CN-121981306-A - PSO-GWO hybrid algorithm-based grid enterprise carbon quota collaborative optimization method, system, equipment and medium

Abstract

The invention discloses a PSO-GWO hybrid algorithm-based power grid enterprise carbon quota collaborative optimization method, a system, equipment and a medium, which belong to the technical field of power grid enterprise carbon quota optimization and comprise the steps of establishing a dynamic carbon emission factor modeling mechanism to obtain a power carbon emission coefficient of a consumption node, constructing a multi-period carbon quota allocation and scheduling optimization model constraint carbon emission total minimization for a power grid enterprise based on the power carbon emission coefficient, carrying out double-layer carbon quota allocation based on a PSO-GWO hybrid algorithm, and carrying out intelligent optimization scheduling based on an algorithm result. The method effectively improves the accuracy and adaptability of carbon quota allocation, realizes the accurate allocation of carbon quota, overcomes the problem that the traditional method is easy to fall into local extremum, improves the global performance of overall quota optimization, realizes the organic integration of emission reduction cost minimization and a carbon emission peak regulation and control path, reduces the overall carbon emission reduction cost of enterprises, and improves the controllability of a carbon peak target.

Inventors

• WANG BIN
• WANG WEI
• LUO NING
• ZHU YONGQING
• WANG YUXIANG
• YIN JIA
• LIN CHAO
• CHEN JULONG
• ZHANG YU
• MOU XUEPENG
• YANG SHIPING
• LUO CHEN
• WANG CE
• HU BIN
• TANG XUEYONG

Assignees

• 贵州电网有限责任公司

Dates

Publication Date

20260505

Application Date

20251128

Claims (10)

1. 1. A power grid enterprise carbon quota collaborative optimization method based on a PSO-GWO hybrid algorithm is characterized by comprising the following steps of, Establishing a dynamic carbon emission factor modeling mechanism, and acquiring carbon emission factors of consumption nodes; Based on the carbon emission factors, constructing a multi-period carbon quota allocation and scheduling optimization model for power grid enterprises to restrict the total carbon emission to be minimized; And performing double-layer carbon quota allocation based on a PSO-GWO hybrid algorithm, and performing intelligent optimal scheduling based on an algorithm result.
2. 2. The PSO-GWO hybrid algorithm-based grid enterprise carbon quota collaborative optimization method as set forth in claim 1, wherein the dynamic carbon emission factor modeling mechanism comprises the steps of constructing the dynamic carbon emission factor modeling mechanism based on a grid running state and power flow distribution; The carbon emission factor is calculated for each node at any time by the combined effect of the emission level of the local power generation unit, the external active power flow and the remaining node emission factor.
3. 3. The PSO-GWO hybrid algorithm-based grid enterprise carbon quota collaborative optimization method is characterized in that the constructing a multi-period carbon quota allocation and scheduling optimization model for a grid enterprise comprises constructing a multi-period carbon quota allocation and scheduling optimization model for the grid enterprise based on the carbon emission factor and with the aim of minimizing total carbon emission; The constraint conditions are set as quota constraint, total quota constraint, operation constraint and engineering constraint.
4. 4. The method for grid enterprise carbon quota collaborative optimization based on PSO-GWO hybrid algorithm as set forth in claim 3, wherein the PSO-GWO hybrid algorithm comprises introducing a global-local dynamic weight allocation mechanism, constructing a double-layer collaborative feedback architecture, and feeding quota allocation results back to intelligent scheduling optimization in real time; and (3) adopting nonlinear convergence weight adjustment, setting a mixed disturbance strategy, and introducing an elite guiding, global disturbance and local fine adjustment triple mechanism.
5. 5. The method for grid enterprise carbon quota collaborative optimization based on the PSO-GWO hybrid algorithm, as set forth in claim 4, wherein the double-layer carbon quota allocation comprises a exploration phase, an exploration-development conversion mechanism phase and a development phase based on the PSO-GWO hybrid algorithm; the exploration stage uses multi-particle parallel search as a core, and combines global particle guidance of PSO and a multi-captain collaboration mechanism of GWO, and each solution vector is regarded as a particle/gray wolf position Using PSO global speed-location update , , wherein, In order to be able to achieve a particle velocity, For the historical optimal location of the individual, As a global optimum position for the device, As the weight of the inertia is given, As the acceleration coefficient of the vehicle, the vehicle is, 、 Is that Is a uniform random number of (a); at part GWO, the collaborative correction is guided using the three leader Alpha, beta, delta, updated as follows: Wherein, the As a parameter of linear convergence, Is that A uniform random number is used for the random number, 、 、 The three leader leads the collaborative correction result respectively to Alpha, beta, delta, 、 、 Is a correction factor guided by different guidance leader, adjusts the position update of the individual, Is that A control factor, determining the step size and the individual update amplitude during the search, 、 、 Guidance correction of three captchas to integrate influence In order to guide the coefficients of the coefficients, Is the first in the particle swarm algorithm The position of the individual particles is determined, Optimizing seed of wolf The position of the individual wolves is determined, 、 、 The positions of the wolves are Alpha, beta, delta respectively, Represents the optimal individual in the population, Represents the second most optimal individual in the population, Representing the third best individual in the population. Two search schemes are fused , wherein, 、 Is self-adaptive, i.e ; The exploration-development conversion mechanism is characterized in that an algorithm dynamically adjusts searching and convergence weights, and inertia weights decrease formula: Wherein the convergence parameter Decreasing with iteration; after the development stage converges to the optimal solution neighborhood, focusing on global optimum and high-precision fine adjustment of Alpha wolves, and introducing target disturbance items , wherein, As a perturbation factor, the light-scattering factor, Is a globally optimal solution.
6. 6. The method for grid enterprise carbon quota collaborative optimization based on PSO-GWO hybrid algorithm as set forth in claim 5, wherein the dual-layer carbon quota allocation further comprises selecting top k fitness optimal solutions, and generating a global guide vector by linear combination : When the adaptability change is smaller than the threshold value, the Gaussian disturbance is introduced to jump out the local extremum , wherein, For the amplitude of the disturbance, Is normally distributed, partial individuals carry out offset updating according to global optimum , wherein, Is an adaptive variation factor, and integrates inertia weight and nonlinear disturbance step length : Wherein, the Represents the first The optimal location of the individual(s), Representation and individual The weight factor associated with the weight factor is, Representing individuals Is used for the disturbance location of (a), Dynamic perturbation terms for adjusting the intensity of the perturbation in the search strategy, Representing individuals Is used for the position of variation of (a), Representing the total number of iterations and, Representing the position of the globally optimal solution in the iterative process.
7. 7. The method for grid enterprise carbon quota collaborative optimization based on PSO-GWO hybrid algorithm as set forth in claim 6, wherein the intelligent optimization scheduling comprises the step of directly using a multi-period carbon quota allocation and scheduling optimization model for the grid enterprise as a fitness function under the condition that constraint conditions are met The method comprises the following steps: Where N represents a total of N different entities or devices and T represents the total number of time steps, typically used to represent optimization of carbon emissions or resource consumption over a plurality of time periods. Initializing population scale, maximum iteration times and key parameters of upper and lower layer optimization, setting an initial quota scheme, wherein the upper layer generates a plurality of groups of quota allocation solutions by adopting a PSO-GWO mixing mechanism, the lower layer performs dynamic scheduling optimization on the basis of 24-hour electricity load and time-varying carbon factors for each group of quota, if the lower layer scheme does not meet the quota or operation constraint, information is returned, an upper layer quota strategy is adjusted, the current population is adjusted and optimized by elite cooperation, global disturbance and dynamic weight, the current population is prevented from being trapped in local optimization, whether iteration termination conditions are met is judged, if so, the optimal quota allocation and scheduling scheme is output, otherwise, quota allocation is performed on different power grid enterprise nodes and time periods in the step of generating a plurality of groups of quota allocation solutions.
8. 8. The PSO-GWO hybrid algorithm-based grid enterprise carbon quota collaborative optimization system is applied to the PSO-GWO hybrid algorithm-based grid enterprise carbon quota collaborative optimization method, and is characterized by comprising a modeling module, a target optimization module and an optimization module; the modeling module is used for establishing a dynamic carbon emission factor modeling mechanism and acquiring an electric carbon emission coefficient of the consumption node; The target optimization module is used for constructing a multi-period carbon quota allocation and scheduling optimization model for a power grid enterprise to restrict the total carbon emission to be minimized based on the electric power carbon emission coefficient; and the optimizing module is used for carrying out double-layer carbon quota allocation based on a PSO-GWO mixing algorithm and carrying out intelligent optimizing scheduling based on an algorithm result.
9. 9. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of a grid enterprise carbon quota collaborative optimization method based on a PSO-GWO hybrid algorithm as defined in any one of claims 1 to 7.
10. 10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of a grid enterprise carbon quota co-optimization method based on a PSO-GWO hybrid algorithm as claimed in any one of claims 1 to 7.

Description

PSO-GWO hybrid algorithm-based grid enterprise carbon quota collaborative optimization method, system, equipment and medium Technical Field The invention relates to the technical field of power grid enterprise carbon quota optimization, in particular to a power grid enterprise carbon quota collaborative optimization method, system, equipment and medium based on a PSO-GWO hybrid algorithm. Background The power grid enterprises serve as energy source conveying and scheduling hubs, and carbon emission of the power grid enterprises affects emission reduction responsibility division of power generation ends and energy utilization ends and is directly related to realization of regional and even national carbon emission targets. With the gradual perfection of carbon emission right trading markets and quota management policies, the carbon quota management pressure faced by power grid enterprises is increasingly increased, and demands for efficient and intelligent quota allocation and scheduling methods are becoming urgent. Scientific and reasonable carbon quota allocation not only can effectively restrict the upper limit of carbon emission of power grid enterprises, but also can excite the enterprises to optimize the scheduling strategy, improve the energy efficiency level and the carbon emission reduction capacity. How to combine the operation characteristics and dynamic load changes of power grid enterprises and to formulate a carbon quota allocation mechanism with adaptability and excitation effect becomes one of the core technical problems of green transformation. The currently mainstream carbon allotment methods include a historical total amount method, a historical strength method, and a baseline method. The historical strength method is widely adopted in the power grid industry because of being capable of considering the historical emission efficiency and the output level. However, these methods are mainly based on static historical data, and it is difficult to reflect dynamic change characteristics of grid load and carbon emission, so that quota allocation is disjointed from actual operation. In recent years, the academy has developed many searches around the dynamic optimization of carbon quota, for example, a learner utilizes an intelligent model such as gray prediction-PSO-BPNN to dynamically predict and allocate regional power grid carbon emission, so that the fairness and adaptability of multi-region quota allocation are improved. The carbon quota mechanism is also researched to be applied to the scenes of micro-grids, energy storage, demand response and the like, and a time self-adaptive quota allocation method is provided by considering the carbon emission time-varying factors, so that the dual improvement of emission reduction and economy is realized. Despite the progress of related research in local scenarios, there is generally no dual-layer system modeling for the power grid enterprise, the energy and carbon flow hub. In the existing method, quota allocation and operation scheduling splitting treatment are mostly carried out, unified intelligent algorithm cooperative optimization is lacked, and organic fusion of quota design, operation cost control and carbon peak reaching targets is difficult to achieve. The existing power grid enterprise carbon quota allocation method depends on static historical data, dynamic carbon emission characteristics in power grid operation are difficult to reflect in real time, so that quota allocation results are disjointed with actual scheduling and emission reduction effects, the existing method adopts single optimization algorithm or traditional linear programming, is prone to being in local optimum, is difficult to ensure global optimization and fairness of carbon quota allocation, lacks joint modeling of 24-hour daily power grid load distribution and time-varying carbon emission factors, fails to fully mine carbon reduction potential in the power grid operation process, and is insufficient in model closed loop performance due to quota allocation and actual operation scheduling cutting, and the existing scheme generally has the problems of low calculation efficiency, slow convergence speed and the like when large-scale data and long-period optimization are achieved. Disclosure of Invention The present invention has been made in view of the above-described problems. Therefore, the invention aims to provide the double-layer carbon quota optimal allocation method which is suitable for the power grid enterprises and can dynamically respond to the time-varying carbon emission factors and the multidimensional operation constraint of the power grid, so that scientific allocation and refined scheduling of the carbon emission quota of the power grid enterprises are realized, and the fairness, the economy and the carbon peak management and control capability of the carbon quota management are effectively improved. In order to solve the technical problems, the invention prov