Multi-virtual power plant hybrid game optimization method and system considering wind and light uncertainty

Abstract

The invention relates to the technical field of multi-virtual power plant optimization scheduling, and discloses a multi-virtual power plant mixed game optimization method and system considering wind and light uncertainty, and the method comprises the steps: building a joint model combining a master-slave game with a cooperative game based on the game relation of a power grid operator and a virtual power plant alliance; aiming at the output uncertainty of the wind and light generator set, through a data-driven two-stage distributed robust optimization method, a typical wind and light generator set output scene is selected, a discrete scene distribution value is calculated, the probability distribution of the wind and light generator set output scene is calculated in combination with historical data and a confidence coefficient formula, and a wind and light uncertainty probability model is constructed; based on the joint model and the wind-solar uncertainty probability model, the total cost and carbon emission of the virtual power plant alliance are reduced through collaborative optimization scheduling. According to the invention, the responsiveness to the carbon trading market is enhanced, and powerful support is provided for promoting low-carbon transformation and realizing the purposes of energy conservation and emission reduction.

Description

Technical Field

[0001] The present invention relates to an optimal scheduling method for multiple virtual power plants under multi-variety market transactions, belonging to the technical field of operation and control of power systems containing various energy forms. Background Art

[0002] As an independent entity that can aggregate distributed resources within a certain area, a virtual power plant (VPP) aggregates distributed new energy, energy storage, and various loads into the power system, and can fully exploit the flexibility on the source-load-storage side, thereby improving energy utilization efficiency.

[0003] In the traditional power market, transactions between power producers and grid operators are usually carried out through simple sales contract relationships. However, with the popularization of VPP technology, the relationships among market participants have become more complex. Multiple VPPs can be jointly operated through a cooperative alliance to achieve resource sharing and complementarity, thereby reducing costs and improving overall efficiency, and forming stronger competitiveness in the market. Under this cooperation model, energy interaction needs to be coordinated among the members of the virtual power plant alliance, and at the same time, they can also play games with grid operators to optimize the purchase and sale prices of electricity. Therefore, a master-slave game is formed between the VPP alliance and the grid operator, and a cooperative game is formed within the alliance. Moreover, incorporating a carbon trading mechanism and a green certificate trading mechanism into the VPP can reduce the carbon emissions of the VPP. However, due to the presence of new energy power generation units within the VPP, their output has volatility and randomness, which in turn leads to inaccurate calculation of green certificate costs and inability to obtain market-adapted cost results, making the scheduling decision of the VPP complicated, that is, the existing technology does not fully consider the impact of wind-solar uncertainty on VPP decision-making.

[0004] Currently, Xu Huihui et al. proposed a hybrid game optimal scheduling method for multiple virtual power plants considering green certificate-carbon trading. First, a hybrid game optimization framework was established to analyze the game relationship between grid operators and virtual power plant alliances. Then, a hybrid game optimization model with grid operators as leaders and virtual power plants as followers was constructed, and based on the Nash bargaining theory, minimizing the cost of virtual power plants was equivalent to two sub-problems: maximizing the alliance benefit and allocating cooperative benefits. Finally, the bisection method and the alternating direction multiplier method were combined to solve the hybrid game model. However, this method has the following limitations: 1. In terms of wind-solar uncertainty. When analyzing the output of wind-solar power generation units, this method only considers the constraint conditions of the maximum and minimum outputs, and does not involve the inherent uncertainty factors of wind-solar power generation. In fact, due to the influence of spatio-temporal differences, the output of wind-solar power generation units has volatility and unpredictability, and this output imbalance will cause deviations in calculating costs, and ultimately affect the calculation accuracy of the objective function, resulting in a difference between the optimization result and the actual working conditions.

[0005] 2. In terms of the green certificate trading market, the method adopts a fixed green certificate price, which fails to reflect the dynamic impact of market supply and demand changes on prices, and does not conform to the price fluctuation characteristics of the actual green certificate market. At the same time, it does not consider the difference in the time value of green certificate trading, lacks the distinction between current selling and future selling strategies, and leads to the inability to optimize intertemporal benefits. The relationship between the green certificate supply and the renewable energy output is too simplified, only established through the liability weight coefficient (β), and fails to fully reflect the direct impact of the volatility of wind and solar power generation on the green certificate supply. These defects make the model have obvious deficiencies in aspects such as price formation mechanism, time dimension, and market interaction.

[0006] 3. In terms of the carbon trading model, the linear calculation method of the carbon trading cost model in this method is too simplified. It neither considers the market volatility of carbon prices, nor sets the carbon emission quota and coefficient as fixed values. At the same time, the carbon capture volume is disconnected from the trading mechanism. Summary of the Invention

[0007] To solve the above problems, the present invention proposes a multi-virtual power plant hybrid game optimization method and system considering wind and solar uncertainties. Multiple VPPs form a cooperative alliance to form a master-slave game with the grid operator, optimize the power purchase and sale prices of the alliance from the operator and the power interaction prices among the alliance members, and use the Nash bargaining theory to allocate the cooperative benefits of the alliance, which is equivalent to two sub-problems of maximizing the alliance benefits and allocating the cooperative benefits. In addition, considering that the VPP contains renewable energy with volatility, wind and solar uncertainties are considered in the model, and a distributionally robust optimization (DRO) method under comprehensive norm constraints is proposed.

[0008] The technical solution adopted by the present invention is as follows: A multi-virtual power plant hybrid game optimization method considering wind and solar uncertainties, including: Based on the game relationship between the grid operator and the virtual power plant alliance, a joint model combining the master-slave game and the cooperative game is constructed; Aiming at the uncertainty of the output of wind and solar power generation units, through a data-driven two-stage distributed robust optimization method, typical output scenarios of wind and solar power generation units are selected and the discrete scenario distribution values are calculated. Combining historical data and the confidence formula, the probability distribution of the output scenarios of wind and solar power generation units is calculated, and a wind and solar uncertainty probability model is constructed; Based on the joint model and the wind and solar uncertainty probability model, the total cost and carbon emissions of the virtual power plant alliance are reduced through collaborative optimal scheduling.

[0009] Further, the constructing a joint model combining the master-slave game and the cooperative game based on the game relationship between the grid operator and the virtual power plant alliance includes: Establish a leader - grid operator model in the master - slave game, with the maximization of the grid operator's own revenue as the objective function; Establish a follower - virtual power plant alliance model in the master - slave game, with the minimization of the virtual power plant alliance members' own costs as the objective function; Establish a Nash bargaining model of the virtual power plant alliance based on cooperative game, and respond to the grid operator's decision through cooperation, with the maximization of the overall benefit as the objective function.

[0010] Furthermore, in the virtual power plant alliance model, the virtual power plant alliance members' own costs at least include carbon trading costs, and the calculation method of the carbon trading costs includes:

[0011] Among them, is the carbon quota of the unit, T is the total number of time periods, n is the number of virtual power plants participating in the game, is the output electric power of the unit of the virtual power plant at time t, is the power supply reference value, is the heating correction coefficient; is the actual carbon emission of the unit, is the carbon emission factor of the unit; is the heating ratio; is the carbon trading cost, is the carbon trading price.

[0012] Furthermore, in the virtual power plant alliance model, the virtual power plant alliance members' own costs at least include green certificate trading costs, and the calculation method of the green certificate trading costs includes:

[0013] Among them, is the green certificate trading cost; is the current green certificate price, is the future green certificate price; is the number of green certificates that can be sold, is the number of green certificates sold currently, is the number of green certificates sold in the future; T is the total number of time periods; is the estimated price ratio of green certificates in the future market.

[0014] Furthermore, in the virtual power plant alliance model, the constraint conditions at least include green certificate constraints, and the green certificate constraints include:

[0015] Among them, is the green certificate price in the market, is the initial green certificate price; and are two positive parameters of the inverse function of the Cournot model price; is the number of green certificates that can be sold on the day; is the green certificate trading price ratio coefficient calculated based on historical data; is the new energy consumption ratio in this region; is the total number of green certificates in a day; is t the output of the wind turbines of the virtual power plant at time is t the output of the photovoltaic panels of the virtual power plant at time is the dispatching unit duration; is the number of green certificates currently sold, is the number of green certificates to be sold in the future.

[0016] Furthermore, for the uncertainty of the output of wind and light generating units, a data-driven two-stage distributed robust optimization method is used to select typical output scenarios of wind and light generating units and calculate the discrete scenario distribution values, including: Select M typical output scenarios from the actually obtained X output samples of wind and light generating units by the K-means algorithm to represent the output uncertainty of wind and light generating units, and obtain the discrete scenario probability distribution values of the output of each wind and light generating unit ( k = 1, 2, …, X ); Construct a feasible region of probability distribution values centered on the discrete scenario probability distribution value with the 1-norm and ∞-norm sets as constraint conditions Ω :

[0017] wherein, is the actual value of the i th probability distribution; is the probability error value under the 1-norm distribution, is the probability error value under the ∞-norm distribution.

[0018] Furthermore, the calculation of the probability distribution of the output scenarios of wind and light generating units by combining historical data and confidence formulas includes: The probability of the output scenarios of wind and light generating units needs to meet the following inequality requirements:

[0019] wherein, is The probability that the inequality holds is the initial probability value selected from historical data M is the number of output samples of wind-solar generating units Let the right sides of the inequality signs in the above two equations be equal to the confidence levels that can make them hold and , respectively, to obtain

[0020] Thus, the probability distribution of the output scenarios of wind-solar generating units is obtained

[0021] Furthermore, based on the joint model and the wind-solar uncertainty probability model, the total cost and carbon emissions of the virtual power plant alliance are reduced through collaborative optimal scheduling, including According to the Nash bargaining model of the virtual power plant alliance based on cooperative game in the joint model, combined with the wind-solar uncertainty probability model for collaborative optimal scheduling, a hybrid game model is obtained

[0022] wherein x is the transaction decision variable in the Nash bargaining model of the virtual power plant alliance y is the adjustable decision variable is the power purchase and sale cost of the virtual power plant alliance, and the superscript represents the transpose operation is the operating cost corresponding to the probability distribution of the worst-case scenario a, b, c, d, g, w, A, G, E, F, U, V is the coefficient matrix is the mathematical representation of wind-solar uncertainty

[0023] Furthermore, for the hybrid game model, the master-slave game model is solved iteratively by the bisection method; the cooperative game model, that is, the Nash bargaining model of the virtual power plant alliance, is solved by the CCG-ADMM algorithm (Constrained Generation - Alternating Direction Method of Multipliers)

[0024] A multi-virtual power plant hybrid game optimization system considering wind-solar uncertainty includes A joint model construction module configured to construct a joint model combining master-slave game and cooperative game based on the game relationship between the grid operator and the virtual power plant alliance A wind-solar uncertainty probability model construction module configured to, for the output uncertainty of wind-solar generating units, select typical output scenarios of wind-solar generating units and calculate discrete scenario distribution values through a data-driven two-stage distributed robust optimization method, calculate the probability distribution of the output scenarios of wind-solar generating units in combination with historical data and the confidence formula, and construct a wind-solar uncertainty probability model The collaborative optimization scheduling module is configured to reduce the total cost and carbon emissions of the virtual power plant alliance through collaborative optimization scheduling based on the joint model and the wind-solar uncertainty probability model.

[0025] The beneficial effects of the present invention are as follows: The present invention combines the game relationship between the power grid operator and the virtual power plant alliance to construct an optimization framework that combines the master-slave game and the cooperative game; considering the uncertainty of new energy sources such as wind and solar in the virtual power plant alliance, a data-driven DRO model is proposed. Through optimized scheduling, the total cost and carbon emissions of the virtual power plant alliance can be effectively reduced. The present invention uses the Nash bargaining theory to fairly distribute the cooperative benefits, promotes the efficient cooperation of the virtual power plant alliance, and realizes the reasonable distribution of interests. Finally, while reducing costs, the members of the virtual power plant alliance enhance their response ability to the carbon trading market, providing strong support for promoting the low-carbon transformation and achieving the energy conservation and emission reduction goals.

[0026] Compared with the prior art, the present invention has the following advantages: 1. In terms of wind-solar uncertainty. The present invention adopts a data-driven two-stage distributionally robust optimization (DRO) method for two typical distributed energy sources, namely wind turbines and photovoltaic units. First, the K-means clustering algorithm is used to select X typical output scenarios from a number of actual wind-solar output samples to represent the uncertainty of wind-solar output. Through these typical scenarios, the discrete scenario probability distribution values corresponding to each wind-solar output can be obtained. Subsequently, based on the probability distribution values screened from historical data and combined with the confidence formula, the probability distribution of the wind-solar output scenario is further calculated. Based on the above results, the minimum operating cost of the alliance in the worst-case scenario can be calculated.

[0027] 2. In terms of the green certificate trading market. In the present invention, the virtual power plant (VPP) can choose to sell the green certificates obtained through new energy power generation on the same day or at a certain future time. The relationship between the price and output of the green certificates satisfies the inverse demand function relationship in economics. In the process of pursuing the maximum benefit of the VPP, the Cournot model based on the quantity competition theory is adopted to solve the quantity of green certificates sold on the same day and in the future. Under this model, the decision-making of the quantity of green certificates sold can reflect the game between market supply and demand, thus realizing the optimization of the overall benefit.

[0028] 3. In terms of the carbon trading model. The present invention dynamically adjusts the carbon emissions of combined heat and power through the heating correction coefficient and the heating ratio, accurately distinguishes the calculation of carbon quotas and actual emissions, and establishes a market-based trading framework based on the real-time carbon price. At the same time, the unit differential modeling is realized by using the power supply benchmark value and the emission factor, comprehensively improving the accuracy and market applicability of the model. Description of the Drawings

[0029] Figure 1It is a flowchart of a multi-virtual power plant hybrid game optimization method considering the uncertainty of wind and light in Embodiment 1 of the present invention. Detailed implementation manners

[0030] In order to have a clearer understanding of the technical features, objectives, and effects of the present invention, the specific implementation manners of the present invention will now be described. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention, that is, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0031] Embodiment 1 As Figure 1 shown, this embodiment provides a multi-virtual power plant hybrid game optimization method considering the uncertainty of wind and light, including: Based on the game relationship between the power grid operator and the virtual power plant alliance, a combined model of master-slave game and cooperative game is constructed; Regarding the uncertainty of the output of wind and light generating units, through a data-driven two-stage distributed robust optimization method, typical output scenarios of wind and light generating units are selected and the discrete scenario distribution values are calculated, and the probability distribution of the output scenarios of wind and light generating units is calculated in combination with historical data and the confidence formula, and a probability model of wind and light uncertainty is constructed; Based on the combined model and the probability model of wind and light uncertainty, the total cost and carbon emissions of the virtual power plant alliance are reduced through collaborative optimal scheduling.

[0032] Preferably, based on the game relationship between the power grid operator and the virtual power plant alliance, a combined model of master-slave game and cooperative game is constructed, including: Establish a leader - power grid operator model in the master-slave game, with the maximum profit of the power grid operator itself as the objective function; Establish a follower - virtual power plant alliance model in the master-slave game, with the minimum cost of the members of the virtual power plant alliance themselves as the objective function; Establish a Nash bargaining model of the virtual power plant alliance based on cooperative game, and respond to the decision of the power grid operator through cooperation, with the maximum overall benefit as the objective function.

[0033] Specifically, the multi-virtual power plant hybrid game optimization method of this embodiment can be implemented by the following steps: (1) Establish a leader - power grid operator model in the master-slave game.

[0034] (1-1) Objective function: With the maximum profit of the power grid operator itself as the goal, specifically including the costs and revenues of power quantity transactions with the superior power grid and the VPP alliance.

[0035] (1) In the formula, , are respectively t the grid electricity price and the feed-in tariff at time , are respectively t the electricity purchase and sale prices of the VPP alliance from / to the grid operator at time , are respectively t the electricity purchase and sale volumes of the VPP alliance from / to the grid operator at time T is the total number of time periods.

[0036] Among them, the electricity purchase and sale volume of the VPP alliance is: (2) (3) In the formula, , are respectively the electricity purchase and sale volumes of the VPP i from / to the grid operator at time t ; n is the number of VPPs participating in the game.

[0037] Constraints (1 - 2).

[0038] The electricity purchase and sale prices set by the grid operator should be within a certain range: (4) (5) In the formula: , are respectively the upper and lower limits of the electricity purchase price of the VPP alliance; , are respectively the upper and lower limits of the electricity sale price of the VPP alliance.

[0039] To maximize its own benefits, the grid operator will set the electricity purchase price of the alliance to the highest and the electricity sale price of the alliance to the lowest. To avoid this problem, the average value constraint of the alliance's electricity purchase and sale prices is set as: (6) (7) In the formula, , are respectively the average values of the VPP's electricity purchase and sale prices.

[0040] (2) Establish the follower - VPP alliance model in the master - slave game.

[0041] It should be noted that in this embodiment, a common VPP is taken as an example for illustration. This VPP mainly includes wind turbines (WT), photovoltaic units (PV), combined heat and power units (CHP), gas boilers (GB), electrical energy storage (EES), and thermal energy storage tanks (TES).

[0042] (2-1) Objective function: The members of the VPP alliance aim to minimize their own costs, specifically including the costs of buying and selling electricity, the interaction costs between VPPs, carbon trading costs, gas costs, demand response costs, and energy storage operation and maintenance costs: (8) In the formula, and are the costs of buying and selling electricity, the interaction costs between VPPs, carbon trading costs, green certificate trading costs, gas purchase costs, demand response costs, and energy storage operation and maintenance costs respectively.

[0043] (2-1-1) Follower's cost of buying and selling electricity: (9) (2-1-2) Interaction costs between VPPs: (10) Among them, is t the electricity trading price between VPP i and VPP j at time ; t is i the electricity trading volume between VPP j and VPP

[0044] at time

[0045] In the carbon market, VPP can obtain a certain amount of carbon quotas according to the type and installed capacity of power generation equipment, and then choose to buy or sell carbon emission rights from the carbon market in combination with the actual carbon emissions.

[0046] (11) (12) (13) (14) Among them, , They are the carbon quota of the unit and the actual carbon emissions of the unit respectively; is the CHP output electric power of VPPi at time t; is the power supply reference value; is the heating correction coefficient; is the heating ratio; is the carbon emission factor of the unit; is the carbon trading cost, is the carbon trading price.

[0047] (2-1-4) Green certificate trading cost.

[0048] In the green certificate market, the VPP can choose to sell the green certificates obtained through new energy power generation on the same day or in the future on the same day.

[0049] (15) (16) (17) Among them, is the green certificate trading cost; , are the numbers of green certificates sold currently and in the future respectively; , are the green certificate prices currently and in the future respectively; is the estimated price ratio of green certificates in the future market; is the number of green certificates that can be sold.

[0050] (2-1-5) Gas purchase cost: (18) Among them, is the unit gas purchase cost, are the gas consumption of CHP and GB of VPPi at time t respectively.

[0051] (2-1-6) Demand response cost: (19) Among them, is the electric load transfer cost, are the electric load curtailment cost and the heat load curtailment cost respectively; are respectively t time VPP i can curtail and transfer the electric load, is t time VPP i can curtail the heat load.

[0052] (2-1-7) Energy storage operation and maintenance cost: (20) Among them, are the EES charge and discharge powers of VPPi at time t, respectively; are the TES heat charge and discharge powers of VPPi at time t, respectively; are the operation and maintenance costs of electrical energy storage and thermal energy storage, respectively.

[0053] (2-2) Constraint conditions.

[0054] (2-2-1) The electrical energy trading constraint of P2P is: (21) (22) In the formula, is the maximum interactive electricity quantity between VPPs; is the electricity quantity sold by VPPi to VPPj, is the electricity quantity sold by VPPj to VPPi.

[0055] (2-2-2) The demand response constraint is:

[0056] In the formula, and are the upper limits of the reducible and transferable electrical loads, respectively; is the upper limit of the reducible thermal load.

[0057] (2-2-3) The power balance constraint is:

[0058] In the formula: , are the electricity purchase and sale quantities of VPP i from / to the grid operator at time t respectively; is the fan output power of the virtual power plant at time t , is the PV output power of the virtual power plant at time t ; is the CHP output electrical power of VPPi at time t , are the EES charge and discharge powers of VPPi at time t, respectively, is the electrical load of VPPi at time t ; are the numbers of electric vehicles for discharging and charging, respectively; are the discharging and charging quantities of the nth electric vehicle of VPP i at time t, respectively; respectively t the electric boiler output heat power, CHP output heat power at time VPPi, t the heat load at time VPPi, respectively the TES charge and discharge power of VPPi at time t.

[0059] (2-2-4)The green certificate constraint is:

[0060] where, is the green certificate price in the market; is the initial green certificate price; and are two positive parameters of the price inverse function of the Cournot model; is the number of green certificates that can be sold on the day; is the green certificate trading price ratio coefficient calculated based on historical data; is the new energy consumption ratio in this region; is the total number of green certificates in a day; is the dispatching unit duration, which can be set to 1h.

[0061] (3)Establish the VPP alliance Nash bargaining model.

[0062] The VPP alliance responds to the decisions of the grid operator through cooperation, and the cooperation goal is to maximize the overall benefit. Preferably, the VPP alliance Nash bargaining model is: (34) In the formula: is the benefit obtained by VPP i participating in the negotiation, is the benefit obtained by VPP i not participating in the negotiation.

[0063] The VPP alliance Nash bargaining model is a non-convex and non-linear problem with multi-variable coupling. Therefore, the model is converted into a VPP alliance cost minimization sub-problem (P1) and a benefit distribution sub-problem (P2), and solved sequentially.

[0064] (3-1)Sub-problem (P1) - VPP alliance cost minimization: (35) (3-2)Sub-problem (P2) - Cooperative benefit distribution: (36) In the formula: is the benefit of VPP obtained in sub-problem P1 i benefit; is t at time VPPi and VPP j electricity trading price is t the VPP at time i and VPP j trading electricity volume; , are the electricity purchase and sale prices at time t obtained in sub - problem P1 respectively.

[0065] (4) Establish a data - driven DRO model.

[0066] In this embodiment, the K - means algorithm is used to select M from the actually obtained X wind and solar power output samples ( k = 1, 2, …, X ) typical output scenarios to represent the output uncertainty of wind and solar power accordingly, and obtain the discrete scenario probability distribution values of each wind and solar power output Since the obtained initial probability distribution is still uncertain, a feasible region of probability distribution values centered on the discrete probability distribution value Ω is constructed with the 1 - norm and ∞ - norm sets as constraint conditions: (37) In the formula, is the actual value of the i th probability distribution; , are the probability error values under the 1 - norm and ∞ - norm distributions respectively.

[0067] The probability of wind and solar power output scenarios satisfies the following inequality requirements:

[0068] In the formula: is the probability that the inequality in holds; and are the initial probability values screened from historical data. Let the right - hand sides of the above two inequalities be equal to the confidence levels (40) Preferably, the Nash bargaining model is combined with the constructed wind and solar uncertainty probability model for collaborative optimal scheduling to obtain the following mixed - game model in matrix form: (41) In the formula: x is the trading decision variable in the Nash bargaining model,y is the adjustable decision variable including renewable energy output, output of equipment within the VPP alliance, etc.; is the electricity purchase and sale cost of the VPP alliance; is the operating cost corresponding to the probability distribution of the worst-case scenario; a, b, c, A, G, E, F, U, V, g, d, w is the coefficient matrix, is the mathematical representation of the uncertainty of wind and light.

[0069] Preferably, after constructing the above hybrid game model, the master-slave game model can be solved iteratively by using the bisection method, and the cooperative game model can be solved by using the CCG-ADMM algorithm (constraint generation - alternating direction method of multipliers).

[0070] In summary, the method of this embodiment combines the game relationship between the power grid operator and the virtual power plant alliance, and constructs an optimization framework combining the master-slave game and the cooperative game. For the uncertainty of the output of wind and light generating units, through the data-driven two-stage distributed robust optimization method, typical new energy unit output scenarios are selected and the discrete scenario distribution values are calculated. Combining historical data and the confidence formula, the uncertainty of wind and light is accurately characterized to ensure that the alliance can still minimize the operating cost under the worst-case scenario. On this basis, the total cost and carbon emissions of the virtual power plant alliance are effectively reduced through optimal scheduling. This method further uses the Nash bargaining theory to fairly distribute the cooperative benefits, which not only promotes the efficient cooperation of the virtual power plant alliance, but also realizes the reasonable distribution of interests. Finally, while reducing costs, the alliance members significantly improve their response ability to the carbon trading market, providing strong support for promoting the low-carbon transformation and achieving the energy conservation and emission reduction goals.

[0071] Embodiment 2 This embodiment is based on Embodiment 1: This embodiment provides a multi-virtual power plant hybrid game optimization system considering the uncertainty of wind and light, including: A joint model construction module configured to construct a joint model combining the master-slave game and the cooperative game based on the game relationship between the power grid operator and the virtual power plant alliance; A wind and light uncertainty probability model construction module configured to, for the uncertainty of the output of wind and light generating units, select typical wind and light generating unit output scenarios and calculate discrete scenario distribution values through a data-driven two-stage distributed robust optimization method, calculate the probability distribution of the wind and light generating unit output scenarios in combination with historical data and the confidence formula, and construct a wind and light uncertainty probability model; A collaborative optimal scheduling module configured to reduce the total cost and carbon emissions of the virtual power plant alliance through collaborative optimal scheduling based on the joint model and the wind and light uncertainty probability model.

[0072] Embodiment 3 This embodiment is based on Embodiment 1: This embodiment provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the multi-virtual power plant hybrid game optimization method considering wind and light uncertainties in Embodiment 1. Among them, the computer program can be in the form of source code, object code, executable file, or some intermediate form, etc.

[0073] Embodiment 4 This embodiment is based on Embodiment 1: This embodiment provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, it implements the multi-virtual power plant hybrid game optimization method considering wind and light uncertainties in Embodiment 1. Among them, the computer program can be in the form of source code, object code, executable file, or some intermediate form, etc. The storage medium includes: any entity or device capable of carrying computer program code, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the storage medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the storage medium does not include electrical carrier signals and telecommunication signals.

[0074] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are expressed as a series of action combinations. However, those skilled in the art should know that this application is not limited by the described action sequence, because according to this application, some steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential for this application.

Claims

1. A multi-virtual power plant hybrid game optimization method considering the uncertainty of wind and light, characterized in that Including: Based on the game relationship between the power grid operator and the virtual power plant alliance, a combined model integrating master-slave game and cooperative game is constructed; Aiming at the output uncertainty of wind and photovoltaic generating units, through a data-driven two-stage distributed robust optimization method, typical output scenarios of wind and photovoltaic generating units are selected and the discrete scenario distribution values are calculated. Combining historical data and the confidence formula, the probability distribution of the output scenarios of wind and photovoltaic generating units is calculated, and a probability model of wind and photovoltaic uncertainty is constructed; Based on the combined model and the probability model of wind and photovoltaic uncertainty, the total cost and carbon emissions of the virtual power plant alliance are reduced through collaborative optimal scheduling.

2. The multi-virtual power plant hybrid game optimization method considering the uncertainty of wind and light according to claim 1, characterized in that The combined model integrating master-slave game and cooperative game constructed based on the game relationship between the power grid operator and the virtual power plant alliance includes: Establish a leader - power grid operator model in the master-slave game, with the maximum profit of the power grid operator itself as the objective function; Establish a follower - virtual power plant alliance model in the master-slave game, with the minimum cost of the members of the virtual power plant alliance themselves as the objective function; Establish a Nash bargaining model of the virtual power plant alliance based on cooperative game, and respond to the decision of the power grid operator through cooperation, with the maximum overall benefit as the objective function.

3. The multi-virtual power plant hybrid game optimization method considering wind and light uncertainties according to claim 2, characterized in that, In the virtual power plant alliance model, the cost of the members of the virtual power plant alliance themselves at least includes the carbon trading cost, and the calculation method of the carbon trading cost includes: Among them, is the carbon quota of the unit, T is the total number of time periods, n is the number of virtual power plants participating in the game, is the output electric power of the unit of the virtual power plant at time t, is the power supply reference value, is the heating correction coefficient; is the actual carbon emission of the unit, is the carbon emission factor of the unit; is the heating ratio; is the carbon trading cost, is the carbon trading price.

4. The multi-virtual power plant hybrid game optimization method considering the uncertainty of wind and light according to claim 2, characterized in that, In the virtual power plant alliance model, the cost of the members of the virtual power plant alliance themselves at least includes the green certificate trading cost, and the calculation method of the green certificate trading cost includes: Among them, is the cost of green certificate trading; is the current green certificate price, is the future green certificate price; is the number of green certificates that can be sold, is the number of currently sold green certificates, is the number of future sold green certificates; T is the total number of time periods; is the estimated price ratio of green certificates in the future market.

5. The multi-virtual power plant hybrid game optimization method considering the uncertainty of wind and light according to claim 2, characterized in that In the virtual power plant alliance model, the constraint conditions at least include the green certificate constraint, and the green certificate constraint includes: Among them, is the green certificate price in the market, is the initial green certificate price; and are two positive parameters of the inverse function of the Cournot model price; is the number of green certificates that can be sold on the day; is the green certificate trading price ratio coefficient calculated based on historical data; is the new energy consumption ratio in this region; is the total number of green certificates in a day; is t the fan output of the virtual power plant at time is t the PV output of the virtual power plant at time is the dispatching unit duration; is the number of green certificates currently sold, is the number of green certificates to be sold in the future.

6. The hybrid game optimization method for multi-virtual power plants considering the uncertainty of wind and light according to claim 1, characterized in that Regarding the output uncertainty of wind and photovoltaic generating units, through a data-driven two-stage distributed robust optimization method, typical output scenarios of wind and photovoltaic generating units are selected and the discrete scenario distribution values are calculated, including: Select X typical output scenarios from the actually obtained M output samples of wind-solar generating units through the K-means algorithm to represent the output uncertainty of wind-solar generating units, and obtain the discrete scenario probability distribution values of the outputs of each wind-solar generating unit M ( X = 1, 2, …, ) k X ) Construct a feasible region of probability distribution values centered on the discrete scenario probability distribution values and bounded by the 1-norm and ∞-norm sets Ω : Among them, is the actual value of the i th probability distribution; is the probability error value under the 1-norm distribution, is the probability error value under the ∞-norm distribution.

7. The multi-virtual power plant hybrid game optimization method considering the uncertainty of wind and light according to claim 6, characterized in that, The calculation of the probability distribution of the output scenarios of wind and photovoltaic generating units by combining historical data and the confidence formula includes: The probability of the output scenarios of wind and photovoltaic generating units needs to meet the following inequality requirements: Among them, is the probability that the inequality in is the initial probability value screened from historical data, M is the number of output samples of the wind-solar generating unit; Let the right sides of the above two inequalities be equal to the confidence levels that can make them hold respectively and , and we get: Thus, the probability distribution of the output scenarios of wind and photovoltaic generating units is obtained.

8. The multi-virtual power plant hybrid game optimization method considering the uncertainty of wind and light according to claim 7, characterized in that, Based on the combined model and the probability model of wind and photovoltaic uncertainty, the total cost and carbon emissions of the virtual power plant alliance are reduced through collaborative optimal scheduling, including: According to the Nash bargaining model of the virtual power plant alliance based on cooperative game in the combined model, combined with the probability model of wind and photovoltaic uncertainty, collaborative optimal scheduling is carried out to obtain a hybrid game model: Among them, x is the transaction decision variable in the Nash negotiation model of the virtual power plant alliance, y is the adjustable decision variable; is the power purchase and sale cost of the virtual power plant alliance, and the superscript represents the transpose operation; is the operating cost corresponding to the probability distribution of the worst scenario; a, b, c, d, g, w, A, G, E, F, U, V is the coefficient matrix, is the mathematical representation of the uncertainty of wind and light.

9. The multi-virtual power plant hybrid game optimization method considering the uncertainty of wind and light according to claim 8, characterized in that, For the hybrid game model, the master-slave game model is solved iteratively by the bisection method; the cooperative game model, that is, the Nash bargaining model of the virtual power plant alliance, is solved by the constraint generation - alternating direction multiplier method.

10. A multi-virtual power plant hybrid game optimization system considering the uncertainty of wind and light, characterized in that, Including: A combined model construction module, configured to construct a combined model integrating master-slave game and cooperative game based on the game relationship between the power grid operator and the virtual power plant alliance; The wind-solar uncertainty probability model construction module is configured to select typical wind-solar generator output scenarios and calculate discrete scenario distribution values for the output uncertainty of wind-solar generating units through a data-driven two-stage distributed robust optimization method, calculate the probability distribution of wind-solar generating unit output scenarios in combination with historical data and a confidence formula, and construct a wind-solar uncertainty probability model; The collaborative optimization scheduling module is configured to reduce the total cost and carbon emissions of the virtual power plant alliance through collaborative optimization scheduling based on the joint model and the wind-solar uncertainty probability model.

Images