A multi-virtual power plant point-to-point transaction method, electronic equipment and storage medium

Abstract

The application discloses a kind of multi-virtual power plant point-to-point transaction method, electronic equipment and storage medium.First, collect and aggregate virtual power plant internal distributed resources, construct multi-virtual power plant electricity-carbon-backup point-to-point transaction model;Set conditional value at risk model confidence and risk preference coefficient, construct multi-virtual power plant electricity-carbon-backup point-to-point transaction model based on conditional value at risk;Adaptive alternating direction multiplier algorithm (ADMM) based on consensus is used to solve the model in distributed manner, and the transaction strategy of each virtual power plant is obtained.The application considers that virtual power plant aggregates photovoltaic, fuel cell, energy storage, central air conditioning, flexible load and other distributed resources to participate in electricity-carbon-backup point-to-point transaction, which widens the transaction channel of virtual power plant;Value at risk quantifies the potential risk brought by photovoltaic randomness, so that virtual power plant decision makers can weigh benefits and risks;Adaptive ADMM algorithm based on consensus effectively improves the calculation efficiency of model solving.

Description

Technical Field

[0001] This invention belongs to the field of power system dispatching and optimization, and specifically relates to a multi-virtual power plant point-to-point trading method, electronic equipment, and storage medium. Background Technology

[0002] To achieve low-carbon goals and build a new type of power system, the widespread integration of distributed resources has become the new normal in my country's power system. However, the small capacity, large number, and geographically dispersed nature of distributed resources make them difficult for the power grid to manage effectively, resulting in problems such as difficulties in renewable energy consumption, low energy utilization, and insufficient power system security. To tap the regulation potential of massive distributed resources, virtual power plant technology has emerged. Through advanced communication, control, and metering methods, it achieves effective aggregation and efficient management of distributed photovoltaic, energy storage, and flexible load resources, mitigates the uncertainty of renewable energy output such as photovoltaics, facilitates renewable energy consumption, reduces the volatility risks of photovoltaics, and assists distributed resources in participating in power system dispatch and electricity market transactions.

[0003] Virtual power plants can participate in both centralized and distributed trading. Compared to centralized trading, distributed peer-to-peer trading offers decentralization and greater flexibility. However, current research on peer-to-peer trading models among virtual power plants focuses on electricity trading. Research on electricity-reserve and electricity-carbon joint trading largely employs centralized trading methods, with little attention paid to how to achieve peer-to-peer trading of multiple resources such as electricity, carbon emissions, and reserves from virtual power plants. In centralized trading, virtual power plants only trade with the main market, resulting in fixed transaction costs and a lack of incentive to utilize internal flexible resources, leading to high transaction costs. Regarding transaction uncertainty, existing risk mitigation methods for virtual power plants are applied to centralized trading; how to handle risk issues in peer-to-peer trading requires further research. For solving peer-to-peer trading models, existing alternating direction multiplier algorithms rely on the selection of iteration step sizes, resulting in low efficiency, poor convergence, and poor stability. Summary of the Invention

[0004] To address the technical problems mentioned in the background section, this invention proposes a multi-virtual power plant peer-to-peer trading method, electronic device, and storage medium.

[0005] To achieve the above-mentioned technical objectives, the technical solution of the present invention is as follows:

[0006] A peer-to-peer trading method for multiple virtual power plants includes the following steps:

[0007] S1. Collect and aggregate distributed resources within virtual power plants to construct a multi-virtual power plant electricity-carbon-reserve point-to-point trading model;

[0008] S2. Construct a conditional value at risk model, set the confidence level and risk preference coefficient of the conditional value at risk model, and add the conditional value at risk model to the multi-virtual power plant electricity-carbon-reserve point-to-point trading model in step S1 to construct a multi-virtual power plant electricity-carbon-reserve point-to-point trading model based on conditional value at risk.

[0009] S3. The consensus-based adaptive ADMM algorithm is used to solve the multi-virtual power plant electricity-carbon-reserve point-to-point trading model based on conditional value of risk in step S2 in a distributed manner to obtain the optimal decision for virtual power plants to participate in market trading.

[0010] Preferably, step S1 includes the following steps:

[0011] S11. Establish the objective function of the multi-virtual power plant electricity-carbon-reserve point-to-point trading model, expressed by the following formula:

[0012]

[0013] Where: N i The total number of virtual power plants participating in the transaction; s represents the photovoltaic output scenario; t represents the transaction period; ρ s Let be the probability of photovoltaic scenario s; and These represent the transaction costs of virtual power plant i in the electricity, carbon, and reserve markets during time period t; The energy storage cost of virtual power plant i during time period t in scenario s; and These represent the discomfort costs incurred by virtual power plant i adjusting central air conditioning and flexible loads during time period t in scenario s; C represents the fuel cell operating cost of virtual power plant i during time period t in scenario s; i,t The cost of peer-to-peer transactions for virtual power plant i within time period t;

[0014] S12. Establish constraints for the multi-virtual power plant electricity-carbon-backup point-to-point trading model, including fuel cell constraints, energy storage constraints, central air conditioning constraints, flexible load constraints, carbon emission constraints, and point-to-point trading constraints.

[0015] Preferably, in step S11,

[0016] The formula for electricity market transaction costs is expressed as follows:

[0017]

[0018] In the formula: and These represent the prices for purchasing and selling electricity in the electricity market during time period t; and These represent the amount of electricity that virtual power plant i purchases and sells from the electricity market during time period t;

[0019] The formula for carbon market transaction costs is expressed as follows:

[0020]

[0021] In the formula: and These represent the purchase and sale prices of carbon emissions in the carbon market during time period t; and These represent the carbon emissions that virtual power plant i purchases and sells from the carbon market during time period t;

[0022] The formula for transaction costs in the standby market is expressed as follows:

[0023]

[0024] In the formula: and These represent the purchase and sale prices of backup services in the backup market during time period t; and These represent the standby capacity that virtual power plant i purchases and sells from the standby market during time period t;

[0025] The energy storage cost formula is expressed as follows:

[0026]

[0027] In the formula: and These represent the charging and discharging amounts of energy stored by virtual power plant i during time period t in scenario s, respectively. and These are the charging and discharging dissipation coefficients of the energy stored in the virtual power plant i, respectively.

[0028] The cost formulas for central air conditioning and flexible load dispatch are expressed as follows:

[0029]

[0030]

[0031] In the formula: m and n are user discomfort coefficients; The indoor temperature of the user in the virtual power plant i during the time period t in scenario s; T i ref The most comfortable temperature for users within the virtual power plant; The flexible load value of the user for virtual power plant i during time period t in scenario s; The load baseline value for users of virtual power plant i within time period t;

[0032] The cost formula for fuel cells is expressed as follows:

[0033]

[0034] In the formula: The unit power generation cost of the fuel cell within virtual power plant i; Let be the power generation capacity of the fuel cell in virtual power plant i during time period t in scenario s;

[0035] The formula for peer-to-peer transaction costs is as follows:

[0036]

[0037] In the formula: c ij d ij e ij P represents the bilateral transaction coefficient between virtual power plant i and virtual power plant j, which reflects product differentiation; i,j,t E i,j,t R i,j,t These represent the trading volume of electricity, carbon emissions, and reserves for virtual power plant i and virtual power plant j during time period t.

[0038] Preferably, in step S12,

[0039] The constraint formula for fuel cells is expressed as follows:

[0040]

[0041]

[0042]

[0043]

[0044] In the formula: P i min and P i max These represent the minimum and maximum output power of the fuel cell within virtual power plant i, respectively. Let be the power generation capacity of the fuel cell in virtual power plant i during time period t in scenario s; The backup capacity provided by the fuel cell of virtual power plant i during time period t in scenario s; and These represent the lower and upper limits of the fuel cell reserve capacity within virtual power plant i, respectively; r i u and r i d For the upward and downward ramp rates of the fuel cells within the virtual power plant i;

[0045] The energy storage constraint formula is expressed as follows:

[0046]

[0047]

[0048]

[0049]

[0050] In the formula: P i c,max and P i d,max These are the maximum charging and discharging power of the energy storage within virtual power plant i, respectively; and These represent the charging and discharging amounts of energy stored in virtual power plant i during time period t in scenario s; S i,s,t The state of charge of the energy stored in the virtual power plant i during time period t in scenario s; and These represent the minimum and maximum energy storage capacities within virtual power plant i, respectively. and The charging and discharging efficiency of energy storage within the virtual power plant i;

[0051] The constraint formula for central air conditioning is expressed as follows:

[0052]

[0053] In the formula: α i,t β i γ i The parameters for virtual power plant i to describe the building's cold storage characteristics and weather conditions during time period t are related to the building characteristics of the building walls, windows and floors, as well as the outdoor temperature. σ represents the indoor temperature of the user in the virtual power plant i during time period t in scenario s; i The energy efficiency ratio of the central air conditioning chiller unit within the virtual power plant i; The cooling power of the central air conditioning in virtual power plant i during time period t in scenario s;

[0054] To maintain the indoor temperature within a specified range, the formula is as follows:

[0055]

[0056] In the formula: T in,min and T in,max These are the minimum and maximum permissible indoor temperatures, respectively.

[0057] The formula for flexible load constraints is expressed as follows:

[0058]

[0059]

[0060]

[0061]

[0062] In the formula: The flexible load value of the user in the virtual power plant i within the scenario s time period t; The load baseline value for users of virtual power plant i within time period t; The reserve capacity provided by virtual power plant i for flexible loads during time period t in scenario s; and These are the adjustable lower and upper limits of the flexible load within virtual power plant i, respectively.

[0063] The carbon emission constraint formula is expressed as follows:

[0064]

[0065] Where: G i,s,t For virtual power plant i, the photovoltaic certified emission reduction during time period t in scenario s; χ represents the photovoltaic power generation of virtual power plant i during time period t in scenario s; χ is the baseline emission factor, which is obtained by weighted average of the marginal emission factor of electricity and the marginal emission factor of capacity in the region where virtual power plant i is located.

[0066] The carbon emissions of the fuel cell within virtual power plant i are expressed by the following formula:

[0067]

[0068] In the formula: F i,s,t υ represents the carbon emissions of the fuel cell in virtual power plant i during time period t in scenario s; i Carbon emission intensity per unit output of fuel cells within virtual power plant i;

[0069] The formula for peer-to-peer transaction constraints is expressed as follows:

[0070] P i,j,t +P j,i,t =0 j≠i

[0071] E i,j,t +E j,i,t =0 j≠i

[0072] R i,j,t +R j,i,t =0 j≠i

[0073]

[0074]

[0075]

[0076] In the formula: P i,j,t E i,j,t R i,j,t These represent the trading volume of electricity, carbon emissions, and reserves for virtual power plant i and virtual power plant j during time period t, respectively. and These represent the amount of electricity that virtual power plant i purchases and sells from the electricity market during time period t; and These represent the carbon emissions that virtual power plant i purchases and sells from the carbon market during time period t; and These represent the standby capacity that virtual power plant i purchases and sells from the standby market during time period t; Carbon emission limits for fuel cells within virtual power plant i; The reserve demand of virtual power plant i during time period t in scenario s;

[0077] Preferably, the building is treated as a flexible resource, and the coupling relationship between the central air conditioning load and its reserve capacity is considered. The central air conditioning constraint is then rewritten as follows:

[0078]

[0079]

[0080]

[0081] In the formula: α i,t β i γ i σ is a parameter describing the building's cold storage characteristics and weather conditions for the virtual power plant i during time period t, and is related to the building characteristics of the building walls, windows, and floors, as well as the outdoor temperature; i The energy efficiency ratio of the central air conditioning chiller unit within the virtual power plant i; The cooling power of the central air conditioning in virtual power plant i during time period t in scenario s; The backup capacity provided by the central air conditioning of virtual power plant i during time period t in scenario s; Provide the backup indoor temperature for the central air conditioning system of the virtual power plant i during the time period t in scenario s; T in,min and T in,max These are the minimum and maximum permissible indoor temperatures, respectively.

[0082] Preferably, step S2 specifically includes the following steps:

[0083] S21. Define the confidence level and risk preference coefficient for the conditional value at risk model, and establish the conditional value at risk model, expressed by the following formula:

[0084]

[0085] Where: δ i The conditional risk value of the cost of virtual power plant i; φ i Value at risk for virtual power plant i; ξ is the confidence level; z i,s This represents the amount by which the cost of virtual power plant i exceeds the value at risk within scenario s;

[0086] z i,s The formula is expressed as follows:

[0087] z i,s ≥0

[0088]

[0089] In the formula: and These represent the transaction costs of virtual power plant i in the electricity, carbon, and reserve markets during time period t; The energy storage cost of virtual power plant i during time period t in scenario s; and These represent the discomfort costs incurred by virtual power plant i adjusting central air conditioning and flexible loads during time period t in scenario s; C represents the fuel cell operating cost of virtual power plant i during time period t in scenario s; i,t The cost of peer-to-peer transactions for virtual power plant i within time period t;

[0090] S22. Incorporate the Conditional Value at Risk (CQR) model into the multi-virtual power plant electricity-carbon-reserve peer-to-peer trading model in step S1, as expressed in the following formula:

[0091]

[0092] In the formula: C all The total cost of multiple virtual power plants based on conditional value at risk; the risk preference coefficient L represents the investor's attitude towards risk, and its value ranges from L∈[0,1]. The larger the value of L, the more risk-averse the virtual power plant decision-maker is.

[0093] Preferably, step S3 includes the following steps:

[0094] S31. The augmented Lagrangian form of the objective function of the multi-virtual power plant electricity-carbon-reserve peer-to-peer trading model based on conditional value of risk is as follows:

[0095]

[0096] In the formula: C all The overall total cost of multiple virtual power plants based on conditional risk value; λ i,j,t η i,j,t and τ i,j,t The dual variables are defined as the electricity trading P of virtual power plant i and virtual power plant j within time period t. i,j,t Carbon Emissions Trading i,j,t and standby transactions R i,j,t The price; ω、 The penalty factor is the step size;

[0097] S32, Regarding electricity trading P i,j,t Carbon Emissions Trading i,j,t Alternate Transaction R i,j,t and dual variable λ i,j,t η i,j,t τ i,j,t The iterative process is expressed by the following formula:

[0098]

[0099]

[0100]

[0101]

[0102] In the formula: k is the iteration number. The superscripts k and k+1 represent the k-th and k+1-th iterations, respectively;

[0103] S33. Calculate the original residual and dual residual after each iteration. The formula is expressed as follows:

[0104]

[0105] In the formula: and These are the original residuals of electricity, carbon emissions, and reserves of virtual power plant i and virtual power plant j in the (k+1)th iteration peer-to-peer transaction, respectively, within time period t. and These are the dual residuals of electricity, carbon emissions, and reserves of virtual power plant i and virtual power plant j in the (k+1)th iteration peer-to-peer transaction, respectively, within time period t.

[0106] S34. Determine whether the ADMM algorithm has converged by the iteration stopping condition. The formula for the iteration stopping condition is expressed as follows:

[0107]

[0108] Where: ε pri and εdual These are the upper tolerance limits for the original residual and the dual residual, respectively;

[0109] S35. Iteratively update the step size to accelerate the algorithm's convergence speed and stability. The step size update formula is expressed as follows:

[0110]

[0111]

[0112]

[0113] Where: μ is the proportionality constant between the original residual and the dual residual; κ incr κ decr For acceleration / deceleration factors, μ > 1, κ incr >1, κ decr >1;

[0114] S36. The model is solved using GAMS software, and the trading decisions of the virtual power plant are obtained after iterative convergence.

[0115] An electronic device includes a memory and a processor, the memory storing a computer program executable by the processor, the processor executing the computer program to implement the above-described multi-virtual power plant peer-to-peer trading method.

[0116] A storage medium storing a computer program, which, when read and executed, implements the above-described multi-virtual power plant peer-to-peer trading method.

[0117] The beneficial effects of adopting the above technical solution are as follows:

[0118] This invention proposes a peer-to-peer trading model for the electricity, carbon emissions, and reserve resources of virtual power plants, broadening the trading channels for virtual power plants, achieving resource sharing, and reducing transaction costs. It considers the uncertainty of photovoltaic output within virtual power plants and applies conditional value-at-risk theory to mitigate risk, enabling decision-makers to weigh risks and benefits. A consensus-based adaptive alternating direction multiplier algorithm is employed to achieve distributed optimization of multiple virtual power plants. This algorithm automatically updates the step size during iterative solving, thereby improving the computational efficiency of the constructed model. Attached Figure Description

[0119] Figure 1 This is a flowchart of the method of the present invention;

[0120] Figure 2 This is a schematic diagram of the main market electricity price in the example;

[0121] Figure 3This is a schematic diagram illustrating the cost and conditional value of risk of a virtual power plant under different risk preferences in the examples. Detailed Implementation

[0122] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings.

[0123] This invention designs a peer-to-peer trading method for multiple virtual power plants, such as... Figure 1 As shown, the steps are as follows:

[0124] Step 1: Collect and aggregate distributed resources within the virtual power plant to construct a peer-to-peer trading model for electricity-carbon-reserve power from multiple virtual power plants;

[0125] Step 2: Set the confidence level and risk preference coefficient of the conditional value at risk model, add the conditional value at risk model to the model in Step 1, and construct a multi-virtual power plant electricity-carbon-reserve peer-to-peer trading model based on conditional value at risk;

[0126] Step 3: Use the consensus-based adaptive ADMM algorithm to solve the model from Step 2 in a distributed manner to obtain the optimal decision for the virtual power plant to participate in market transactions.

[0127] This invention verifies the effectiveness of the proposed method using a simulation example consisting of three virtual power plants. Virtual power plant 1 includes a fuel cell, photovoltaic system, energy storage, central air conditioning, and flexible loads. Virtual power plants 2 and 3 both include photovoltaic system, energy storage, central air conditioning, and flexible loads. The parameters for the fuel cell, energy storage, central air conditioning, and photovoltaic system are shown in Table 1. The selling prices of the virtual power plants to the primary markets (electricity market, carbon market, and reserve market) are as follows: Figure 2 As shown. To prevent arbitrage by virtual power plants, the purchase price of virtual power plants from the main market is set to 1.5 times the selling price.

[0128] To better analyze the peer-to-peer transactions between virtual power plants, the example only shows the optimization results from 9:00 to 19:00 (the photovoltaic operating period).

[0129] Table 1 Virtual Power Plant Parameters

[0130]

[0131] To verify the impact of multi-product trading and peer-to-peer trading models on the revenue of each virtual power plant, this invention sets up six virtual power plant trading schemes as shown below (the trading method proposed in this invention corresponds to scheme 6):

[0132] Option 1: Virtual power plants engage in electricity trading with the main market.

[0133] Option 2: Virtual power plants engage in electricity-reserve trading with the primary market.

[0134] Option 3: Virtual power plants engage in electricity-carbon-sustainment trading with the primary market.

[0135] Option 4: Virtual power plants engage in electricity trading with the main market and other virtual power plants.

[0136] Option 5: Virtual power plants engage in power-reserve transactions with the main market and other virtual power plants.

[0137] Option 6: Virtual power plants engage in electricity-carbon-sustainment trading with the main market and other virtual power plants.

[0138] Table 2 shows the costs of the three virtual power plants participating in the electricity-carbon-reserve trading. A comparison of schemes 1, 2, and 3, and schemes 4, 5, and 6 shows that considering multi-product trading of virtual power plants effectively reduces their total cost. A comparison of schemes 1 and 4, 2 and 5, and 3 and 6 shows that considering point-to-point trading between virtual power plants reduces the total cost of multi-virtual power plant trading. Table 1 also shows that scheme 6 has the lowest total cost for virtual power plants, demonstrating that the trading method proposed in this invention can broaden the trading channels for virtual power plants, better leverage their resource management and market participation functions, and thus improve their revenue and social benefits.

[0139] Table 2 Virtual Power Plant Transaction Costs

[0140]

[0141] The impact of risk aversion value L on the total cost of the virtual power plant in the conditional value-at-risk model is as follows: Figure 3 As shown, it can be seen that as L increases, the virtual power plant becomes more conservative, reflected in the continuous increase of the total cost of the virtual power plant and the continuous decrease of the conditional risk value. Figure 3 This curve represents the relationship between the cost and conditional value of a virtual power plant. It quantifies the relationship between returns and risks, allowing virtual power plant decision-makers to determine their risk preferences based on their psychological expectations and the efficient frontier curve.

[0142] Table 3 shows a cost comparison between the deterministic approach and the conditional value-at-risk (VAT) approach for virtual power plants. In day-ahead decision-making, the deterministic approach does not consider the uncertainty of photovoltaic (PV) output; therefore, the cost of each virtual power plant in the day-ahead phase using this approach is lower than that of the conditional VAT approach. However, the deterministic approach ignores the uncertainty of PV output, which means that when the actual daily PV output is lower than the predicted value, each virtual power plant must purchase the insufficient power generation at a higher price in the main market, resulting in a significant increase in both intraday and total costs. In contrast, the conditional VAT approach considers PV uncertainty in day-ahead decision-making, thus its intraday dispatch costs and total costs are much lower than those of the deterministic approach, demonstrating the economic viability of the conditional VAT approach.

[0143] Table 3. Cost Comparison of Deterministic Model and Conditional Value at Risk Model

[0144]

[0145] Table 4 shows a comparison between the traditional ADMM and the adaptive step-size ADMM algorithm. The traditional ADMM algorithm is highly dependent on the step size, with the number of iterations and time varying significantly with the step size. In contrast, the adaptive step-size ADMM algorithm reduces this dependence on the initial step size by adjusting the step size, resulting in a relatively consistent iteration time for any step size. This significantly improves the stability of convergence. The comparison of the two algorithms demonstrates that, with the same initial step size, the adaptive step-size ADMM algorithm requires fewer iterations and has a shorter iteration time, proving its effectiveness.

[0146] Table 4 Comparison between Traditional ADMM and Adaptive Step Size ADMM

[0147]

[0148]

[0149] The embodiments are merely illustrative of the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of this invention.

Claims

1. A multi-virtual power plant point-to-point trading method, characterized in that, Comprise the following steps: S1, collect and aggregate the distributed resources in the virtual power plant, and construct a multi-virtual power plant electricity-carbon-backup point-to-point transaction model; S2, construct a conditional value at risk model, set the conditional value at risk model confidence and risk preference coefficient, add the conditional value at risk model to the multi-virtual power plant electricity-carbon-backup point-to-point transaction model in step S1, and construct a multi-virtual power plant electricity-carbon-backup point-to-point transaction model based on conditional value at risk; S3, the multi-virtual power plant electricity-carbon-backup point-to-point transaction model based on conditional value at risk in step S2 is solved by using an adaptive ADMM algorithm based on consensus, and the optimal decision of the virtual power plant participating in market transaction is obtained; Step S1 includes the following steps: S11, the objective function of the multi-virtual power plant electricity-carbon-backup point-to-point transaction model is established, and the formula is expressed as follows: where N i is the total amount of virtual power plants participating in the transaction; s is the photovoltaic output scenario; t is the transaction period; p s is the probability of photovoltaic scenario s; and are the transaction costs of virtual power plant i in the power, carbon, and reserve markets, respectively, within period t; is the energy storage cost of virtual power plant i within scenario s period t; and are the discomfort costs of virtual power plant i in adjusting central air conditioning and flexible load within scenario s period t, respectively; is the fuel cell operation cost of virtual power plant i within scenario s period t; C i,t is the point-to-point transaction cost of virtual power plant i within period t; S12, the constraint conditions of the multi-virtual power plant electricity-carbon-backup point-to-point transaction model are established, including fuel cell constraint, energy storage constraint, central air conditioning constraint, flexible load constraint, carbon emission constraint and point-to-point transaction constraint.

2. The method of claim 1, wherein, In step S11, The formula of the electricity market transaction cost is expressed as follows: wherein: and are the price of electricity purchase and sale in the power market in period t, respectively; and are the amount of electricity purchased and sold by the virtual power plant i from the power market in period t, respectively. The formula of the carbon market transaction cost is expressed as follows: wherein: and are the purchase and sale price of carbon emissions on the carbon market for period t, respectively; and are the carbon emissions purchased and sold by the virtual power plant i from the carbon market for period t, respectively. The formula of the backup market transaction cost is expressed as follows: wherein: and are the purchase and sale price of the reserve service in the reserve market for the period t, respectively; and are the reserve capacity purchased and sold by the virtual power plant i from the reserve market for the period t, respectively. The formula of the energy storage cost is expressed as follows: wherein: and are the charging and discharging quantities of the energy storage of the virtual power plant i in the time period t of the scenario s, respectively; and are the charging and discharging dissipation coefficients of the energy storage of the virtual power plant i, respectively. The formula of the central air conditioning and flexible load calling cost is expressed as follows: wherein: m and n are user discomfort coefficients; is the indoor temperature of the user in the virtual power plant i at the scenario s in the time period t; T i ref is the most comfortable temperature of the user in the virtual power plant i; is the flexible load value of the user in the virtual power plant i at the scenario s in the time period t; is the load reference value of the user in the virtual power plant i in the time period t; The formula of the fuel cell cost is expressed as follows: wherein: is the unit generation cost of fuel cells in virtual power plant i; is the generation power of fuel cells in virtual power plant i at time period t of scenario s. The formula of the point-to-point transaction cost is expressed as follows: wherein: c ij , d ij , e ij is the bilateral transaction coefficient between virtual power plant i and virtual power plant j, which reflects the product differentiation; P i,j,t , E i,j,t , R i,j,t are the transaction amounts of electricity, carbon emissions and reserve of virtual power plant i and virtual power plant j in time period t, respectively.

3. The method of claim 2, wherein, In step S12, The formula of the fuel cell constraint is expressed as follows: where P i min and P i max Pmin and Pmax are the minimum and maximum fuel cell output power in the virtual power plant i, respectively; Pgen is the generated power of the fuel cell in the virtual power plant i at time period t of scenario s; Psp is the spinning reserve capacity provided by the fuel cell in the virtual power plant i at time period t of scenario s; and Pmin and Pmax are the minimum and maximum fuel cell output power in the virtual power plant i, respectively; r i u and r i d is the up and down ramp rate of the fuel cell in the virtual power plant i; The formula of the energy storage constraint is expressed as follows: where P i c,max and P i d,max are the maximum charge and discharge power of the energy storage in the virtual power plant i, respectively; and are the charge and discharge amount of the energy storage in the virtual power plant i in the time period t of the scenario s, respectively; S i,s,t is the state of charge of the energy storage in the virtual power plant i in the time period t of the scenario s; and are the minimum and maximum storage amount of the energy storage in the virtual power plant i, respectively; and are the charge and discharge efficiency of the energy storage in the virtual power plant i, respectively; The formula of the central air conditioning constraint is expressed as follows: wherein: a i,t , β i , γ i are parameters of the virtual power plant i describing the building cold storage characteristics and weather conditions in the time period t, related to the building characteristics of the building walls, windows and floors and the outdoor temperature; is the indoor temperature of the users of the virtual power plant i in the time period t of the scenario s; σ i is the energy efficiency ratio of the central air conditioning refrigeration unit in the virtual power plant i; is the refrigeration power of the central air conditioning in the virtual power plant i in the time period t of the scenario s. The formula of the indoor temperature keeping in the temperature range is expressed as follows: where: T in,min and T in,max are the minimum and maximum allowed indoor temperatures, respectively; The formula of the flexible load constraint is expressed as follows: wherein: is the flexible load value of a user in a virtual power plant i at a time period t in a scenario s; is the load reference value of a user in a virtual power plant i at a time period t; is the reserve capacity provided by the flexible load of a virtual power plant i at a time period t in a scenario s; and are the lower and upper adjustable limits of the flexible load in a virtual power plant i, respectively. The formula of the carbon emission constraint is expressed as follows: wherein: G i,s,t is the PV certified emission reduction amount of virtual power plant i in scenario s during period t; is the PV generation amount of virtual power plant i in scenario s during period t; and γ is the baseline emission factor, which is obtained by weighted average of the marginal emission factor of electricity and the marginal emission factor of capacity in the region where virtual power plant i is located. The formula of the carbon emission of the fuel cell in the virtual power plant i is expressed as follows: wherein: F i,s,t is the carbon emission of the fuel cell of the virtual power plant i in the scenario s during the time period t; υ i is the carbon emission intensity of the fuel cell of the virtual power plant i per unit output. The formula of the point-to-point transaction constraint is expressed as follows: P i,j,t +P j,i,t =0 j≠i E i,j,t +E j,i,t =0 j≠i R i,j,t +R j,i,t =0 j≠i wherein: P i,j,t , E i,j,t , R i,j,t are the traded quantities of electricity, carbon emissions and reserve of virtual power plant i and virtual power plant j in time period t, respectively; and are the electricity purchased and sold by virtual power plant i from the electricity market in time period t, respectively; and are the carbon emissions purchased and sold by virtual power plant i from the carbon market in time period t, respectively; and are the reserve capacity purchased and sold by virtual power plant i from the reserve market in time period t, respectively; is the carbon emission limit of fuel cells in virtual power plant i; is the reserve demand of virtual power plant i in time period t in scenario s.

4. The method of claim 3, wherein, The building is regarded as a flexible resource, the coupling relationship between the central air conditioning load and the backup capacity provided by the building is considered, the central air conditioning constraint is rewritten, and the formula is expressed as follows: wherein: a i,t , β i , γ i are parameters of the virtual power plant i describing the building cold storage characteristics and weather conditions in time period t; σ i is the central air conditioning chiller energy efficiency ratio in the virtual power plant i; is the central air conditioning refrigeration power in the virtual power plant i in scenario s time period t; is the central air conditioning reserve capacity provided by the virtual power plant i in scenario s time period t; is the indoor temperature after the central air conditioning reserve provided by the virtual power plant i in scenario s time period t; T in,min and T in,max are the minimum and maximum allowable indoor temperatures, respectively.

5. The method of claim 1, wherein, Step S2 specifically includes the following steps: S21, set the conditional value at risk model confidence and risk preference coefficient, establish the conditional value at risk model, and the formula is expressed as follows: where: δ i is the conditional value at risk of the cost of the virtual power plant i; φ i is the value at risk of the cost of the virtual power plant i; ξ is the confidence level; z i,s represents the amount by which the cost of the virtual power plant i exceeds the value at risk within the scenario s; z i,s The formula is expressed as follows: z i,s ≥0 wherein: and are the transaction costs of virtual power plant i in the electricity, carbon and reserve markets, respectively, in period t; is the energy storage cost of virtual power plant i in scenario s in period t; and are the discomfort costs of virtual power plant i in scenario s in period t from adjusting central air conditioning and flexible loads, respectively; is the fuel cell operation cost of virtual power plant i in scenario s in period t;C i,t is the point-to-point transaction cost of virtual power plant i in period t; S22, add the conditional value at risk model to the multi-virtual power plant electricity-carbon-backup point-to-point transaction model in step S1, and the formula is expressed as follows: In the formula, C all The overall total cost of the multi-virtual power plant based on the conditional value at risk; the risk preference coefficient L represents the attitude of the investor to the risk, and the value range is L∈[0, 1], the greater the value of L, the more risk-averse the virtual power plant decision maker is.

6. The method of claim 1, wherein, Step S3 includes the following steps: S31, the augmented Lagrangian form of the objective function of the multi-virtual power plant electricity-carbon-backup point-to-point transaction model based on conditional value at risk is as follows: wherein: C all Total cost of the multi-virtual power plant based on the conditional value at risk; λ i,j,t η i,j,t and τ i,j,t are dual variables defined as the price of electricity trading P i,j,t , carbon emission trading E i,j,t and reserve trading R i,j,t for virtual power plant i and virtual power plant j in time period t, respectively; ω、 is the penalty factor, i.e. the step size; S32, on electricity trading P i,j,t , carbon emission trading E i,j,t , reserve trading R i,j,t and the dual variable λ i,j,t , η i,j,t , τ i,j,t iteration, formula expression as follows: In the formula, k is the iteration number; Subscripts k and k+1 represent the kth and (k+1)th iterations, respectively; S33, calculate the original residual and dual residual after each iteration, and the formula is expressed as follows: wherein: and are the primal residuals of the virtual power plant i and the virtual power plant j, respectively, for electrical energy, carbon emissions and reserve in the time period t in the point-to-point trade in the k+1th iteration; and are the dual residuals of the virtual power plant i and the virtual power plant j, respectively, for electrical energy, carbon emissions and reserve in the time period t in the point-to-point trade in the k+1th iteration. S34, judge whether the ADMM algorithm converges or not through the iteration stopping condition, and the formula of the iteration stopping condition is expressed as follows: where: ε pri and ε dual are the upper tolerance limits for the primal and dual residuals, respectively. S35, update the step length by iteration to speed up the convergence speed and stability of the algorithm, and the formula of the step length updating is expressed as follows: where: μ is a scale constant between the original residual and the dual residual; κ incr is an acceleration / deceleration factor, μ > 1, κ decr > 1, κ incr > 1; and decr > 1. S36, the model is solved by using GAMS software, and the transaction decision of the virtual power plant is obtained after iteration convergence.

7. An electronic device, comprising: Comprise: A memory and a processor, the memory stores a computer program executable by the processor, and the processor executes the computer program to implement the multi-virtual power plant point-to-point transaction method of any one of claims 1-6.

8. A storage medium, characterized by The storage medium stores a computer program, and the computer program is read and executed to implement the multi-virtual power plant point-to-point transaction method of any one of claims 1-6.

Images