A low-carbon scheduling method for multi-agent integrated energy systems considering wind and solar correlation

Abstract

The present invention discloses a low-carbon scheduling method for a multi-agent integrated energy system that considers the correlation between wind and solar power. The method comprises the following steps: S1: establishing an integrated energy system structure and constructing an integrated energy system energy trading process that includes a green certificate-tiered carbon trading joint trading market mechanism; S2: within the integrated energy system structure, energy trading service providers are set as upper-level leaders, and energy suppliers and users are set as lower-level followers, establishing a multi-agent master-slave game scheduling mechanism; S3: generating wind-solar combined output scenarios based on historical distributed energy data, and introducing wind-solar correlation into the integrated energy system structure; S4: constructing a decision-making model for each stakeholder in the integrated energy system, including energy trading service providers, energy suppliers, and users; and S5: introducing a hierarchical differential evolution algorithm as a solution method and performing game-based solutions on the decision-making models of each subject in different scenarios to generate optimized scheduling results.

Description

Technical Field

[0001] The present invention relates to the field of integrated energy system optimization, and in particular to a low-carbon scheduling method for a multi-agent integrated energy system taking into account the correlation between wind and solar power. Background Art

[0002] An integrated energy system integrates multiple energy sources, achieving coordinated optimization and achieving full energy utilization and energy conservation and emission reduction. To enhance energy conservation and emission reduction and renewable energy consumption, my country has successively formulated carbon trading policies and green certificate trading policies, stimulating the carbon reduction capabilities of the integrated energy system.

[0003] Furthermore, the uncertainty of renewable energy also impacts the integrated energy system's ability to absorb new energy. In recent years, Copula theory has begun to be applied to the study of uncertainty in renewable energy generation, and considering the correlation between wind and solar power has shown potential in enhancing the ability of integrated energy systems to absorb new energy.

[0004] The operation and optimization of integrated energy systems rely on the coordinated efforts of multiple stakeholders. Game theory is a key approach to resolving conflicts of interest among these various market players. The Stackelberg game, a framework that describes the interactive relationship between leaders and followers, plays a crucial role in resolving these conflicts.

[0005] The differential evolution algorithm can dynamically track the current search status and adjust the search strategy, demonstrating strong global convergence and robustness. The master-slave game framework is a two-layer optimization and scheduling model. The two-layer differential evolution algorithm can nest lower-layer optimization within the upper-layer optimization, enabling rapid optimization within the master-slave game framework. Summary of the Invention

[0006] In order to promote the consumption of renewable energy in the integrated energy system model, the present invention provides a low-carbon scheduling method for a multi-agent integrated energy system considering the correlation between wind and solar power, taking into account the correlation characteristics between distributed energy sources and the green certificate-carbon joint trading mechanism, and verifying the effectiveness of the model under the master-slave game scheduling framework.

[0007] The above purpose is achieved through the following technical solutions:

[0008] The present invention provides a low-carbon scheduling method for a multi-agent integrated energy system considering wind and solar correlation, the method comprising the following steps:

[0009] S1: Establish an integrated energy system structure and construct an integrated energy system energy trading process including a green certificate-tiered carbon trading joint trading market mechanism;

[0010] S2: In the integrated energy system structure, energy trading service providers are set as upper-level leaders, energy suppliers and users are set as lower-level followers, and a multi-agent master-slave game scheduling mechanism is established;

[0011] S3: Generate wind-solar combined output scenarios based on historical distributed energy data, and introduce wind-solar correlation into the integrated energy system structure;

[0012] S4: Build a decision-making model for all stakeholders in the integrated energy system, including energy trading service providers, energy suppliers, and users;

[0013] S5: Introduce the hierarchical differential evolution algorithm as a solution method and perform game-based solution on the decision-making models of various stakeholders in different scenarios to generate optimized scheduling results.

[0014] Furthermore, the green certificate-ladder carbon trading joint trading market mechanism described in step S1 is constructed as follows:

[0015] First, we design the carbon trading mechanism CET, adopt a stepped carbon trading cost mechanism, and construct the carbon trading cost as shown below: :

[0016] ,

[0017] Where, is the carbon trading base price, is the growth rate of carbon trading price, is the length of the carbon emission interval, is the actual carbon emissions of the integrated energy system;

[0018] Secondly, a green certificate trading mechanism GCT is designed, where the green certificate transaction cost calculation method is:

[0019] ,

[0020] Where, represents the transaction cost of green certificates, is the unit price of green certificate transaction, The green certificate quota coefficient required for the integrated energy system is is the electric load of the integrated energy system at time t, The quantitative coefficient of converting renewable energy power generation into green certificates, 、 is the output power of distributed photovoltaic and wind power at time t;

[0021] Finally, a green certificate-ladder carbon trading mechanism is designed. Energy suppliers can obtain carbon emission permits in addition to green certificates. The calculation method for converting renewable energy into green certificates is as follows:

[0022] ,

[0023] Where, In order to absorb the carbon emission quota of new energy, is the conversion coefficient, the conversion coefficient The conversion coefficient is determined by the base price of the green certificate market and the carbon trading market. Expressed as .

[0024] Furthermore, the specific method of step S2 is to analyze the pursuit of the optimal decision of the energy trading service provider ETSP, the energy supplier ES and the user USER based on the Stackelberg game theory. In the multi-agent Stackelberg game framework, the master-slave game consists of three parts: the participant set N; the game strategy set of the energy trading service provider ETSP, the energy supplier ES and the user USER 、 、 ; The interests of energy trading service providers ETSP, energy suppliers ES and users USER 、 、 ; The Stackelberg game model is specifically expressed as follows:

[0025] ,

[0026] The set of game participants N includes ETSP, ES, and USER, then Expressed as:

[0027] ,

[0028] Where, For the leader ETSP; 、 They are follower ES and USER respectively.

[0029] Furthermore, the specific method of step S3 is to use the kernel density estimation method to estimate the marginal distribution functions of wind and solar output in different time periods based on historical data, and use the Copula function to model the wind-solar joint distribution function. Finally, random sampling is performed through inverse transformation to generate typical scenarios. The specific scenario generation steps are as follows:

[0030] S3-1. Use kernel density estimation method to generate the probability density function of wind and solar output for each 24-hour period based on historical photovoltaic and wind power data;

[0031] S3-2. Establish a joint probability distribution function of photovoltaic and wind power based on the Frank-Copula function for each time period;

[0032] S3-3. Sample the photovoltaic and wind power probability distribution functions based on the Frank-Copula function for each time period, perform an inverse transformation on the joint probability distribution function based on the sampling results, and finally obtain the wind and photovoltaic outputs for each time period, that is, generate the wind and solar outputs considering the correlation;

[0033] S3-4. Use the K-means clustering algorithm to cluster the sampling results to generate a wind-solar combined power scenario. The specific steps are as follows: randomly select k data points as the initial cluster centers; assign data points, assign each data point in the data set to the nearest cluster center, and use Euclidean distance to measure the distance; update the cluster center, calculate the average value of all points in each cluster, and update the average value as the new cluster center; iterate, repeating the above steps until the cluster center no longer changes or the preset number of iterations is reached;

[0034] S3-5. Verify the effectiveness of the wind-solar correlation for the integrated energy system model. At the same time, introduce wind curtailment, solar curtailment, and wind-solar fluctuation standard deviation as measurement indicators. The calculation formula for wind-solar fluctuation standard deviation is as follows:

[0035] σ= 1 24 ∑ t=1 24 [( P pv,t + P w,t ) - 1 24 ∑ t=1 24 ( P pv,t + P w,t ) ] ,

[0036] Where, is the standard deviation of wind and solar fluctuations, 、 are the photovoltaic and wind power outputs at time t respectively.

[0037] Furthermore, the specific method of step S4 is to first design the objective function of each stakeholder. Among them, the objective function of the energy trading service takes the time-of-use electricity price in different time periods as a reference basis, and its benefit comes from the energy complementarity between other stakeholder groups. The objective function of the energy trading service is expressed as:

[0038] ,

[0039] Where, The proceeds from energy sales to ETSP; The cost of energy purchase when ETSP interacts with ES; The cost of electricity purchase for ETSP to interact with the grid;

[0040] Energy suppliers establish an objective function based on the energy prices set by energy trading service providers and their own interests. The energy supplier objective function is as follows:

[0041] ,

[0042] Where, Profit from energy sales for ES; Operation and maintenance costs; 、 They are carbon trading and green certificate transaction costs respectively;

[0043] The user side also aims to maximize its own benefits and establishes an objective function:

[0044] ,

[0045] Where, User satisfaction index, It is the payment indicator for energy purchase.

[0046] Furthermore, in step S5, the hierarchical differential evolution algorithm is introduced as a solution method and game solving is performed to generate an optimized scheduling result. Specifically, the optimization problem is decomposed into two independent but interactive optimization levels, the upper level and the lower level. The upper level includes ETSP, and the lower level includes ES and USER. Each level uses the differential evolution algorithm and calls the CPLEX solver for solution.

[0047] The beneficial effects of the present invention compared to the prior art are:

[0048] The invention discloses a method for optimizing and scheduling an integrated energy system containing distributed energy resources based on a green certificate-tiered carbon trading mechanism. Currently, few integrated energy system optimization and scheduling models consider the impact of the correlation between the green certificate-tiered carbon trading mechanism and distributed energy resources on the absorption of new energy in the integrated energy system. The method for optimizing and scheduling an integrated energy system containing distributed energy resources based on a green certificate-tiered carbon trading mechanism provided by the present invention takes into account the correlation characteristics between distributed energy resources and the green certificate-tiered carbon trading mechanism, and verifies the effectiveness of the model under a master-slave game scheduling framework. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 The regional integrated energy system structure constructed for the present invention;

[0050] Figure 2 Schematic diagram of the two-layer differential evolution algorithm;

[0051] Figure 3 This is a schematic diagram of the multi-agent master-slave game framework;

[0052] Figure 4 To produce a landscape scene diagram that takes into account correlation, Figure 4 (a) shows the wind speed scenario, and (b) shows the photovoltaic output scenario.

[0053] Figure 5 A line chart showing the carbon reduction capacity of each scenario;

[0054] Figure 6 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0055] Figure 6 This is a flow chart of a low-carbon scheduling method for a multi-agent integrated energy system considering the correlation between wind and solar power. The specific implementation method of the present invention is as follows:

[0056] The present invention considers the three energy forms of cooling, heating and electricity combined heat and cooling unit, and its structure is as follows Figure 1 As shown in the figure, the energy equipment contained in the system mainly includes photovoltaic (PV), wind turbine (WT), gas turbine (GT) and gas boiler (GB); the energy storage equipment mainly includes electric energy storage (EES), cold energy storage (CES) and heat energy storage (HES); the energy conversion equipment includes waste heat boiler (WHB), ice-storage air-condition (ISAC) and absorption refrigerator (AR).

[0057] The mathematical models of each unit are shown below.

[0058] A1: The output constraints and climbing constraints of the GT device are:

[0059] ,

[0060] Where, is the exhaust waste heat of GT at time t; is the electric power output by GT at time t; 、 They are the upper and lower limits of GT's climbing rate, 、 are the lower and upper limits of GT’s output power at time t, is the electric energy conversion coefficient of the GT device, is the waste coefficient of the GT device.

[0061] A2: GB device output constraints are:

[0062] ,

[0063] Where, is the output thermal power of GB at time t; is the amount of natural gas consumed by GB at time t; For GB efficiency, 、 They are the lower and upper limits of heat production of GB device respectively.

[0064] A3: The battery device charging and discharging constraints are:

[0065] ,

[0066] Where, 、 They are the charge and discharge status marks of the battery respectively; 、 are the battery charging and discharging power respectively; 、 They are the upper limits of battery charge and discharge power respectively; 、 They are the lower limits of battery charging and discharging power respectively.

[0067] Taking into account the energy loss and efficiency of charging and discharging, the energy storage state constraints of the energy storage device satisfy:

[0068] ,

[0069] Where, is the SOC value of the battery; 、 are the charging and discharging power of the battery respectively; 、 They are the lower and upper limits of SOC respectively.

[0070] During actual operation, the battery needs to meet the charge and discharge ramp rate constraints:

[0071] ,

[0072] Where, 、 and 、 They are the upper and lower limits of the climbing grade under the battery charging and discharging status respectively.

[0073] A4: Heat storage and release constraints of heat storage tank devices:

[0074] ,

[0075] Where, 、 They are the charge and discharge status marks of the battery respectively; 、 are the battery charging and discharging power respectively; 、 They are the upper limits of battery charge and discharge power respectively; 、 They are the lower limits of battery charging and discharging power respectively.

[0076] Taking into account the energy loss and efficiency of charging and discharging thermal energy, the energy storage state constraints of the thermal storage device satisfy:

[0077] ,

[0078] Where, is the heat storage value of the heat storage tank; 、 are the charging and discharging powers of the heat storage tank respectively; 、 are the lower and upper limits of the heat storage tank power respectively; is the energy self-loss rate of the heat storage tank.

[0079] During actual operation, the heat storage tank needs to meet the charging and discharging ramp rate constraints:

[0080] ,

[0081] Where, 、 and 、 They are the upper and lower limits of climbing under the battery charging and discharging state, is the charging power of the heat storage tank at time t, is the heat release power of the heat storage tank at time t.

[0082] A5: ISAC system cooling constraints:

[0083] This invention uses a parallel ISAC system. Cold storage can only be performed during electricity price valleys, and the refrigerator can simultaneously cool and store ice. Specific constraints are:

[0084] ,

[0085] ,

[0086] Where, 、 are the output cooling power and ice storage power of the refrigerator respectively; is the ice melting power of the cold storage tank; 、 They are respectively the refrigeration and cold storage mark positions of the refrigerator. 、 Respectively represent the minimum and maximum power of the refrigerator, It is the maximum ice melting power of the cold storage tank.

[0087] like Figure 2 As shown in the multi-agent master-slave game framework, the objective function of the upper-layer energy trading service is:

[0088] ,,

[0089] Where, The proceeds from energy sales to ETSP; The cost of energy purchase when ETSP interacts with ES; is the electricity purchase cost of ETSP interacting with the grid.

[0090] The revenue from energy sales is as follows:

[0091] ,

[0092] Where, 、 、 Respectively expressed as income from selling electricity, heat and cooling energy; 、 、 are the actual electricity, heating and cooling loads on the user side respectively; 、 、 are the prices of electricity, heating and cooling energy sold by ETSP respectively.

[0093] The energy purchase cost is expressed as follows:

[0094] ,

[0095] Where, 、 、 are the electricity, heating and cooling power purchased by ETSP from ES respectively; 、 、 represent the ETSP purchase price of electricity, heat, and cooling energy respectively; 、 、 Represents the cost of purchasing electricity, heating and cooling energy respectively.

[0096] To ensure the interests of all stakeholders, the energy purchase and sale prices of ETSP must meet the following constraints:

[0097] ,

[0098] Where, 、 They are respectively the electricity purchase and electricity sales prices when ETSP interacts with the external large power grid. 、 The objective function of energy suppliers in the lower tier of energy selling and purchasing prices formulated for ETSP is:

[0099] ,

[0100] Where, Profit from energy sales for ES; Operation and maintenance costs; 、 They are respectively the transaction costs of carbon trading and green certificates.

[0101] ,

[0102] Where, 、 、 These are the operation and maintenance costs of IES power generation, cooling, and heating power, respectively; 、 、 are the electricity, heating and cooling cost coefficients respectively.

[0103] The electrical, thermal, and cooling power output at ES time t satisfies the following expressions:

[0104] ,

[0105] Where, 、 、 are the electrical, heating and cooling power values ​​produced by the CCHP system respectively; is the heat generation power of GB, 、 They are the power of discharging and charging heat energy of the heat storage tank respectively; 、 They are energy storage discharge and charging power respectively; They are the power consumption of ISAC; 、 、 、 They are the cooling power and ice melting power output of ISAC respectively. 、 are the wind power and photovoltaic power consumption constraints at time t, respectively, as shown in the following formula:

[0106] ,

[0107] Where, 、 are the wind power and power curtailment power of distributed energy at time t, 、 are the predicted wind and solar power values ​​at time t respectively.

[0108] The user side also aims to maximize its own benefits and establishes an objective function:

[0109] ,

[0110] Where, User satisfaction index, It is the payment indicator for energy purchase. 、 The calculation formula is as follows:

[0111] ,

[0112] Where, ; 、 、 、 The present invention sets the user's preference coefficients for electricity, heat and cooling energy to be 、 、 、 .

[0113] The solution of the present invention is further illustrated below with a specific case.

[0114] Simulation examples and calculation parameters

[0115] Figure 1 The regional integrated energy system structure constructed by the present invention is based on the historical photovoltaic, wind power and load data of a certain region in eastern China and selects a certain industrial park as the unit parameters. The green certificate trading base price is designed to be 210 yuan / book, the carbon trading base price is 80 yuan / t, the carbon trading price growth rate is 0.25, and the carbon emission interval length is 50t.

[0116] Table 1 Parameters of each device

[0117]

[0118] This invention sets up four scenarios to verify that the green certificate-carbon joint trading mechanism and the wind-solar combined power scenario have the effect of reducing carbon emissions and increasing distributed energy consumption in the integrated energy system. Scenario 1: Considering the master-slave game framework, without considering the green certificate trading market, only considering the carbon trading market; Scenario 2: Considering the master-slave game framework, considering the carbon trading mechanism and the green certificate market trading mechanism; Scenario 3: Considering the master-slave game framework, considering the green certificate-carbon joint trading mechanism; Scenario 4: Considering the master-slave game framework, considering the green certificate-carbon joint trading mechanism and the wind-solar combined power scenario.

[0119] The present invention takes historical wind and solar load data as the basis, constructs a green certificate-ladder carbon trading joint trading market mechanism, sets energy trading service providers as upper-level leaders, energy suppliers and users as lower-level followers, establishes a multi-agent master-slave game scheduling mechanism, generates wind and solar joint output scenarios for distributed energy historical data, constructs decision-making models for various stakeholders including energy trading service providers, energy suppliers, and users, introduces a hierarchical differential evolution algorithm as a solution method and performs a solution to generate an optimized scheduling result, wherein the master-slave game framework is as follows: Figure 3 shown.

[0120] like Figure 4 As shown, the present invention adopts Frank-Copula function to describe the wind-solar joint output scenario and generates the illustrated scenarios based on the K-means clustering method.

[0121] The optimized scheduling benefits and GCT and CET costs of scenarios 1-4 in the present invention are shown in Table 2. It can be seen from the table that in scenario 1, the cost factor of the green certificate trading market is not included in the optimized scheduling process, which limits its effectiveness in promoting carbon emission reduction. In the context of scenario 2, when considering the GCT generated by the ES's consumption of new energy during operation, this mechanism effectively reduces the CET cost. In scenario 3, by comprehensively considering the interaction between GCT and CET, the efficiency of the joint market is significantly improved, which has a positive impact on the ES in the consumption of new energy. This strategy not only promotes the efficient use of new energy such as photovoltaic and wind power, but also achieves a reduction in CET costs by optimizing the cost structure. In the context of scenario 4, when the correlation between wind and light is taken into account, the fluctuation of new energy in the system can be reduced, and the new energy consumption rate can be increased, thereby achieving the goal of reducing carbon emissions and maximizing the benefits of each entity.

[0122] like Figure 5 As shown in the figure, the carbon emission cost and carbon emissions of each scenario. It can be seen from the figure that the changing trends of carbon emissions and carbon trading amount under different scenarios are plotted in the figure. The red and blue curves in the figure both show a downward trend to varying degrees, verifying that the GCT mechanism alone or the GCT-CET joint mechanism has a certain carbon reduction effect in the regional integrated energy system, and the wind and solar combined output characteristics have also been proven to have the ability to reduce carbon emissions and improve the new energy absorption capacity in the regional integrated energy system, which can stimulate the system's carbon reduction potential at a deeper level.

[0123] Table 2 Optimized scheduling benefits and GCT and CET costs for scenarios 1-4

[0124] Scenario 1 2 3 4 ETSP / yuan 29393 32181 33136 33282 ES / yuan 35094 35228 36340 38171 USER / yuan 20203 20816 20796 20433 ES carbon emission cost / yuan 967.12 868.29 784.46 742.98 Carbon emissions / t 8.659 7.835 7.598 7.501 Green certificate cost / yuan 0 -1572.6 -2099.7 -2217.3 Wind and solar fluctuation standard deviation / kW 112.3 112.3 112.3 87.8 Abandoned light rate / % 1.3 0 0 0 Wind curtailment rate / % 10.1 7.3 4.8 2.5

[0125] The above embodiments are only for illustrating the technical ideas of the present invention and cannot be used to limit the scope of protection of the present invention. Any changes made on the basis of the technical solutions in accordance with the technical ideas proposed by the present invention shall fall within the scope of protection of the present invention; any technologies not covered by the present invention may be implemented by existing technologies.

Claims

1. A low-carbon scheduling method for a multi-agent integrated energy system considering the correlation between wind and solar power, characterized in that: The method comprises the following steps: S1: Establish an integrated energy system structure and construct an integrated energy system energy trading process including a green certificate-tiered carbon trading joint trading market mechanism; S2: In the integrated energy system structure, energy trading service providers are set as upper-level leaders, energy suppliers and users are set as lower-level followers, and a multi-agent master-slave game scheduling mechanism is established; S3: Generate wind-solar combined output scenarios based on historical distributed energy data, and introduce wind-solar correlation into the integrated energy system structure; S4: Build a decision-making model for all stakeholders in the integrated energy system, including energy trading service providers, energy suppliers, and users; S5: Introducing the hierarchical differential evolution algorithm as a solution method and performing game-based solutions to the decision-making models of various stakeholders in different scenarios to generate optimized scheduling results; The specific method of step S3 is to use the kernel density estimation method to estimate the marginal distribution functions of wind and solar output in different periods of time based on historical data, and use the Copula function to model the wind-solar joint distribution function. Finally, random sampling is performed through inverse transformation to generate typical scenarios. The specific scenario generation steps are as follows: S3-1. Use kernel density estimation method to generate the probability density function of wind and solar output for each 24-hour period based on historical photovoltaic and wind power data; S3-2. Establish a joint probability distribution function of photovoltaic and wind power based on the Frank-Copula function for each time period; S3-3. Sample the photovoltaic and wind power probability distribution functions based on the Frank-Copula function for each time period, perform an inverse transformation on the joint probability distribution function based on the sampling results, and finally obtain the wind and photovoltaic outputs for each time period, that is, generate the wind and solar outputs considering the correlation; S3-4. Use the K-means clustering algorithm to cluster the sampling results to generate a wind-solar combined power scenario. The specific steps are as follows: randomly select k data points as the initial cluster centers; assign data points, and for each data point in the data set, assign it to the nearest cluster center, using Euclidean distance to measure distance; Update the cluster center, calculate the average value of all points in each cluster, and update the average value as the new cluster center; iterate and repeat the above steps until the cluster center no longer changes or the preset number of iterations is reached; S3-5. Verify the effectiveness of the wind-solar correlation for the integrated energy system model. At the same time, introduce wind curtailment, solar curtailment, and wind-solar fluctuation standard deviation as measurement indicators. The calculation formula for wind-solar fluctuation standard deviation is as follows: ， Where, is the standard deviation of wind and solar fluctuations, 、 are the photovoltaic and wind power outputs at time t respectively.

2. A low-carbon scheduling method for a multi-agent integrated energy system considering wind and solar correlation according to claim 1, characterized in that: The specific method for constructing the green certificate-ladder carbon trading joint trading market mechanism described in step S1 is as follows: First, we design the carbon trading mechanism CET, adopt a stepped carbon trading cost mechanism, and construct the carbon trading cost as shown below: : ， Where, is the carbon trading base price, is the growth rate of carbon trading price, is the length of the carbon emission interval, is the actual carbon emissions of the integrated energy system; Secondly, a green certificate trading mechanism GCT is designed, where the green certificate transaction cost calculation method is: ， Where, represents the transaction cost of green certificates, is the unit price of green certificate transaction, The green certificate quota coefficient required for the integrated energy system is is the electric load of the integrated energy system at time t, The quantitative coefficient of converting renewable energy power generation into green certificates, 、 is the output power of distributed photovoltaic and wind power at time t; Finally, a green certificate-ladder carbon trading mechanism is designed. Energy suppliers can obtain carbon emission permits in addition to green certificates. The calculation method for converting renewable energy into green certificates is as follows: ， Where, In order to absorb the carbon emission quota of new energy, is the conversion coefficient, the conversion coefficient The conversion coefficient is determined by the base price of the green certificate market and the carbon trading market. Expressed as .

3. The low-carbon scheduling method for a multi-agent integrated energy system considering wind and solar correlation according to claim 1 is characterized in that: The specific method of step S2 is to analyze the pursuit of optimal decisions by energy trading service provider ETSP, energy supplier ES and user USER based on Stackelberg game theory. In the multi-agent Stackelberg game framework, the master-slave game consists of three parts: the participant set N; the game strategy set of energy trading service provider ETSP, energy supplier ES and user USER. 、 、 ; The interests of energy trading service providers ETSP, energy suppliers ES and users USER 、 、 ; The Stackelberg game model is specifically expressed as follows: ， The set of game participants N includes ETSP, ES, and USER, then Expressed as: ， Where, For the leader ETSP; 、 They are follower ES and USER respectively.

4. The low-carbon scheduling method for a multi-agent integrated energy system considering wind and solar correlation according to claim 1 is characterized in that: The specific method of step S4 is to first design the objective function of each stakeholder. Among them, the objective function of the energy trading service takes the time-of-use electricity price in different time periods as a reference basis, and its benefit comes from the energy complementarity between other stakeholder groups. The objective function of the energy trading service is expressed as: ， Where, The proceeds from energy sales to ETSP; The cost of energy purchase when ETSP interacts with ES; The cost of electricity purchase for ETSP to interact with the grid; Energy suppliers establish an objective function based on the energy prices set by energy trading service providers and their own interests. The energy supplier objective function is as follows: ， Where, Profit from energy sales for ES; Operation and maintenance costs; 、 They are carbon trading and green certificate transaction costs respectively; The user side also aims to maximize its own benefits and establishes an objective function: ， Where, User satisfaction index, It is the payment indicator for energy purchase.

5. The low-carbon scheduling method for a multi-agent integrated energy system considering wind and solar correlation according to claim 1 is characterized in that: In step S5, the hierarchical differential evolution algorithm is introduced as a solution method and game solving is performed to generate an optimized scheduling result. Specifically, the optimization problem is decomposed into two independent but interactive optimization levels, the upper level and the lower level. The upper level includes ETSP, and the lower level includes ES and USER. Each level uses the differential evolution algorithm and calls the CPLEX solver for solution.

Images