiple sub-acquisition functions in multiple low-dimensional subspaces. This reduces computational complexity, alleviates the curse of dimensionality, improves the optimization efficiency of the acquisition function, and enhances the performance of the final solution in balancing convergence and diversity.

[0153] In some embodiments of this specification, in step S109, refer to Figure 5 The optimization of the candidate solution set includes:

[0154] S501: Substitute the candidate solution set into the multi-objective function to perform function evaluation, and obtain the objective value corresponding to each candidate solution;

[0155] S502: Update the set of known observation points of the addable Gaussian model according to the target value;

[0156] S503: Construct a new additive Gaussian model based on the updated set of observation points, and re-solve the multi-objective function based on the new additive Gaussian model;

[0157] S504: Repeat the above steps until the preset termination condition is met to obtain the approximate Pareto optimal solution set and Pareto front of the multi-objective function.

[0158] This can be understood as follows: In the embodiments of this specification, after obtaining B candidate solutions by solving the upper boundary confidence function of the addable structure and the dual-objective acquisition function of the addable structure, the B candidate solutions are substituted into the multi-objective function for true function evaluation to obtain their corresponding true objective values. Then, a uniformly distributed weight vector, different from the previous iteration, is selected for objective function aggregation to obtain the corresponding scalar cost value. At the same time, the known observation points for establishing the Gaussian model are updated, that is, the B new candidate solutions are added to D. 1:t D was obtained from 1:t+1 This method is used to construct a new additive Gaussian model. Through repeated iterations of function evaluation and objective function aggregation, decision variable space partitioning learning, additive Gaussian model learning, and candidate solution recommendation, the algorithm terminates when the termination condition is met. Finally, an approximate Pareto optimal solution set and Pareto front of the multi-objective function are obtained, and the approximate Pareto optimal solution set is used as the optimal natural gas production scheduling scheme.

[0159] This specification's embodiments achieve the partitioning of the decision variable space by grouping decision variables based on prior knowledge, thereby reducing the dimensionality of the decision variable space. Furthermore, by utilizing an additive bi-objective acquisition function and an additive upper boundary confidence function, parallel function evaluation in the high-dimensional decision variable space can be achieved, improving the balance between convergence and diversity. Moreover, in the subspace of the decision variable space, an additive Gaussian model with scalar cost for the original optimization problem is obtained through weighted summation, reducing the computational complexity of the Gaussian model. In addition, the additive bi-objective acquisition function and the additive upper boundary confidence function are established in the decision subspace, obtaining multiple candidate solutions through group-by-group optimization. This not only improves the optimization efficiency of the acquisition function but also enables batch function evaluation, promoting convergence; simultaneously, it better balances the relationship between exploitation and exploration in Bayesian optimization. Thus, to a certain extent, it alleviates the curse of dimensionality problem, improves the optimization efficiency of high-dimensional problems, and accelerates the convergence speed.

[0160] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the acquisition, storage, use, and processing of data in the technical solutions described in the embodiments of this application all comply with relevant regulations.

[0161] Based on the above-described method for optimizing natural gas production scheduling, this specification also provides a corresponding device for optimizing natural gas production scheduling. The device may include a system (including a distributed system), software (application), module, component, server, client, etc., using the method described in this specification, combined with necessary hardware implementation. Based on the same innovative concept, the devices in one or more embodiments provided in this specification are as described in the following embodiments. Since the implementation schemes and methods for solving the problem are similar, the implementation of specific devices in this specification can refer to the implementation of the aforementioned method, and repeated details will not be repeated. As used below, the terms "unit" or "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0162] Specifically, Figure 6 This is a schematic diagram of the module structure of one embodiment of a natural gas production scheduling optimization device provided in this specification. (Refer to...) Figure 6 As shown in the embodiments of this specification, a natural gas production scheduling optimization device includes:

[0163] The acquisition module 601 is used to acquire natural gas production process data, which includes raw material data, production operation data, product quality data, and storage and transportation data.

[0164] The optimization model construction module 602 is used to construct a natural gas production scheduling optimization model based on the natural gas production process data and production scheduling optimization objectives. The natural gas production scheduling optimization model includes constraints, multi-objective functions, and decision variables.

[0165] The aggregation module 603 is used to obtain several sub-objective functions from the multi-objective function, and to aggregate the several sub-objective functions using the augmented Chebyshev function to obtain the scalar cost function.

[0166] The partitioning module 604 is used to spatially partition the decision variables based on pre-set prior knowledge using the Bayesian variational inference method, thereby obtaining several spatial groups of decision variables.

[0167] Gaussian model construction module 605 is used to construct an additive Gaussian model of the scalar value function based on the scalar value function and several decision variable space groupings.

[0168] The function construction module 606 is used to construct an additive structure biobjective acquisition function and an additive structure upper boundary confidence function for each decision variable spatial grouping based on the additive Gaussian model.

[0169] The solver module 607 is used to solve the upper boundary confidence function of the addable structure and the bi-objective acquisition function of the addable structure to obtain candidate solutions for each decision variable space grouping;

[0170] The candidate solution acquisition module 608 is used to aggregate the candidate solutions of each decision variable space group to obtain the candidate solution set of the natural gas production scheduling optimization model;

[0171] The optimization scheme acquisition module 609 is used to optimize the candidate solution set to obtain the optimal natural gas production scheduling scheme.

[0172] The beneficial effects obtained by the apparatus provided in the embodiments of this specification are consistent with the beneficial effects obtained by the methods described above, and will not be repeated here.

[0173] Reference Figure 7 As shown, based on the above-described method for optimizing natural gas production scheduling, one embodiment of this specification also provides a computer device 702, wherein the above method operates on the computer device 702. The computer device 702 may include one or more processors 704, such as one or more central processing units (CPUs), each of which can implement one or more hardware threads. The computer device 702 may also include any memory 706 for storing any kind of information such as code, settings, data, etc. Non-limitingly, for example, the memory 706 may include any type of RAM, any type of ROM, flash memory, hard disk, optical disk, etc. More generally, any memory can use any technology to store information. Further, any memory can provide volatile or non-volatile retention of information. Further, any memory can represent a fixed or removable component of the computer device 702. In one case, when the processor 704 executes associated instructions stored in any memory or combination of memories, the computer device 702 can perform any operation of the associated instructions. The computer device 702 also includes one or more drive mechanisms 708 for interacting with any memory, such as a hard disk drive mechanism, an optical disk drive mechanism, etc.

[0174] Computer device 702 may also include an input / output module 710 (I / O) for receiving various inputs (via input device 712) and providing various outputs (via output device 714). A specific output mechanism may include a presentation device 716 and an associated graphical user interface (GUI) 718. In other embodiments, the input / output module 710 (I / O), input device 712, and output device 714 may be omitted, and the device may function solely as a computer device within a network. Computer device 702 may also include one or more network interfaces 720 for exchanging data with other devices via one or more communication links 722. One or more communication buses 724 couple the components described above together.

[0175] Communication link 722 can be implemented in any way, such as via a local area network, a wide area network (e.g., the Internet), a point-to-point connection, or any combination thereof. Communication link 722 may include any combination of hardwired links, wireless links, routers, gateway functions, name servers, etc., governed by any protocol or combination of protocols.

[0176] Corresponding to, for example Figures 1 to 5 In addition to the method shown, embodiments of this specification also provide a computer-readable storage medium storing a computer program that, when executed by a processor, performs the steps of the above-described method.

[0177] This specification also provides computer-readable instructions, wherein when a processor executes the instructions, the program therein causes the processor to perform the following... Figures 1 to 5 The method shown.

[0178] This specification also provides a computer program product, including at least one instruction or at least one program segment, wherein the at least one instruction or the at least one program segment is loaded and executed by a processor to achieve the following: Figures 1 to 5 The method shown.

[0179] It should be understood that in the various embodiments of this specification, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this specification.

[0180] It should also be understood that, in the embodiments of this specification, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this specification generally indicates that the preceding and following related objects have an "or" relationship.

[0181] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this specification can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this specification.

[0182] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0183] In the several embodiments provided in this specification, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the couplings or direct couplings or communication connections shown or discussed may be indirect couplings or communication connections through some interfaces, devices, or units, or they may be electrical, mechanical, or other forms of connection.

[0184] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments described in this specification, depending on actual needs.

[0185] Furthermore, the functional units in the various embodiments of this specification can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0186] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this specification, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this specification. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0187] This specification uses specific embodiments to illustrate the principles and implementation methods of this specification. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this specification. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this specification. Therefore, the content of this specification should not be construed as a limitation of this specification.

Claims

1. A method for optimizing natural gas production scheduling, characterized in that, The method includes: Acquire natural gas production process data, which includes raw material data, production operation data, product quality data, and storage and transportation data; A natural gas production scheduling optimization model is constructed based on the natural gas production process data and production scheduling optimization objectives. The natural gas production scheduling optimization model includes constraints, multi-objective functions, and decision variables. Several sub-objective functions are obtained from the multi-objective function, and the several sub-objective functions are aggregated using the augmented Chebyshev function to obtain the scalar cost function; The decision variables are spatially partitioned based on pre-defined prior knowledge using Bayesian variational inference methods, resulting in several spatial groupings of decision variables. Based on the scalar cost function and the spatial grouping of several decision variables, an additive Gaussian model of the scalar cost function is constructed. Based on the additive Gaussian model, construct the additive structure biobjective acquisition function and the additive structure upper boundary confidence function for each decision variable spatial grouping; The upper boundary confidence function and the biobjective acquisition function of the additive structure are solved to obtain candidate solutions for each decision variable spatial grouping; The candidate solutions of each decision variable space group are aggregated to obtain the candidate solution set of the natural gas production scheduling optimization model; The candidate solution set is optimized to obtain the optimal natural gas production scheduling scheme.

2. The method according to claim 1, characterized in that, The scalar cost function includes: The scalar cost function can be expressed using the following formula: Among them, f λ (x) represents the scalar cost function, λ i f represents the weight of the sub-objective function of the i-th sub-objective function. i (x) represents the sub-objective function corresponding to the i-th optimization objective, M represents the number of sub-objective functions, and x is the decision variable.

3. The method according to claim 1, characterized in that, The step of using Bayesian variational inference to spatially partition the decision variables based on pre-defined prior knowledge includes: Based on a pre-set number of spatial groups, a latent variable array is constructed, wherein the latent variable parameters in the latent variable array are used to represent the spatial group to which each decision variable belongs; Using Bayesian variational inference methods, the latent variable array is inferred based on pre-defined prior knowledge to obtain the posterior distribution function of the latent variable array; Gibbs sampling is used to sample the latent variable parameters in the latent variable array of the posterior distribution function, and the decision variable spatial grouping of each decision variable is determined based on the sampled values.

4. The method according to claim 1, characterized in that, The step of constructing an additive Gaussian model of the scalar value function based on the scalar value function and the spatial grouping of several decision variables includes: By taking the decision variables in each decision variable space grouping as variables in the scalar value function, several sub-scalar value functions are constructed. Construct a Gaussian model of the cost function of each subscalar; The additive Gaussian model of the scalar cost function is obtained based on the Gaussian model of each subscalar cost function.

5. The method according to claim 1, characterized in that, The step of constructing the additive structure biobjective acquisition function and the additive structure upper boundary confidence function for each decision variable spatial grouping based on the additive Gaussian model includes: Construct a bi-objective acquisition function and an upper boundary confidence function for spatial grouping of each decision variable; The additive Gaussian model is used to transform the bi-objective acquisition function and the upper boundary confidence function to obtain the additive bi-objective acquisition function and the additive upper boundary confidence function for each spatial group.

6. The method according to claim 1, characterized in that, The additable biobjective acquisition function includes: The additable structure biobjective acquisition function is expressed by the following formula: in, This represents an additable biobjective acquisition function. This represents the additable biobjective acquisition subfunction corresponding to the nth spatial group. This represents the mean function corresponding to the nth spatial group. Let t represent the decision variables included in the nth spatial group, t represent the iteration number, and N represent the number of spatial groups. Let represent the covariance function corresponding to the nth spatial group during the t-th iteration. This represents the decision variables included in another spatial grouping in the t-th iteration.

7. The method according to claim 1, characterized in that, The upper boundary confidence function of the additable structure includes: The confidence function of the upper boundary of the additable structure is expressed by the following formula: in, Let β represent the upper boundary confidence function of the additive structure. t For the pre-set coefficients, This represents the mean function corresponding to the nth spatial group. Let represent the square root of the covariance function corresponding to the nth spatial grouping during the t-th iteration. Let t represent the decision variables included in the nth spatial group, t represent the iteration number, and N represent the number of spatial groups. This represents the decision variables included in another spatial grouping in the t-th iteration.

8. The method according to claim 1, characterized in that, The optimization of the candidate solution set includes: Substitute the candidate solution set into the multi-objective function to perform function evaluation, and obtain the objective value corresponding to each candidate solution; Update the set of known observation points of the addable Gaussian model according to the target value; A new Gaussian model is constructed based on the updated set of observations, and the multi-objective function is re-solved based on the new Gaussian model. Repeat the above steps until the preset termination condition is met to obtain the approximate Pareto optimal solution set and Pareto front of the multi-objective function.

9. A natural gas production scheduling optimization device, characterized in that, The device includes: The acquisition module is used to acquire natural gas production process data, which includes raw material data, production operation data, product quality data, and storage and transportation data. The optimization model construction module is used to construct a natural gas production scheduling optimization model based on the natural gas production process data and production scheduling optimization objectives. The natural gas production scheduling optimization model includes constraints, multi-objective functions, and decision variables. An aggregation module is used to obtain several sub-objective functions from the multi-objective function, and to aggregate the several sub-objective functions using an augmented Chebyshev function to obtain a scalar cost function. The partitioning module is used to spatially partition the decision variables based on pre-defined prior knowledge using the Bayesian variational inference method, resulting in several spatial groups of decision variables. The Gaussian model construction module is used to construct an additive Gaussian model of the scalar value function based on the scalar value function and several decision variable space groupings. The function construction module is used to construct an additive structure biobjective acquisition function and an additive structure upper boundary confidence function for each decision variable spatial grouping based on the additive Gaussian model. The solution module is used to solve the upper boundary confidence function of the additive structure and the bi-objective acquisition function of the additive structure to obtain candidate solutions for each decision variable space grouping; The candidate solution acquisition module is used to aggregate the candidate solutions of each decision variable space group to obtain the candidate solution set of the natural gas production scheduling optimization model; The optimization scheme acquisition module is used to optimize the candidate solution set to obtain the optimal natural gas production scheduling scheme.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method as described in any one of claims 1 to 8.

12. A computer program product, characterized in that, It includes at least one instruction or at least one program segment, said at least one instruction or said at least one program segment being loaded and executed by a processor to implement the method as claimed in any one of claims 1 to 8.

Images