Target enterprise carbon emission reduction scheme optimization method and device, computer equipment and readable storage medium

Abstract

The invention relates to the technical field of carbon emission reduction, and provides a target enterprise carbon emission reduction scheme optimization method and device, computer equipment and a readable storage medium. The method comprises the steps of obtaining a carbon emission reduction optimization model; obtaining a current carbon emission reduction scheme according to the initial parameters of the artificial travel mouse optimization algorithm and the carbon emission reduction optimization model; adaptively adjusting the Brown disturbance intensity according to the current optimization progress; if the current stage is in a global search stage, executing a Brownian motion search mode according to the Brownian disturbance intensity, and updating the current carbon emission reduction scheme to obtain a first carbon emission reduction scheme; if the current stage is in a switching stage of switching from the global search stage to the local development stage, selecting a Brownian motion search mode or a heavy tail distribution random walk search mode according to a random probability, updating the first carbon emission reduction scheme, and obtaining a second carbon emission reduction scheme, and an optimal carbon emission reduction scheme meeting an optimization target and an operation constraint condition is obtained. By adopting the method, the accuracy of global search and local development can be enhanced.

Description

Technical Field

[0001] The present application relates to the field of carbon emission reduction technology, and in particular to a method, apparatus, computer equipment, computer-readable storage medium, and computer program product for optimizing a carbon emission reduction plan for a target enterprise. Background Art

[0002] Key industries such as steel, cement, chemicals, ceramics, and textiles are gradually transitioning to green and low-carbon development. Traditional high-carbon development models are unsustainable, and optimizing carbon reduction strategies is crucial for businesses to achieve sustainable development through this green and low-carbon transformation.

[0003] To optimize carbon reduction solutions, swarm intelligence algorithms such as particle swarm optimization (PSO), differential evolution (DE), and genetic algorithms (GA) have been introduced into enterprise carbon emission and load optimization scenarios, achieving promising results at medium scales. However, when faced with complex boundary conditions, severe load fluctuations, and long operating cycles, traditional swarm intelligence algorithms still face challenges such as premature search, local optimality, and insufficient adaptability. These algorithms struggle to meet the "multi-timescale, dynamic, and global optimization" requirements for optimizing carbon emission pathways for target enterprises in key industries. Summary of the Invention

[0004] Based on this, it is necessary to provide a method, device, computer equipment, computer-readable storage medium and computer program product for optimizing the carbon emission reduction plan of a target enterprise in response to the above technical problems.

[0005] In the first aspect, this application provides a method for optimizing a target enterprise's carbon emission reduction plan, including:

[0006] Based on the unit carbon emission factor of each target enterprise and the power load at each moment, a carbon emission reduction optimization model is obtained;

[0007] Randomly generate an initial carbon emission reduction plan that meets the operation constraints based on the initial parameters of the artificial lemming optimization algorithm and the carbon emission reduction optimization model, and obtain a current carbon emission reduction plan;

[0008] Adaptively adjust the Brownian perturbation intensity according to the current optimization progress;

[0009] If the current stage is in the global search stage, executing the Brownian motion search mode according to the Brownian perturbation intensity, updating the current carbon emission reduction plan, and obtaining a first carbon emission reduction plan;

[0010] If the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, the first carbon emission reduction plan is updated, and the second carbon emission reduction plan is obtained; the switching stage corresponds to switching from the global search stage to the local development stage;

[0011] If the current stage is in the local development stage, then according to the spiral update strategy, the second carbon emission reduction plan is updated to obtain the third carbon emission reduction plan;

[0012] If the current optimization progress reaches the maximum number of function evaluations or the convergence threshold, an optimal carbon emission reduction plan that meets the optimization objectives and operation constraints is obtained according to the third carbon emission reduction plan.

[0013] In one embodiment, obtaining a carbon emission reduction optimization model based on the unit carbon emission factor of each target enterprise and the power load at each moment includes:

[0014] According to the grid topology, the unit carbon factor of each target enterprise and the output power at each time, the unit carbon emission factor of each target enterprise is obtained;

[0015] For each target enterprise, obtain the carbon emission reduction assessment value of the target enterprise based on the unit carbon emission factor of the target enterprise and the power load at each time;

[0016] The carbon emission reduction assessment values ​​of each target enterprise are added together to obtain a carbon emission reduction optimization model.

[0017] In one embodiment, the adaptively adjusting the Brownian perturbation intensity according to the current optimization progress includes:

[0018] According to the current optimization progress, get the current function evaluation times;

[0019] Obtaining a nonlinear convergence factor according to the current function evaluation number and the maximum function evaluation number;

[0020] The Brownian perturbation intensity is adaptively adjusted according to the nonlinear convergence factor.

[0021] In one embodiment, the method further comprises:

[0022] According to the current optimization progress, get the current function evaluation times;

[0023] Obtaining a nonlinear convergence factor according to the current function evaluation number and the maximum function evaluation number;

[0024] When the nonlinear convergence factor is less than a first preset value, determining that the current stage is a global search stage;

[0025] When the nonlinear convergence factor is greater than a first preset value and less than a second preset value, determining that the current stage is a switching stage corresponding to the global search stage to the local development stage;

[0026] When the nonlinear convergence factor is greater than a second preset value, the current stage is determined to be a local development stage.

[0027] In one embodiment, if the current stage is in the global search stage, a Brownian motion search mode is executed according to the Brownian perturbation intensity to update the current carbon emission reduction plan to obtain a first carbon emission reduction plan, including:

[0028] If the current stage is in the global search stage, a Brownian motion search mode is executed according to the Brownian perturbation intensity, and the current carbon emission reduction plan is iteratively updated multiple times in the first stage to obtain a first carbon emission reduction plan in the first stage;

[0029] If the current optimization progress converges and stagnates or falls into a local area, the first carbon emission reduction plan of the first stage is iteratively updated multiple times in the second stage according to the heavy-tailed distribution random walk search mode to obtain the first carbon emission reduction plan of the second stage;

[0030] If the current optimization progress does not converge and stagnates or falls into a local area, the Brownian motion search mode is executed according to the Brownian perturbation intensity, and the first carbon emission reduction plan of the second stage is iteratively updated for multiple stages to obtain the first carbon emission reduction plan of the third stage;

[0031] According to the first carbon emission reduction plan of the third stage, a first carbon emission reduction plan is obtained.

[0032] In one embodiment, if the current stage is in the switching stage, selecting the Brownian motion search mode or the heavy-tailed distribution random walk search mode according to random probability, updating the first carbon emission reduction plan, and obtaining the second carbon emission reduction plan include:

[0033] If the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, and the first carbon emission reduction plan is iteratively updated multiple times in the first stage to obtain the second carbon emission reduction plan in the first stage;

[0034] If the current optimization progress converges and stagnates or falls into a local area, the second carbon emission reduction plan of the first stage is updated multiple times in the second stage according to the heavy-tailed distribution random walk search mode to obtain the second carbon emission reduction plan of the second stage;

[0035] If the current optimization progress does not converge, stagnates, or falls into a local area, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, and the second carbon emission reduction plan of the second stage is updated multiple times in the third stage to obtain the second carbon emission reduction plan of the third stage;

[0036] According to the second carbon emission reduction plan of the third stage, a second carbon emission reduction plan is obtained.

[0037] In one embodiment, if the current stage is the local development stage, then according to the spiral update strategy, the second carbon emission reduction plan is updated to obtain a third carbon emission reduction plan, including:

[0038] If the current stage is in the local development stage, according to the spiral update strategy, the second carbon emission reduction plan is iteratively updated multiple times in the first stage to obtain the third carbon emission reduction plan in the first stage;

[0039] If the current optimization progress converges and stagnates or falls into a local area, the third carbon emission reduction plan of the first stage is updated multiple times in the second stage according to the heavy-tailed distribution random walk search mode to obtain the third carbon emission reduction plan of the second stage;

[0040] If the current optimization progress does not converge and stagnates or falls into a local area, the third carbon emission reduction plan of the second stage is updated multiple times in the third stage according to the spiral update strategy to obtain the third carbon emission reduction plan of the third stage;

[0041] According to the third carbon emission reduction plan of the third stage, a third carbon emission reduction plan is obtained.

[0042] Secondly, this application also provides a device for optimizing a target enterprise's carbon emission reduction plan, comprising:

[0043] A carbon emission reduction optimization model acquisition module is used to obtain a carbon emission reduction optimization model based on the unit carbon emission factor of each target enterprise and the power load at each moment;

[0044] A current carbon emission reduction plan acquisition module is used to randomly generate an initial carbon emission reduction plan that meets the operation constraints based on the initial parameters of the artificial lemming optimization algorithm and the carbon emission reduction optimization model, and obtain the current carbon emission reduction plan;

[0045] Brownian perturbation intensity adjustment module, used to adaptively adjust the Brownian perturbation intensity according to the current optimization progress;

[0046] A current carbon emission reduction scheme update module is configured to, if the current stage is in the global search stage, execute a Brownian motion search mode according to the Brownian perturbation intensity to update the current carbon emission reduction scheme to obtain a first carbon emission reduction scheme; if the current stage is in the switching stage, select a Brownian motion search mode or a heavy-tailed distribution random walk search mode according to random probability to update the first carbon emission reduction scheme to obtain a second carbon emission reduction scheme; the switching stage corresponds to switching from the global search stage to the local development stage; if the current stage is in the local development stage, update the second carbon emission reduction scheme according to the spiral update strategy to obtain a third carbon emission reduction scheme;

[0047] The optimal carbon emission reduction scheme acquisition module is used to obtain the optimal carbon emission reduction scheme that meets the optimization objectives and operation constraints according to the third carbon emission reduction scheme if the current optimization progress reaches the maximum number of function evaluations or the convergence threshold.

[0048] In a third aspect, the present application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program and the processor executes the above method.

[0049] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program is used by a processor to execute the above method.

[0050] In a fifth aspect, the present application further provides a computer program product, wherein the computer program product includes a computer program, and the computer program is executed by a processor to execute the above method.

[0051] The above-mentioned target enterprise carbon emission reduction plan optimization method, device, computer equipment, computer-readable storage medium and computer program product obtain a carbon emission reduction optimization model based on the unit carbon emission factor of each target enterprise and the power load at each moment; randomly generate an initial carbon emission reduction plan that meets the operating constraints based on the initial parameters of the artificial lemming optimization algorithm and the carbon emission reduction optimization model to obtain the current carbon emission reduction plan; adaptively adjust the Brownian perturbation intensity according to the current optimization progress; if the current stage is in the global search stage, execute the Brownian motion search mode according to the Brownian perturbation intensity, update the current carbon emission reduction plan, and obtain a first carbon emission reduction plan; if the current stage is in the switching stage, select the Brownian motion search mode or the heavy-tailed distribution random walk search mode according to random probability, update the first carbon emission reduction plan, and obtain a second carbon emission reduction plan; the switching stage corresponds to switching from the global search stage to the local development stage; if the current stage is in the local development stage, update the second carbon emission reduction plan according to the spiral update strategy to obtain a third carbon emission reduction plan; if the current optimization progress reaches the maximum number of function evaluations or the convergence threshold, obtain the optimal carbon emission reduction plan that meets the optimization objectives and operating constraints based on the third carbon emission reduction plan. This application adaptively adjusts the Brownian perturbation intensity according to the current optimization progress; if the current stage is in the global search stage, the Brownian motion search mode is executed according to the Brownian perturbation intensity, the current carbon emission reduction plan is updated, and the first carbon emission reduction plan is obtained; if the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, the first carbon emission reduction plan is updated, and the second carbon emission reduction plan is obtained, and then the optimal carbon emission reduction plan that meets the optimization objectives and operating constraints is obtained, which significantly enhances the accuracy of global search and local development, and is suitable for carbon emission reduction optimization scheduling scenarios under high-dimensional complex and multi-constraint conditions, and can meet the optimization requirements of "multi-time scale, dynamic adjustment, and global optimization" for the carbon emission paths of target enterprises in key industries. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments of the present application or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying any creative work.

[0053] Figure 1 This is an application environment diagram of a target enterprise carbon emission reduction plan optimization method in one embodiment;

[0054] Figure 2 A flowchart of a method for optimizing a target enterprise's carbon emission reduction plan in one embodiment;

[0055] Figure 3A flow chart of a method for optimizing a target enterprise's carbon emission reduction plan in another embodiment;

[0056] Figure 4 A flowchart of an artificial lemming optimization algorithm in one embodiment;

[0057] FIG5 (a) is a graph showing the weekly carbon reduction potential assessment results of a target enterprise in a key industry based on the artificial lemming optimization algorithm in one embodiment;

[0058] FIG5( b ) is a graph showing the monthly carbon reduction potential assessment results of a target enterprise in a key industry based on the artificial lemming optimization algorithm in one embodiment;

[0059] FIG5( c ) is a graph showing the quarterly carbon reduction potential assessment results of a target enterprise in a key industry based on the artificial lemming optimization algorithm in one embodiment;

[0060] Figure 6 This is a structural block diagram of a device for optimizing a target enterprise's carbon emission reduction plan in one embodiment;

[0061] Figure 7 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0063] It should be noted that the terms "first", "second", etc. used in this application may be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "including" and "having" used in this application and any variations thereof are intended to cover non-exclusive inclusions. The term "plurality" used in this application refers to two or more. The term "and / or" used in this application refers to one of the solutions or any combination of multiple solutions.

[0064] The target enterprise carbon emission reduction program optimization method provided in the embodiment of the present application can be applied to Figure 1In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store data that server 104 needs to process. The data storage system can be integrated with server 104, or placed on a cloud or other network server. Terminal 102 can obtain a carbon reduction optimization model based on the unit carbon emission factor of each target enterprise and the power load at each time. Based on the initial parameters of the artificial lemming optimization algorithm and the carbon reduction optimization model, it randomly generates an initial carbon reduction plan that meets the operational constraints, obtains the current carbon reduction plan, and optimizes it. If the current optimization progress reaches the maximum number of function evaluations or the convergence threshold, it obtains the optimal carbon reduction plan that meets the optimization objectives and operational constraints based on the third carbon reduction plan. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, drones, low-altitude aircraft, IoT devices, and portable wearable devices. Server 104 can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server providing cloud computing services.

[0065] In an exemplary embodiment, Figure 2 As shown in the figure, a method for optimizing carbon emission reduction schemes for target enterprises is provided. Figure 1 The terminal 102 in the example is used as an example to illustrate the process, including the following steps S201 to S207.

[0066] Step S201 : obtaining a carbon emission reduction optimization model based on the unit carbon emission factor of each target enterprise and the power load at each moment.

[0067] Based on the grid topology, the unit carbon factor of each target enterprise, and the output power at each time, the unit carbon emission factor of each target enterprise is obtained. Based on the unit carbon emission factor of each target enterprise and the power load at each time, the carbon emission reduction assessment value of each target enterprise can be obtained, and then the carbon emission reduction optimization model can be obtained.

[0068] This carbon emission reduction optimization model is suitable for carbon emission reduction optimization scheduling scenarios under high-dimensional complex and multi-constraint conditions, and can meet the optimization requirements of "multi-time scale, dynamic adjustment, and global optimization" for the carbon emission paths of target enterprises in key industries.

[0069] Step S202 : randomly generating an initial carbon emission reduction plan that meets the operation constraints based on the initial parameters of the artificial lemming optimization algorithm and the carbon emission reduction optimization model, and obtaining the current carbon emission reduction plan.

[0070] Parameters of an Artificial Lemming Algorithm (ALA) can be initialized to obtain initial parameters of the artificial lemming algorithm.

[0071] The parameters of the artificial lemming optimization algorithm include population size N, maximum function evaluation times F MEs , Brownian perturbation coefficient, heavy-tailed distribution random flight jump factor, nonlinear convergence factor (adaptive convergence factor).

[0072] Operational constraints include at least one of: total energy consumption constraints, equipment operation constraints, emission boundary constraints, and operational stability constraints. The total energy consumption constraint requires that the total electricity consumption of each target enterprise must not exceed the production plan budget, and that total electricity consumption remains unchanged before and after optimization. The equipment operation constraint requires that the start / stop characteristics, minimum operating time, and maximum load limits of each type of equipment be met. The emission boundary constraint requires that each target enterprise must not exceed the set carbon emission cap within each time period. The operational stability constraint requires that the task scheduling of the carbon reduction plan be temporally continuous and executable.

[0073] Based on the initial parameters of the artificial lemming optimization algorithm and the carbon emission reduction optimization model, an initial solution population that meets the operating constraints can be randomly generated, the fitness can be calculated, and the current global optimal individual can be selected. That is, an initial carbon emission reduction plan that meets the operating constraints can be randomly generated to obtain the current carbon emission reduction plan.

[0074] Step S203: adaptively adjust the Brownian perturbation intensity according to the current optimization progress.

[0075] The nonlinear convergence factor can be obtained according to the current optimization progress; the Brownian perturbation intensity can be adaptively adjusted according to the nonlinear convergence factor. , as shown in formula (1).

[0076] (1)

[0077] in, represents the Brownian perturbation intensity, represents a random number, represents the nonlinear convergence factor.

[0078] According to the current optimization progress, the Brownian perturbation intensity is adaptively adjusted. The global perturbation ability can be enhanced in the early stage and the perturbation intensity can be reduced in the later stage, thereby achieving an efficient transition from global to local search.

[0079] Step S204: If the current stage is in the global search stage, a Brownian motion search mode is executed according to the Brownian perturbation intensity, and the current carbon emission reduction plan is iteratively updated multiple times to obtain a first carbon emission reduction plan.

[0080] If the current stage is in the global search stage, the Brownian motion search mode is executed according to the Brownian perturbation intensity to perform a wide-area search, update the current carbon emission reduction plan, and obtain the first carbon emission reduction plan.

[0081] Specifically, individuals in the population perform wide-area search through random Brownian motion, mimicking the lemming population, to effectively explore the entire solution space. The position update strategy of each individual is shown in Equation (2).

[0082] (2)

[0083] in, Indicates the Daidi The location of an individual, Indicates the Daidi the location of each individual; Indicates the current global optimal position; are other randomly selected individual positions; is a random number in the range [0, 1]; represents Brownian random perturbation, simulating the random search behavior of lemming groups exploring the environment; Indicates the direction adjustment factor (value is ).

[0084] Step S205: If the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, the first carbon emission reduction plan is updated, and the second carbon emission reduction plan is obtained; the switching stage corresponds to switching from the global search stage to the local development stage.

[0085] If the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode (also known as the Lévy flight strategy) is selected according to random probability, and the first carbon emission reduction plan is iteratively updated multiple times to obtain the second carbon emission reduction plan, which can improve the robustness of the artificial lemming optimization algorithm in complex scenarios.

[0086] Step S206: If the current stage is in the local development stage, the second carbon emission reduction plan is updated according to the spiral update strategy to obtain a third carbon emission reduction plan.

[0087] If the current stage is in the local development stage, that is, in the late stage of the artificial lemming optimization algorithm, individuals gradually gather in the neighborhood of the current optimal solution and conduct a refined search. Then, according to the spiral update strategy based on radius perturbation, the second carbon emission reduction plan can be iteratively updated multiple times to obtain the third carbon emission reduction plan, as shown in formula (3).

[0088] (3)

[0089] in, Indicates the Daidi The location of an individual, Indicates the Daidi the location of each individual; Indicates the current global optimal position, spiral search radius Defined as: ; Represents the Euclidean distance between the current individual and the optimal solution; Represents a random number; Indicates the direction adjustment factor (value is ).

[0090] Step S207: If the current optimization progress reaches the maximum number of function evaluations or the convergence threshold, an optimal carbon emission reduction plan that meets the optimization objectives and operation constraints is obtained according to the third carbon emission reduction plan.

[0091] If the current optimization progress reaches the maximum number of function evaluations or the convergence threshold, the third carbon emission reduction plan will be used as the optimal carbon emission reduction plan that meets the optimization objectives and operation constraints.

[0092] Among them, the optimization goal is to minimize the carbon emission level while ensuring the completion of production tasks and the satisfaction of basic operating constraints; the maximum number of function evaluations and the convergence threshold can be set according to actual conditions.

[0093] In the above-mentioned method for optimizing the carbon emission reduction plans of target enterprises, the Brownian perturbation intensity is adaptively adjusted according to the current optimization progress; if the current stage is in the global search stage, the Brownian motion search mode is executed according to the Brownian perturbation intensity, the current carbon emission reduction plan is updated, and the first carbon emission reduction plan is obtained; if the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, the first carbon emission reduction plan is updated, and the second carbon emission reduction plan is obtained, and then the optimal carbon emission reduction plan that meets the optimization objectives and operating constraints is obtained, which significantly enhances the accuracy of global search and local development, and is suitable for carbon emission reduction optimization scheduling scenarios under high-dimensional complex and multi-constraint conditions, and can meet the optimization requirements of "multi-time scale, dynamic adjustment, and global optimization" for the carbon emission paths of target enterprises in key industries.

[0094] In one embodiment, a carbon emission reduction optimization model is obtained based on the unit carbon emission factor of each target enterprise and the power load at each moment. The specific steps are as follows: based on the grid topology, the unit power generation carbon factor of each target enterprise and the output power at each moment, the unit carbon emission factor of each target enterprise is obtained; for each target enterprise, based on the unit carbon emission factor of the target enterprise and the power load at each moment, the carbon emission reduction assessment value of the target enterprise is obtained; the carbon emission reduction assessment values ​​of each target enterprise are added together to obtain the carbon emission reduction optimization model.

[0095] The unit carbon emission factor of each target enterprise can be obtained according to the grid topology, the unit power generation carbon factor of each target enterprise and the output power at each time, as shown in formula (4).

[0096] (4)

[0097] in, express node Equivalent carbon emission factor; Indicates the output power of the local power generation unit at that moment; is the unit carbon factor of electricity generated by the local power generation unit; Representation node To the node Active power transmitted; For nodes The set of all connected energy supply nodes; Represents the carbon emission factor of the adjacent node.

[0098] The unit carbon emission factor model shown in formula (4) is applicable to target enterprises in key industries, covering multiple time scales such as hours, days, months and quarters. It can more accurately capture the fluctuation characteristics of carbon emission intensity during the operation of target enterprises, significantly improve the practical adaptability and accuracy of carbon emission reduction potential assessment, and realize dynamic tracking of carbon emission intensity of target enterprises, providing a highly reliable input basis for the subsequent optimization of carbon emission reduction plans.

[0099] For each target enterprise, the carbon emission reduction assessment value of the target enterprise can be obtained based on the unit carbon emission factor of the target enterprise and the power load at each moment; the carbon emission reduction assessment values ​​of each target enterprise are added together to obtain the carbon emission reduction optimization model under multiple time scales, as shown in Formula (5).

[0100] (5)

[0101] Where N represents the number of target enterprises; Indicates the number of discrete time periods within the optimization cycle; Indicates the Enterprises at the moment power load; For the Enterprises at the moment The corresponding unit carbon emission factor.

[0102] In this embodiment, the unit carbon emission factor of each target enterprise is obtained based on the grid topology, the unit power generation carbon factor of each target enterprise and the output power at each moment, and then the carbon emission reduction assessment value of each target enterprise is obtained; the carbon emission reduction assessment values ​​of each target enterprise are added together to obtain a carbon emission reduction optimization model. This carbon emission reduction optimization model is suitable for carbon emission reduction optimization scheduling scenarios under high-dimensional complex and multi-constraint conditions, and can meet the optimization requirements of "multi-time scale, dynamic adjustment, and global optimization" for the carbon emission paths of target enterprises in key industries.

[0103] In one embodiment, the Brownian perturbation intensity is adaptively adjusted according to the current optimization progress. The specific steps are as follows: according to the current optimization progress, the current function evaluation number is obtained; according to the current function evaluation number and the maximum function evaluation number, a nonlinear convergence factor is obtained; according to the nonlinear convergence factor, the Brownian perturbation intensity is adaptively adjusted.

[0104] The current number of function evaluations can be obtained based on the current optimization progress; the nonlinear convergence factor can be obtained based on the current number of function evaluations and the maximum number of function evaluations, as shown in formula (6).

[0105] (6)

[0106] in, represents the nonlinear convergence factor, Indicates the current function evaluation times; Indicates the maximum number of function evaluations.

[0107] The Brownian perturbation intensity can be adaptively adjusted according to the nonlinear convergence factor , as shown in formula (1).

[0108] (1)

[0109] in, represents the Brownian perturbation intensity, represents a random number, represents the nonlinear convergence factor.

[0110] In this embodiment, by adaptively adjusting the Brownian perturbation intensity according to the current optimization progress, the global perturbation capability can be enhanced in the early stage and the perturbation intensity can be reduced in the later stage, thereby achieving an efficient global to local search transition.

[0111] In one embodiment, the method provided in the present application also includes: obtaining the current number of function evaluations based on the current optimization progress; obtaining the nonlinear convergence factor based on the current number of function evaluations and the maximum number of function evaluations; when the nonlinear convergence factor is less than a first preset value, determining that the current stage is the global search stage; when the nonlinear convergence factor is greater than the first preset value and less than the second preset value, determining that the current stage is the switching stage corresponding to the global search stage to the local development stage; when the nonlinear convergence factor is greater than the second preset value, determining that the current stage is the local development stage.

[0112] In order to achieve a smooth transition from the global exploration stage to the local development stage during the optimization process of the target enterprise's carbon emission reduction plan, a nonlinear convergence factor as shown in Equation (6) can be introduced to dynamically adjust the weight ratio of global exploration and local development to control the behavior conversion of different optimization stages and realize the switch from the global search stage to the local development stage; among them, the smaller the nonlinear convergence factor, the higher the weight ratio of local development.

[0113] (6)

[0114] in, represents the nonlinear convergence factor, Indicates the current function evaluation times; Indicates the maximum number of function evaluations; early artificial lemming optimization algorithm A larger value promotes global search; later Lower, populations tend to be more fine-grained, localized exploitation.

[0115] The first preset value and the second preset value can be set according to actual conditions.

[0116] When the nonlinear convergence factor is less than a first preset value, the current stage is determined to be the global search stage; when the nonlinear convergence factor is greater than the first preset value and less than a second preset value, the current stage is determined to be the switching stage corresponding to the global search stage to the local development stage; when the nonlinear convergence factor is greater than the second preset value, the current stage is determined to be the local development stage.

[0117] In this embodiment, a nonlinear convergence factor may be obtained according to the current optimization progress; and the current stage may be determined according to the magnitude relationship between the nonlinear convergence factor and the first preset value and the second preset value.

[0118] In one embodiment, if the current stage is in the global search stage, the Brownian motion search mode is executed according to the Brownian perturbation intensity, the current carbon emission reduction plan is updated, and the first carbon emission reduction plan is obtained. The specific steps are as follows: if the current stage is in the global search stage, the Brownian motion search mode is executed according to the Brownian perturbation intensity, and the current carbon emission reduction plan is updated for multiple iterations in the first stage to obtain the first carbon emission reduction plan in the first stage; if the current optimization progress converges and stagnates or falls into a local area, the first carbon emission reduction plan in the first stage is updated for multiple iterations in the second stage according to the heavy-tailed distribution random walk search mode to obtain the first carbon emission reduction plan in the second stage; if the current optimization progress does not converge and stagnate or falls into a local area, the Brownian motion search mode is executed according to the Brownian perturbation intensity, and the first carbon emission reduction plan in the second stage is updated for multiple iterations in the third stage to obtain the first carbon emission reduction plan in the third stage; based on the first carbon emission reduction plan in the third stage, the first carbon emission reduction plan is obtained.

[0119] If the current stage is in the global search stage, the Brownian motion search mode can be executed according to the Brownian perturbation intensity, and the current carbon emission reduction plan can be iteratively updated multiple times in the first stage to obtain the first carbon emission reduction plan in the first stage; among which, the number of iterative updates in the first stage can be set according to actual conditions.

[0120] After obtaining the first carbon emission reduction plan in the first stage, it can be determined whether the current optimization progress has converged and stagnated or fallen into a local area. If the current optimization progress has converged and stagnated or fallen into a local area, the first carbon emission reduction plan in the first stage is updated multiple times in the second stage according to the heavy-tailed random walk search mode to obtain the first carbon emission reduction plan in the second stage, as shown in Equation (7). The number of iterative updates in the second stage can be set according to actual conditions.

[0121] (7)

[0122] in, Indicates the Daidi The location of an individual, Indicates the Daidi the location of each individual; Indicates the current global optimal position; Indicates the jump factor that is dynamically adjusted as the search progresses; Represents a random step size based on generation, with dimension ; Indicates the direction adjustment factor (value is ).

[0123] If the current optimization progress converges and stagnates or falls into a local area, selecting the heavy-tailed distribution random walk search mode can help the population jump out of the local optimal area and achieve a leapfrog global jump.

[0124] After obtaining the first carbon emission reduction plan of the first stage, if the current optimization progress has not converged and stagnated or is stuck in a local area, the Brownian motion search mode is still executed according to the Brownian perturbation intensity, and the first carbon emission reduction plan of the first stage is updated multiple times in the second stage to obtain the first carbon emission reduction plan of the second stage;

[0125] After the first carbon reduction plan in the second phase, it can be determined whether the current optimization progress has converged, stagnated, or fallen into a local area. If the current optimization progress has not converged, stagnated, or fallen into a local area, the Brownian motion search mode is executed according to the Brownian perturbation intensity, and the first carbon reduction plan in the second phase is iteratively updated multiple times in the third phase to obtain the first carbon reduction plan in the third phase; and the first carbon reduction plan is obtained based on the first carbon reduction plan in the third phase. The number of iterative updates in the second phase can be set according to actual conditions.

[0126] After the first carbon reduction plan in the second phase, if the current optimization progress converges and stagnates or is stuck in a local area, the first carbon reduction plan in the second phase is iteratively updated in the third phase based on the heavy-tailed random walk search mode to obtain the first carbon reduction plan in the third phase; and the first carbon reduction plan is obtained based on the first carbon reduction plan in the third phase. The number of iterative updates in the second phase can be set according to actual conditions.

[0127] In this embodiment, if the current stage is in the global search stage, after obtaining the carbon emission reduction plan corresponding to each stage, it can be determined whether the current optimization progress has converged and stagnated or fallen into a local area. If the current optimization progress has converged and stagnated or fallen into a local area, the current stage carbon emission reduction plan is iteratively updated multiple times according to the heavy-tailed distribution random walk search mode to help the population jump out of the local optimal area and achieve a leapfrog global jump; if the current optimization progress has not converged and stagnated or fallen into a local area, the Brownian motion search mode is executed according to the Brownian perturbation intensity, and the current stage carbon emission reduction plan is iteratively updated multiple times to obtain the first carbon emission reduction plan.

[0128] In one embodiment, if the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, the first carbon emission reduction plan is updated, and the second carbon emission reduction plan is obtained. The specific steps are as follows: If the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, and the first carbon emission reduction plan is updated for multiple iterations in the first stage to obtain the second carbon emission reduction plan in the first stage; if the current optimization progress converges and stagnates or falls into a local area, the second carbon emission reduction plan in the first stage is updated for multiple iterations in the second stage according to the heavy-tailed distribution random walk search mode to obtain the second carbon emission reduction plan in the second stage; if the current optimization progress does not converge and stagnate or falls into a local area, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, and the second carbon emission reduction plan in the second stage is updated for multiple iterations in the third stage to obtain the second carbon emission reduction plan in the third stage; based on the second carbon emission reduction plan in the third stage, the second carbon emission reduction plan is obtained.

[0129] If the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, and the first carbon emission reduction plan is iteratively updated multiple times in the first stage to obtain the second carbon emission reduction plan in the first stage; among which, the number of iterative updates in the first stage can be set according to actual conditions.

[0130] After obtaining the second carbon emission reduction plan in the first stage, it can be determined whether the current optimization progress has converged and stagnated or fallen into a local area. If the current optimization progress has converged and stagnated or fallen into a local area, the second carbon emission reduction plan in the first stage is updated multiple times in the second stage according to the heavy-tailed random walk search mode to obtain the second carbon emission reduction plan in the second stage, as shown in Equation (7). The number of iterative updates in the second stage can be set according to actual conditions.

[0131] (7)

[0132] in, Indicates the Daidi The location of an individual, Indicates the Daidi the location of each individual; Indicates the current global optimal position; Indicates the jump factor that is dynamically adjusted as the search progresses; Represents a random step size based on generation, with dimension ; Indicates the direction adjustment factor (value is ).

[0133] If the current optimization progress converges and stagnates or falls into a local area, selecting the heavy-tailed distribution random walk search mode can help the population jump out of the local optimal area and achieve a leapfrog global jump.

[0134] After obtaining the second carbon emission reduction plan of the first stage, if the current optimization progress has not converged and stagnated or is stuck in a local area, the Brownian motion search mode is still executed according to the Brownian perturbation intensity, and the second carbon emission reduction plan of the first stage is updated multiple times in the second stage to obtain the second carbon emission reduction plan of the second stage;

[0135] After the second carbon reduction plan in the second phase, it can be determined whether the current optimization progress has converged, stagnated, or fallen into a local area. If the current optimization progress has not converged, stagnated, or fallen into a local area, a Brownian motion search mode is executed according to the Brownian perturbation intensity, and the second carbon reduction plan in the second phase is iteratively updated multiple times in the third phase to obtain the second carbon reduction plan in the third phase; and the second carbon reduction plan is obtained based on the second carbon reduction plan in the third phase. The number of iterative updates in the second phase can be set according to actual conditions.

[0136] After the second carbon reduction plan in the second phase, if the current optimization progress converges and stagnates or is stuck in a local area, the second carbon reduction plan in the second phase is iteratively updated in the third phase based on the heavy-tailed random walk search mode to obtain the second carbon reduction plan in the third phase; and the second carbon reduction plan is obtained based on the second carbon reduction plan in the third phase. The number of iterative updates in the second phase can be set according to actual conditions.

[0137] In this embodiment, if the current stage is in the switching stage, after obtaining the carbon emission reduction plan corresponding to each stage, it can be determined whether the current optimization progress has converged and stagnated or fallen into a local area. If the current optimization progress has converged and stagnated or fallen into a local area, the current stage carbon emission reduction plan is iteratively updated multiple times according to the heavy-tailed distribution random walk search mode to help the population jump out of the local optimal area and achieve a leapfrog global jump; if the current optimization progress has not converged and stagnated or fallen into a local area, the Brownian motion search mode is executed according to the Brownian perturbation intensity, and the current stage carbon emission reduction plan is iteratively updated multiple times to obtain a second carbon emission reduction plan.

[0138] In one embodiment, if the current stage is in the local development stage, the second carbon emission reduction plan is updated according to the spiral update strategy to obtain the third carbon emission reduction plan. The specific steps are as follows: if the current stage is in the local development stage, the second carbon emission reduction plan is updated with multiple iterations of the first stage according to the spiral update strategy to obtain the third carbon emission reduction plan of the first stage; if the current optimization progress converges and stagnates or falls into a local area, the third carbon emission reduction plan of the first stage is updated with multiple iterations of the second stage according to the heavy-tailed distribution random walk search mode to obtain the third carbon emission reduction plan of the second stage; if the current optimization progress does not converge and stagnate or falls into a local area, the third carbon emission reduction plan of the second stage is updated with multiple iterations of the third stage according to the spiral update strategy to obtain the third carbon emission reduction plan of the third stage; according to the third carbon emission reduction plan of the third stage, the third carbon emission reduction plan is obtained.

[0139] If the current stage is in the local development stage, according to the spiral update strategy, the second carbon emission reduction plan is iteratively updated multiple times in the first stage to obtain the third carbon emission reduction plan in the first stage; among which, the number of iterative updates in the first stage can be set according to actual conditions.

[0140] After obtaining the third carbon emission reduction scheme in the first phase, it can be determined whether the current optimization progress has converged and stagnated or fallen into a local area. If the current optimization progress has converged and stagnated or fallen into a local area, the third carbon emission reduction scheme in the first phase is updated multiple times in the second phase according to the heavy-tailed random walk search mode to obtain the third carbon emission reduction scheme in the second phase, as shown in Equation (7). The number of iterative updates in the second phase can be set according to actual conditions.

[0141] (7)

[0142] in, Indicates the Daidi The location of an individual, Indicates the Daidi the location of each individual; Indicates the current global optimal position; Indicates the jump factor that is dynamically adjusted as the search progresses; Represents a random step size based on generation, with dimension ; Indicates the direction adjustment factor (value is ).

[0143] If the current optimization progress converges and stagnates or falls into a local area, selecting the heavy-tailed distribution random walk search mode can help the population jump out of the local optimal area and achieve a leapfrog global jump.

[0144] After obtaining the third carbon emission reduction plan of the first stage, if the current optimization progress has not converged and stagnated or is stuck in a local area, the Brownian motion search mode is still executed according to the Brownian perturbation intensity, and the third carbon emission reduction plan of the first stage is updated multiple times in the second stage to obtain the third carbon emission reduction plan of the second stage;

[0145] After the third carbon reduction plan in the second phase, it can be determined whether the current optimization progress has converged, stagnated, or fallen into a local area. If the current optimization progress has not converged, stagnated, or fallen into a local area, a Brownian motion search mode is executed according to the Brownian perturbation intensity, and the third carbon reduction plan in the second phase is iteratively updated multiple times in the third phase to obtain the third carbon reduction plan in the third phase; based on the third carbon reduction plan in the third phase, a third carbon reduction plan is obtained. The number of iterative updates in the second phase can be set according to actual conditions.

[0146] After the third carbon reduction plan in the second phase, if the current optimization progress converges and stagnates or is stuck in a local area, the third carbon reduction plan in the second phase is iteratively updated multiple times in the third phase according to the heavy-tailed random walk search mode to obtain the third carbon reduction plan in the third phase; and the third carbon reduction plan is obtained based on the third carbon reduction plan in the third phase. The number of iterative updates in the second phase can be set according to actual conditions.

[0147] In this embodiment, if the current stage is in the local development stage, after obtaining the carbon emission reduction plan corresponding to each stage, it can be determined whether the current optimization progress has converged and stagnated or fallen into a local area. If the current optimization progress has converged and stagnated or fallen into a local area, the current stage carbon emission reduction plan is iteratively updated multiple times according to the heavy-tailed distribution random walk search mode to help the population jump out of the local optimal area and achieve a leapfrog global jump; if the current optimization progress has not converged and stagnated or fallen into a local area, the Brownian motion search mode is executed according to the Brownian perturbation intensity, and the current stage carbon emission reduction plan is iteratively updated multiple times to obtain a third carbon emission reduction plan.

[0148] In order to better understand the above method, an application example of the method for optimizing the carbon emission reduction plan of the target enterprise of this application is described in detail below.

[0149] Key industries such as steel, cement, chemicals, ceramics, and textiles are gradually transitioning to green and low-carbon development. To address these policy requirements and achieve carbon reduction goals, academia and industry have conducted extensive research on energy-carbon coupling models, optimized power load scheduling, and multi-period energy consumption control, aiming to provide businesses with more efficient and actionable carbon reduction technology pathways.

[0150] In actual applications, due to differences in production processes, electricity dependence, and process continuity, the carbon emission reduction plans of enterprises in different industries show diverse characteristics. Taking the steel and cement industries as an example, researchers generally adopt a scheduling method based on dynamic modeling of carbon factors. By adjusting the operating sequence of key links such as smelting, electric furnaces, and grinding mills, high-carbon periods are avoided to achieve decoupling of production and carbon emissions. In the chemical and ceramic industries, some models introduce a time-of-use electricity price and marginal emission linkage mechanism to construct a multi-objective optimization framework that takes into account multiple constraint variables such as cost, energy consumption, and emissions. Industries such as textiles, which are mainly intermittent loads, rely more on flexible scheduling and real-time adjustment strategies to improve the system's carbon responsiveness through flexible load migration.

[0151] To optimize carbon emission reduction solutions, current research has introduced swarm intelligence algorithms such as particle swarm optimization (PSO), differential evolution (DE), and genetic algorithms (GA) into enterprise scheduling and load optimization scenarios, achieving promising results at medium scales. However, when faced with complex boundary conditions, severe load fluctuations, and long operating cycles, traditional swarm intelligence algorithms still face challenges such as premature search, local optimality, and insufficient adaptability. These challenges make it difficult to meet the "multi-timescale, dynamic, and global optimization" requirements for optimizing carbon emission pathways for target enterprises in key industries.

[0152] Specifically, traditional carbon reduction scheme optimization methods simplify carbon emission factors to fixed values, failing to reflect actual carbon emission levels under changing grid structures, affecting assessment accuracy. Furthermore, traditional carbon reduction scheme optimization methods lack multi-timescale carbon reduction assessment mechanisms applicable to key industries; traditional swarm intelligence algorithms have limited global search capabilities and are prone to falling into local optimality; dynamic operational constraint modeling is insufficient, resulting in limited adaptability and flexibility; and optimization results are not easily interpretable or practical, making them difficult to guide actual decision-making.

[0153] Based on the above problems, this embodiment proposes a method for optimizing carbon emission reduction plans for target enterprises in key industries based on the artificial lemming optimization algorithm. This method fully considers the time series characteristics of the target enterprise's operating data and the dynamic changes in the emission coefficient, and constructs a multidimensional scheduling model with the goal of minimizing total carbon emissions. Through the random Brownian perturbation, direction perception mechanism and heavy-tailed distribution random walk search mode in the artificial lemming optimization algorithm, the iterative search of the load regulation path and the adaptive control of the scheduling boundary are collaboratively completed. This method can be applied to the hourly operating data analysis of the target enterprise, and supports the optimization of carbon emission reduction plans at different time scales such as day, month, and quarter.

[0154] This embodiment constructs a carbon emission reduction assessment mechanism applicable to key energy-consuming industries with dynamic adaptability to cope with the problems of ever-changing carbon emission factors and complex operating environments, and realizes carbon emission monitoring and scheduling optimization for different time levels such as day, month, and quarter.

[0155] In order to effectively deal with the challenges of multi-objective coupling, complex operation constraints, and dynamic changes in solution space in the optimization of carbon emission reduction for high-energy-consuming enterprises, this embodiment proposes a method for optimizing carbon emission reduction plans for target enterprises based on the artificial lemming optimization algorithm. Compared with traditional swarm intelligence optimization strategies such as Genetic Algorithm (GA), Simulated Annealing (SA), Artificial Immune Algorithm (IA), and Sled Dog Optimization (SDO), the artificial lemming optimization algorithm exhibits stronger stability and convergence efficiency in terms of global search capabilities, adaptive perturbation mechanisms, and escaping local optimal traps. This method is particularly suitable for complex scheduling scenarios where multiple time scales, dynamic carbon factors, and nonlinear boundary constraints coexist. By integrating with carbon emission models of different time granularities such as daily, monthly, and quarterly, the proposed method can accurately identify carbon emission reduction potential and output the optimal path without affecting the normal operation of the enterprise, providing key industry enterprises with efficient, intelligent, and executable low-carbon operation solutions. Including Figure 3 Steps shown:

[0156] S1: Data preparation stage: Collect basic data such as the target enterprise's power load, carbon factor, electricity price and production operation parameters.

[0157] S2: Parameter import: Construct input parameters such as electricity emission factor model, electricity consumption boundary, output-load correspondence, etc.

[0158] S3: Establish a unit carbon emission factor model: compare the baseline scenario with the optimized scenario and construct an emission reduction potential assessment structure system (Equations 1-7).

[0159] In real-world power systems, changes in node carbon emission factors are affected by a variety of factors and exhibit significant temporal fluctuations. This fluctuation typically stems from the dynamic evolution of the grid's energy supply structure, real-time changes in node loads, and the output characteristics of different types of power generation sources. Using a constant value for modeling will severely impact the accuracy and adaptability of the assessment results. To this end, this embodiment constructs a unit carbon emission factor model (also known as a time-varying carbon emission factor model) based on node power flow. This model fully incorporates information such as the grid topology, power generation unit type, and its carbon intensity, enhancing the model's ability to respond to the actual system's operating status. This unit carbon emission factor model can dynamically track carbon emission intensity, providing a highly reliable input basis for subsequent scheduling optimization. Its mathematical expression is shown in Equation (4).

[0160] (4)

[0161] in, express node Equivalent carbon emission factor; Indicates the output power of the local power generation unit at that moment; is the unit carbon factor of electricity generated by the local power generation unit; Representation node To the node Active power transmitted; For nodes The set of all connected energy supply nodes; Represents the carbon emission factor of the adjacent node.

[0162] S4: Construct a carbon emission reduction optimization model: comprehensively consider carbon emissions, electricity costs and operational stability, and define a multi-objective function form.

[0163] In view of the problems of uneven load distribution and dramatic fluctuations in carbon emissions over time in the actual operation of target enterprises in key industries, a carbon emission reduction optimization model under multiple time scales is constructed as shown in Equation (5). The optimization goal is to minimize the carbon emission level while ensuring the completion of production tasks and the satisfaction of basic operation constraints.

[0164] (5)

[0165] Where N represents the number of target enterprises; Indicates the number of discrete time periods within the optimization cycle; Indicates the Enterprises at the moment power load; For the Enterprises at the moment The corresponding unit carbon emission factor.

[0166] The operating constraints of this carbon emission reduction optimization model include: a) total energy consumption constraint sub-condition: the total electricity consumption of each target enterprise shall not exceed the production plan budget, and the total electricity consumption remains unchanged before and after optimization; b) equipment operation constraint sub-condition: the start and stop characteristics, minimum operating time and maximum load limits of each type of equipment shall be met; c) emission boundary constraint sub-condition: each target enterprise shall not exceed the set carbon emission limit in each time segment; d) operation stability constraint sub-condition: the task scheduling of the carbon emission reduction plan has temporal continuity and executable.

[0167] S5: Start the optimization algorithm solution: introduce the artificial lemming algorithm and iteratively search for the optimal carbon emission reduction solution.

[0168] To address the challenges faced by target enterprises in key industries in optimizing carbon emissions, such as strong operating load fluctuations, complex boundary constraints, and unclear control strategies, this embodiment proposes a multi-period carbon reduction potential identification and path optimization method based on an artificial lemming optimization algorithm. By simulating the collaborative exploration and jumping behavior of lemming groups during migration and combining it with a multi-target dynamic environment, an adaptive search strategy is constructed. This method offers the following four key innovations and improvements: 1) A collective search mechanism based on "leader-follower" behavior is constructed, in which individuals form multiple paths for parallel advancement through group perception and local perturbations, significantly improving the ability to globally identify carbon reduction directions; 2) A heavy-tailed random walk search pattern (also known as a Lévy flight strategy) is introduced to enhance search and jumping capabilities, effectively preventing the population from falling into local optima and maintaining diversity and global exploration intensity in a non-convex carbon emission space; 3) A Brownian motion perturbation mechanism is integrated to achieve flexible adaptation to multi-period emission boundaries, improving the distribution balance and engineering feasibility of the solution while satisfying operational constraints; 4) A dynamic approach factor and an adaptive step-size adjustment function are set to gradually focus the algorithm on the potential space boundary during the convergence phase, enhancing the stability and convergence accuracy of the final output solution.

[0169] 1. Introduction to the algorithm stage:

[0170] 1.1. Exploration phase:

[0171] In the early stages of the artificial lemming optimization algorithm, that is, in the global search phase, individuals in the population perform a wide-area search through the random Brownian motion of the bionic lemming group to effectively explore the entire solution space. The position update strategy of each individual is shown in Equation (2).

[0172] (2)

[0173] in, Indicates the Daidi The location of an individual, Indicates the Daidi the location of each individual; Indicates the current global optimal position; are other randomly selected individual positions; is a random number in the range [0, 1]; represents Brownian random perturbation, simulating the random search behavior of lemming groups exploring the environment; Indicates the direction adjustment factor (value is ).

[0174] 1.2. Heavy-tailed distribution random walk search mode:

[0175] When the population search is trapped in a local area or there is an obvious convergence stagnation, the artificial lemming optimization algorithm introduces a heavy-tailed distribution random walk search mode to help the population jump out of the local optimal area and achieve a leapfrog global jump, as shown in Equation (7).

[0176] (7)

[0177] in, Indicates the Daidi The location of an individual, Indicates the Daidi the location of each individual; Indicates the current global optimal position; Indicates the jump factor that is dynamically adjusted as the search progresses; Represents a random step size based on generation, with dimension ; Indicates the direction adjustment factor (value is ).

[0178] 1.3. Dynamic conversion mechanism from exploration to development:

[0179] In order to achieve a smooth transition from the global exploration stage to the local development stage during the optimization process of the target enterprise's carbon emission reduction plan, a nonlinear convergence factor as shown in Equation (6) can be introduced to dynamically adjust the weight ratio of global exploration and local development to control the behavior conversion of different optimization stages and realize the switch from the global search stage to the local development stage; among them, the smaller the nonlinear convergence factor, the higher the weight ratio of local development.

[0180] (6)

[0181] in, represents the nonlinear convergence factor, Indicates the current function evaluation times; Indicates the maximum number of function evaluations; early artificial lemming optimization algorithm A larger value promotes global search; later Lower, populations tend to be more fine-grained, localized exploitation.

[0182] 1.4. Local development stage:

[0183] In the late stage of the artificial lemming optimization algorithm, that is, when the current stage is in the local development stage, individuals gradually gather in the neighborhood of the current optimal solution and conduct a refined search. Specifically, a spiral update strategy based on radius perturbation is adopted, as shown in Equation (3).

[0184] (3)

[0185] in, Indicates the Daidi The location of an individual, Indicates the Daidi the location of each individual; Indicates the current global optimal position, spiral search radius Defined as: ; Represents the Euclidean distance between the current individual and the optimal solution; Represents a random number; Indicates the direction adjustment factor (value is ).

[0186] 2. Algorithm improvement mechanism:

[0187] To further improve the performance of the artificial lemming optimization algorithm in complex carbon emission optimization problems, the following optimization strategies are proposed:

[0188] 2.1. Adaptive disturbance intensity adjustment:

[0189] The nonlinear convergence factor can be obtained according to the current optimization progress; the Brownian perturbation intensity can be adaptively adjusted according to the nonlinear convergence factor. , as shown in formula (1).

[0190] (1)

[0191] in, represents the Brownian perturbation intensity, represents a random number, represents the nonlinear convergence factor.

[0192] According to the current optimization progress, the Brownian perturbation intensity is adaptively adjusted. The global perturbation ability can be enhanced in the early stage and the perturbation intensity can be reduced in the later stage, thereby achieving an efficient transition from global to local search.

[0193] 2.2. Collaborative behavior guidance mechanism:

[0194] Individuals communicate the current optimal solution information in real time to improve the group's ability to track high-quality search directions and promote rapid convergence.

[0195] 2.3. Dynamic random selection strategy:

[0196] If the current stage is in the switching stage from the global search stage to the local development stage, different search modes (Brownian motion search mode or heavy-tailed distribution random walk search mode) are selected according to random probability to improve the robustness of the artificial lemming optimization algorithm in complex scenarios.

[0197] 2.4 Nonlinear step size control strategy:

[0198] The search step size of the artificial lemming optimization algorithm changes with the number of evaluations without linear decay, as shown in formula (8).

[0199] (8)

[0200] This adaptive step size mechanism effectively balances the search range and solution accuracy, ensuring that the algorithm can operate effectively in each optimization stage.

[0201] The following is based on Figure 4 The artificial lemming optimization algorithm process shown in the figure (T represents the iteration cycle) details the solution process of the artificial lemming optimization algorithm in the carbon emission reduction optimization model:

[0202] Step 1: Initialize the parameters of the Artificial Lemming Algorithm (ALA) to obtain the initial parameters of the artificial lemming algorithm. The parameters of the artificial lemming algorithm include the population size N, the maximum number of function evaluations F, and the number of parameters of the artificial lemming algorithm. MEs , Brownian perturbation coefficient, heavy-tailed distribution random flight jump factor, nonlinear convergence factor (adaptive convergence factor).

[0203] Step 2: Based on the initial parameters of the artificial lemming optimization algorithm and the carbon emission reduction optimization model, an initial solution population that meets the operating constraints can be randomly generated, the fitness can be calculated, and the current global optimal individual can be selected. That is, an initial carbon emission reduction plan that meets the operating constraints can be randomly generated to obtain the current carbon emission reduction plan.

[0204] Step 3: If the current stage is in the global search stage, the group position is updated according to the global exploration mechanism, and Brownian motion random perturbation is performed to ensure the comprehensiveness of the search. That is, the Brownian motion search mode is executed according to the Brownian perturbation intensity, and the current carbon emission reduction plan is iteratively updated multiple times to obtain the first carbon emission reduction plan.

[0205] Step 4: Adjust the algorithm strategy based on the nonlinear convergence factor (adaptive factor), gradually switching from the global search phase to the local development phase. If the current phase is in the switching phase, select the Brownian motion search mode or the heavy-tailed random walk search mode based on random probability, update the first carbon reduction plan, and obtain the second carbon reduction plan.

[0206] Step 5: If the current optimization progress converges and stagnates or falls into a local area, the current carbon emission reduction plan is iteratively updated multiple times according to the heavy-tailed random walk search mode to reactivate the exploration vitality of the population;

[0207] Step 6: In the local development phase, a spiral update strategy is used to conduct a fine search around the current second carbon emission reduction plan to obtain a third carbon emission reduction plan, which can improve the accuracy of the third carbon emission reduction plan;

[0208] Step 7: Determine the termination condition. If the current optimization progress reaches the maximum number of function evaluations or the convergence threshold, obtain the optimal carbon emission reduction plan that meets the optimization objectives and operation constraints based on the third carbon emission reduction plan. Otherwise, return to step 3 to continue iteration.

[0209] It can achieve accurate assessment of the carbon emission reduction potential of target enterprises in key industries and output the optimal carbon emission reduction plan that meets optimization goals and operational constraints.

[0210] This embodiment provides a method for optimizing carbon emission reduction plans for target enterprises in key industries based on an artificial lemming optimization algorithm, which has the following beneficial effects:

[0211] (1) A unit carbon emission factor model covering multiple time scales of hours, days, months and quarters was established for target enterprises in key industries. This model can more accurately capture the fluctuation characteristics of carbon emission intensity during the operation of target enterprises and significantly improve the practical adaptability and accuracy of carbon emission reduction potential assessment.

[0212] (2) The artificial lemming optimization algorithm is introduced to solve the constructed carbon emission reduction optimization model. The artificial lemming optimization algorithm combines the Brownian random perturbation mechanism and the heavy-tailed distribution random walk search mode, which significantly enhances the global search capability and local search accuracy of the model. It is particularly suitable for carbon emission reduction optimization scheduling scenarios under high-dimensional complex and multi-constraint conditions. As shown in Figures 5 (a), 5 (b) and 5 (c), the weekly carbon reduction potential assessment effect diagram, monthly carbon reduction potential assessment effect diagram and quarterly carbon reduction potential assessment effect diagram of a target enterprise in a key industry based on the artificial lemming optimization algorithm are shown. This embodiment provides an optimization method for carbon emission reduction plans of target enterprises in key industries based on the artificial lemming optimization algorithm, which can meet the optimization requirements of "multi-time scale, dynamic adjustment and global optimization" for the carbon emission paths of target enterprises in key industries.

[0213] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily performed in sequence in the order indicated by the arrows. Unless clearly stated herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above may include multiple steps or multiple stages, and these steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of the steps or stages in other steps or other steps. It is understandable that the various steps in different embodiments can be freely combined as needed, and the various non-contradictory schemes formed by the combination all fall within the scope of protection of this application.

[0214] Based on the same inventive concept, the present application also provides a device for optimizing a target enterprise's carbon emission reduction plan for implementing the aforementioned method for optimizing a target enterprise's carbon emission reduction plan. The solution provided by this device is similar to the solution described in the aforementioned method. Therefore, the specific limitations of one or more embodiments of the device for optimizing a target enterprise's carbon emission reduction plan provided below can be found in the aforementioned definition of the method for optimizing a target enterprise's carbon emission reduction plan, and will not be further elaborated here.

[0215] In an exemplary embodiment, Figure 6 As shown, a device for optimizing a target enterprise's carbon emission reduction plan is provided, wherein:

[0216] The carbon emission reduction optimization model acquisition module 601 is used to obtain a carbon emission reduction optimization model based on the unit carbon emission factor of each target enterprise and the power load at each time;

[0217] The current carbon emission reduction plan acquisition module 602 is used to randomly generate an initial carbon emission reduction plan that meets the operation constraints based on the initial parameters of the artificial lemming optimization algorithm and the carbon emission reduction optimization model, and obtain the current carbon emission reduction plan;

[0218] Brownian perturbation intensity adjustment module 603, used to adaptively adjust the Brownian perturbation intensity according to the current optimization progress;

[0219] The current carbon emission reduction scheme update module 604 is configured to, if the current stage is in the global search stage, execute a Brownian motion search mode according to the Brownian perturbation intensity to update the current carbon emission reduction scheme to obtain a first carbon emission reduction scheme; if the current stage is in the switching stage, select a Brownian motion search mode or a heavy-tailed distribution random walk search mode according to random probability to update the first carbon emission reduction scheme to obtain a second carbon emission reduction scheme; the switching stage corresponds to switching from the global search stage to the local development stage; if the current stage is in the local development stage, update the second carbon emission reduction scheme according to the spiral update strategy to obtain a third carbon emission reduction scheme;

[0220] The optimal carbon emission reduction scheme acquisition module 605 is used to obtain the optimal carbon emission reduction scheme that meets the optimization objectives and operation constraints according to the third carbon emission reduction scheme if the current optimization progress reaches the maximum function evaluation times or the convergence threshold.

[0221] In one embodiment, the carbon emission reduction optimization model acquisition module 601 is also used to: obtain the unit carbon emission factor of each target enterprise based on the grid topology, the unit power generation carbon factor of each target enterprise and the output power at each time; for each target enterprise, obtain the carbon emission reduction assessment value of the target enterprise based on the unit carbon emission factor of the target enterprise and the power load at each time; and add the carbon emission reduction assessment values ​​of each target enterprise to obtain a carbon emission reduction optimization model.

[0222] In one embodiment, the Brownian perturbation intensity adjustment module 603 is further used to: obtain the current function evaluation number based on the current optimization progress; obtain the nonlinear convergence factor based on the current function evaluation number and the maximum function evaluation number; and adaptively adjust the Brownian perturbation intensity based on the nonlinear convergence factor.

[0223] In one embodiment, the device also includes a current stage determination module, which is used to: obtain the current function evaluation number based on the current optimization progress; obtain the nonlinear convergence factor based on the current function evaluation number and the maximum function evaluation number; when the nonlinear convergence factor is less than a first preset value, determine that the current stage is the global search stage; when the nonlinear convergence factor is greater than the first preset value and less than the second preset value, determine that the current stage is the switching stage corresponding to the global search stage to the local development stage; when the nonlinear convergence factor is greater than the second preset value, determine that the current stage is the local development stage.

[0224] In one embodiment, the current carbon emission reduction scheme update module 604 is further used to: if the current stage is in the global search stage, execute the Brownian motion search mode according to the Brownian perturbation intensity, perform multiple iterative updates on the current carbon emission reduction scheme in the first stage to obtain the first carbon emission reduction scheme in the first stage; if the current optimization progress converges and stagnates or falls into a local area, perform multiple iterative updates on the first carbon emission reduction scheme in the first stage according to the heavy-tailed distribution random walk search mode to obtain the first carbon emission reduction scheme in the second stage; if the current optimization progress does not converge and stagnate or falls into a local area, execute the Brownian motion search mode according to the Brownian perturbation intensity, perform multiple iterative updates on the first carbon emission reduction scheme in the second stage to obtain the first carbon emission reduction scheme in the third stage; obtain the first carbon emission reduction scheme based on the first carbon emission reduction scheme in the third stage.

[0225] In one embodiment, the current carbon emission reduction scheme update module 604 is further used to: if the current stage is in the switching stage, select the Brownian motion search mode or the heavy-tailed distribution random walk search mode according to random probability, and perform multiple iterative updates on the first carbon emission reduction scheme in the first stage to obtain the second carbon emission reduction scheme in the first stage; if the current optimization progress converges and stagnates or falls into a local area, then according to the heavy-tailed distribution random walk search mode, perform multiple iterative updates on the second carbon emission reduction scheme in the first stage in the second stage to obtain the second carbon emission reduction scheme in the second stage; if the current optimization progress does not converge and stagnate or falls into a local area, then select the Brownian motion search mode or the heavy-tailed distribution random walk search mode according to random probability, and perform multiple iterative updates on the second carbon emission reduction scheme in the second stage in the third stage to obtain the second carbon emission reduction scheme in the third stage; according to the second carbon emission reduction scheme in the third stage, obtain the second carbon emission reduction scheme.

[0226] In one embodiment, the current carbon emission reduction scheme update module 604 is further used to: if the current stage is in the local development stage, perform multiple iterative updates on the second carbon emission reduction scheme in the first stage according to the spiral update strategy to obtain the third carbon emission reduction scheme in the first stage; if the current optimization progress converges and stagnates or falls into a local area, perform multiple iterative updates on the third carbon emission reduction scheme in the first stage according to the heavy-tailed distribution random walk search mode to obtain the third carbon emission reduction scheme in the second stage; if the current optimization progress does not converge and stagnate or falls into a local area, perform multiple iterative updates on the third carbon emission reduction scheme in the third stage according to the spiral update strategy to obtain the third carbon emission reduction scheme in the third stage; obtain the third carbon emission reduction scheme according to the third carbon emission reduction scheme in the third stage.

[0227] Each module in the aforementioned target enterprise carbon emission reduction plan optimization device may be implemented in whole or in part through software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor within a computer device in the form of hardware, or may be stored in a computer device's memory in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0228] In an exemplary embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 6 As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O) and a communication interface. The processor, memory and input / output interface are connected via a system bus, and the communication interface is connected to the system bus via the input / output interface. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store data of an embodiment of a method for optimizing a carbon emission reduction plan for a target enterprise. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, a method for optimizing a carbon emission reduction plan for a target enterprise is implemented.

[0229] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0230] In one embodiment, a computer device is further provided, including a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0231] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0232] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0233] 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, and the collection, use and processing of relevant data must comply with relevant regulations.

[0234] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), quantum computing-based data processing logic devices, artificial intelligence (AI) processors, and the like.

[0235] The technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0236] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. A method for optimizing a target enterprise's carbon emission reduction plan, characterized in that: The method comprises: Based on the unit carbon emission factor of each target enterprise and the power load at each moment, a carbon emission reduction optimization model is obtained; Randomly generate an initial carbon emission reduction plan that meets the operation constraints based on the initial parameters of the artificial lemming optimization algorithm and the carbon emission reduction optimization model, and obtain a current carbon emission reduction plan; Adaptively adjust the Brownian perturbation intensity according to the current optimization progress; If the current stage is in the global search stage, executing the Brownian motion search mode according to the Brownian perturbation intensity, updating the current carbon emission reduction plan, and obtaining a first carbon emission reduction plan; If the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, the first carbon emission reduction plan is updated, and the second carbon emission reduction plan is obtained; the switching stage corresponds to switching from the global search stage to the local development stage; If the current stage is in the local development stage, then according to the spiral update strategy, the second carbon emission reduction plan is updated to obtain the third carbon emission reduction plan; If the current optimization progress reaches the maximum number of function evaluations or the convergence threshold, an optimal carbon emission reduction plan that meets the optimization objectives and operation constraints is obtained according to the third carbon emission reduction plan.

2. The method according to claim 1, characterized in that The carbon emission reduction optimization model is obtained based on the unit carbon emission factor of each target enterprise and the power load at each moment, including: According to the grid topology, the unit carbon factor of each target enterprise and the output power at each time, the unit carbon emission factor of each target enterprise is obtained; For each target enterprise, obtain the carbon emission reduction assessment value of the target enterprise based on the unit carbon emission factor of the target enterprise and the power load at each time; The carbon emission reduction assessment values ​​of each target enterprise are added together to obtain a carbon emission reduction optimization model.

3. The method according to claim 1, characterized in that The adaptive adjustment of the Brownian perturbation intensity according to the current optimization progress includes: According to the current optimization progress, get the current function evaluation times; Obtaining a nonlinear convergence factor according to the current function evaluation number and the maximum function evaluation number; The Brownian perturbation intensity is adaptively adjusted according to the nonlinear convergence factor.

4. The method according to claim 1, wherein The method further comprises: According to the current optimization progress, get the current function evaluation times; Obtaining a nonlinear convergence factor according to the current function evaluation number and the maximum function evaluation number; When the nonlinear convergence factor is less than a first preset value, determining that the current stage is a global search stage; When the nonlinear convergence factor is greater than a first preset value and less than a second preset value, determining that the current stage is a switching stage corresponding to the global search stage to the local development stage; When the nonlinear convergence factor is greater than a second preset value, the current stage is determined to be a local development stage.

5. The method according to claim 1, wherein If the current stage is in the global search stage, executing the Brownian motion search mode according to the Brownian perturbation intensity, updating the current carbon emission reduction plan, and obtaining a first carbon emission reduction plan, including: If the current stage is in the global search stage, a Brownian motion search mode is executed according to the Brownian perturbation intensity, and the current carbon emission reduction plan is iteratively updated multiple times in the first stage to obtain a first carbon emission reduction plan in the first stage; If the current optimization progress converges and stagnates or falls into a local area, the first carbon emission reduction plan of the first stage is iteratively updated multiple times in the second stage according to the heavy-tailed distribution random walk search mode to obtain the first carbon emission reduction plan of the second stage; If the current optimization progress does not converge and stagnates or falls into a local area, the Brownian motion search mode is executed according to the Brownian perturbation intensity, and the first carbon emission reduction plan of the second stage is iteratively updated for multiple stages to obtain the first carbon emission reduction plan of the third stage; According to the first carbon emission reduction plan of the third stage, a first carbon emission reduction plan is obtained.

6. The method according to claim 1, characterized in that If the current stage is in the switching stage, selecting the Brownian motion search mode or the heavy-tailed distribution random walk search mode according to random probability, updating the first carbon emission reduction plan, and obtaining the second carbon emission reduction plan, including: If the current stage is in the switching stage, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, and the first carbon emission reduction plan is iteratively updated multiple times in the first stage to obtain the second carbon emission reduction plan in the first stage; If the current optimization progress converges and stagnates or falls into a local area, the second carbon emission reduction plan of the first stage is updated multiple times in the second stage according to the heavy-tailed distribution random walk search mode to obtain the second carbon emission reduction plan of the second stage; If the current optimization progress does not converge, stagnates, or falls into a local area, the Brownian motion search mode or the heavy-tailed distribution random walk search mode is selected according to random probability, and the second carbon emission reduction plan of the second stage is updated multiple times in the third stage to obtain the second carbon emission reduction plan of the third stage; According to the second carbon emission reduction plan of the third stage, a second carbon emission reduction plan is obtained.

7. The method according to claim 1, characterized in that If the current stage is in the local development stage, the second carbon emission reduction plan is updated according to the spiral update strategy to obtain a third carbon emission reduction plan, including: If the current stage is in the local development stage, according to the spiral update strategy, the second carbon emission reduction plan is iteratively updated multiple times in the first stage to obtain the third carbon emission reduction plan in the first stage; If the current optimization progress converges and stagnates or falls into a local area, the third carbon emission reduction plan of the first stage is updated multiple times in the second stage according to the heavy-tailed distribution random walk search mode to obtain the third carbon emission reduction plan of the second stage; If the current optimization progress does not converge and stagnates or falls into a local area, the third carbon emission reduction plan of the second stage is updated multiple times in the third stage according to the spiral update strategy to obtain the third carbon emission reduction plan of the third stage; According to the third carbon emission reduction plan of the third stage, a third carbon emission reduction plan is obtained.

8. A device for optimizing a target enterprise's carbon emission reduction plan, characterized in that: The device comprises: A carbon emission reduction optimization model acquisition module is used to obtain a carbon emission reduction optimization model based on the unit carbon emission factor of each target enterprise and the power load at each moment; A current carbon emission reduction plan acquisition module is used to randomly generate an initial carbon emission reduction plan that meets the operation constraints based on the initial parameters of the artificial lemming optimization algorithm and the carbon emission reduction optimization model, and obtain the current carbon emission reduction plan; Brownian perturbation intensity adjustment module, used to adaptively adjust the Brownian perturbation intensity according to the current optimization progress; A current carbon emission reduction scheme update module is configured to, if the current stage is in the global search stage, execute a Brownian motion search mode according to the Brownian perturbation intensity to update the current carbon emission reduction scheme to obtain a first carbon emission reduction scheme; if the current stage is in the switching stage, select a Brownian motion search mode or a heavy-tailed distribution random walk search mode according to random probability to update the first carbon emission reduction scheme to obtain a second carbon emission reduction scheme; the switching stage corresponds to switching from the global search stage to the local development stage; if the current stage is in the local development stage, update the second carbon emission reduction scheme according to the spiral update strategy to obtain a third carbon emission reduction scheme; The optimal carbon emission reduction scheme acquisition module is used to obtain the optimal carbon emission reduction scheme that meets the optimization objectives and operation constraints according to the third carbon emission reduction scheme if the current optimization progress reaches the maximum number of function evaluations or the convergence threshold.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images