Virtual power plant operator and producer-consumer collaborative optimization method, system, device and medium

Abstract

This invention belongs to the field of power grid technology and discloses a collaborative optimization method, system, equipment, and medium based on virtual power plant operators and prosumers, including the following steps: S1: Establish a collaborative management framework for electricity carbon emissions between virtual power plant operators and prosumers; S2: Based on the collaborative management framework, construct a two-layer game model; the upper layer is the operator, and the lower layer is the prosumer. The operator calculates the node marginal electricity price and node carbon intensity through optimal power flow and distributes them to the prosumer. The prosumer optimizes scheduling based on the node marginal electricity price and node carbon intensity and submits the power purchase and sale curve; S3: Solve the two-layer game model using a distributed iterative algorithm with a penalty function, and achieve adaptive optimization of the model solution by introducing a penalty term to obtain the optimal collaborative electricity carbon emissions strategy.

Description

Technical Field

[0001] This invention belongs to the field of power grid technology, and in particular relates to a collaborative optimization method, system, equipment and medium based on virtual power plant operators and producers and consumers. Background Technology

[0002] In the field of power grid technology, achieving a low-carbon transformation of the power system has become a core development goal for the industry. The penetration rate of distributed renewable energy (DRG) in the power distribution network is continuously increasing. This trend is driving traditional power users to gradually transform into prosumers who combine power production and consumption capabilities. Prosumers can effectively improve the energy utilization efficiency of the entire power system by integrating their own power generation resources, energy storage systems (ESS), and flexible loads.

[0003] Against this backdrop, bidirectional energy exchange models in distribution networks are becoming increasingly mature, and carbon trading mechanisms are gradually being integrated into distribution network operations. The emergence of virtual power plants provides an important platform for implicit energy sharing between producers and consumers. However, existing distribution network operation and management still lacks a comprehensive electricity-carbon co-management scheme, making it difficult to simultaneously meet the decarbonization needs of local distribution network systems and the goals of efficient absorption and coordinated operation of a large amount of DRGs.

[0004] Traditional centralized management methods are computationally intensive and fail to adequately consider the individual profit-maximizing demands of different stakeholders (prosumers and consumers). Highly distributed management methods, on the other hand, are prone to violating distribution network operation constraints, leading to suboptimal scheduling results. Under a hierarchical management framework, existing research often models the two-layer game between virtual power plant operators and prosumers as an equilibrium problem with equilibrium constraints. However, such methods require detailed equipment operation information within the prosumers, which presents significant challenges in practical applications and fails to effectively address the impact of operational uncertainties among prosumers. Therefore, a collaborative optimization scheme for electricity and carbon emissions between virtual power plant operators and prosumers is urgently needed. Summary of the Invention

[0005] This application addresses the problems existing in the prior art by proposing a collaborative optimization method, system, equipment, and medium based on virtual power plant operators and prosumers. By co-managing the electricity carbon between virtual power plant operators and prosumers, it considers the operational uncertainties of prosumers, the application of the distribution network carbon emission factor model (CEF model), and a privacy-preserving game equilibrium solution algorithm. Furthermore, by introducing a penalty term to achieve adaptive optimization of the model solution, the optimal electricity carbon coordination strategy is obtained.

[0006] To achieve the above objectives, this application provides the following technical solution:

[0007] Firstly, a collaborative optimization method, system, equipment, and medium based on virtual power plant operators and prosumers include the following steps: S1: Establishing a collaborative management framework for electricity carbon emissions between virtual power plant operators and prosumers; S2: Constructing a two-layer game model based on the collaborative management framework; the upper layer is the operator, and the lower layer is the prosumer. The operator calculates the node marginal electricity price and node carbon intensity through optimal power flow and distributes them to the prosumer. The prosumer optimizes scheduling based on the node marginal electricity price and node carbon intensity and submits the electricity purchase and sale curve; S3: Solving the two-layer game model using a distributed iterative algorithm with a penalty function, and achieving adaptive optimization of the model solution by introducing a penalty term to obtain the optimal collaborative electricity carbon emissions strategy.

[0008] Optionally, in the two-level game model, the objective function of the virtual power plant operator includes generation cost, external power purchase cost, transaction cost with prosumers and tiered carbon cost, and the constraints include node power balance, branch power flow limit and generator output upper and lower limits.

[0009] Optionally, producer-consumer resource scheduling optimization includes the coordinated scheduling of photovoltaic, wind power, micro gas turbines, energy storage systems, electric vehicles, and flexible loads, with the goal of minimizing the sum of electricity costs and carbon trading costs.

[0010] Optionally, an opportunity constraint method can be used to address the impact of renewable energy DRG output and load forecasting errors of producers and consumers on their operation. The opportunity constraint transforms the original rigid constraint into a flexible constraint, that is, it sets the probability of the rigid constraint being satisfied to be no less than a certain given confidence level. The uncertainty of the source load is simultaneously transformed into an opportunity constraint, as shown in the following equation:

[0011] (16)

[0012] (17)

[0013] (18)

[0014] in, The probability of an event occurring; The confidence level for producer-consumer m; The net power demand of producer-consumer m at time t; and These are the net load and electricity supply within the producer-consumer group, respectively.

[0015] Optionally, in step S3, a game problem-solving algorithm based on a penalty function is constructed to accelerate the solution process. The penalty function-based method involves adding a penalty term to the objective function of the producers and consumers, as shown in the following equation:

[0016] (20)

[0017] in This represents the number of iterations. Let m be the electricity purchased by producer-consumer m at time t and the kth iteration. Let m be the electricity sold by producer-consumer at time t and in the kth iteration; For consumer m in the first Penalty term in the next iteration; These are the first-order and second-order penalty factors, where The calculation method is defined as follows:

[0018] (twenty one)

[0019] in, is an exponential decay function used to smoothly adjust the second-order penalty factor as the iterative residual decreases, where e is the natural constant; is the residual scaling factor for the k-th iteration, used to adjust the magnitude of the residual index and control the change range of the penalty factor; As an intermediate variable, The value range of is (0,1), which means for The increase of is very sensitive; its value increases rapidly with the increase of the iteration residual, thus accelerating the convergence of the game process. Meanwhile, The definition is as follows:

[0020] (twenty two)

[0021] in, express In the iterative solution of the two-level game problem, when the first... The second iteration and the first When the residuals at each time step between -1 iterations are less than a set threshold, it is considered to have reached an equilibrium state, as shown below:

[0022] (twenty three).

[0023] Optionally, the following steps are also included: S4: Constructing a P2P secondary carbon trading market, in which the carbon allowance CEP bids of producers and consumers are constrained by the net difference between the initial free allowance and the actual emissions. The bidding model takes into account historical transaction prices, local supply and demand differences and carbon emission assessment and compliance pressure. The compliance pressure increases exponentially as the assessment period approaches.

[0024] Optionally, the clearing rules for the secondary carbon trading market are as follows: virtual power plant operators match carbon quota buyers and sellers with the goal of maximizing social welfare, and the transaction price is the arithmetic average of the winning bids.

[0025] Secondly, the present invention provides a collaborative optimization system based on virtual power plant operators and prosumers, for implementing the collaborative optimization method based on virtual power plant operators and prosumers as described in the first aspect.

[0026] Thirdly, the present invention provides a computer device, the computer device including a memory, a processor and a computer program, wherein when the computer program is executed by the processor, it implements the collaborative optimization method based on virtual power plant operators and prosumers as described in the first aspect.

[0027] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the collaborative optimization method based on virtual power plant operators and prosumers as described in the first aspect.

[0028] The beneficial effects of this application are as follows:

[0029] 1. By co-managing the carbon emissions of virtual power plant operators and producers and consumers, considering the operational uncertainties of producers and consumers, the application of the carbon emission factor model (CEF) at the distribution network level, and the game equilibrium solution algorithm based on privacy protection, and by introducing a penalty term to achieve adaptive optimization of the model solution, the optimal carbon emissions coordination strategy is obtained.

[0030] 2. By constructing a two-way P2P secondary carbon trading market, the influence of carbon emission assessment and compliance pressure and local carbon market supply and demand relationship is introduced into the producer-consumer pricing model to more reasonably and accurately simulate the strategic pricing behavior of producers and consumers. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of a collaborative optimization method based on virtual power plant operators and producers / consumers in one embodiment of this application.

[0032] Figure 2 For the purposes of this application Figure 1 The diagram illustrates a collaborative interaction architecture between a virtual power plant operator and a producer-consumer in one embodiment.

[0033] Figure 3 For the purposes of this application Figure 1 The diagram shows a schematic of the carbon emission factor (CEF) model at the distribution network level in one embodiment.

[0034] Figure 4 For the purposes of this application Figure 1 The diagram illustrates a virtual power plant carbon co-optimization method considering local carbon trading in a distribution network environment, as shown in one embodiment.

[0035] Figure 5 For the purposes of this application Figure 1 The diagram shows a schematic of an IEEE 33-node system in one embodiment.

[0036] Figure 6 For the purposes of this application Figure 1 The diagram illustrates a typical daily net consumer load, market electricity price, and carbon emission factor in one embodiment.

[0037] Figure 7 For the purposes of this application Figure 1 The diagram illustrates a virtual power plant operator's power generation decision-making process in one embodiment.

[0038] Figure 8 For the purposes of this application Figure 1 The diagram shows the nodal marginal price (LMP) and nodal carbon intensity (NCI) curves for producers and consumers at various times in one embodiment.

[0039] Figure 9 For the purposes of this application Figure 1 The diagram shows the producer-consumer energy scheduling results in one embodiment.

[0040] Figure 10 For the purposes of this application Figure 1 The diagram shows the carbon trading results under one embodiment.

[0041] Figure 11 For the purposes of this application Figure 1 The diagram shows the residual changes during the solution process of a two-layer game model in one embodiment.

[0042] Figure 12 For the purposes of this application Figure 1 The diagram shows the CEP (Consumer Price Program) transaction price curves for local carbon allowances under different initial free carbon allowances in one embodiment.

[0043] Figure 13 For the purposes of this application Figure 1 The diagram shows the results of local carbon trading conducted over four consecutive days in one embodiment.

[0044] Figure 14 This is a structural diagram of the electronic device shown in Embodiment 3 of this application.

[0045] Figure label:

[0046] 300. Electronic device; 301. Processor; 302. Communication bus; 303. User interface; 304. Network interface; 305. Memory. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description of this application is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely one preferred embodiment of this application and are only used to explain this application. They do not limit the scope of protection of this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0048] Example 1:

[0049] like Figure 1-13 As shown, a collaborative optimization method based on virtual power plant operators and prosumers includes the following steps:

[0050] S1: Establish a collaborative carbon management framework for virtual power plant operators and producers / consumers;

[0051] S2: Based on the aforementioned electricity-carbon collaborative management framework, a two-layer game model is constructed; the upper layer consists of operators and the lower layer consists of producers and consumers. The operators calculate the node marginal electricity price and node carbon intensity through optimal power flow and distribute them to the producers and consumers. The producers and consumers optimize scheduling based on the node marginal electricity price and node carbon intensity and submit the electricity purchase and sale curve.

[0052] S3: Solve the two-layer game model using a distributed iterative algorithm with a penalty function, and achieve adaptive optimization of the model solution by introducing a penalty term to obtain the optimal electric carbon synergy strategy.

[0053] S4: Construct a P2P secondary carbon trading market. In the secondary carbon trading market, the carbon allowance (CEP) quotas quoted by producers and consumers are constrained by the net difference between the initial free allowance and the actual emissions. The pricing model takes into account historical transaction prices, local supply and demand differences, and carbon emission assessment and compliance pressure. The compliance pressure increases exponentially as the assessment period approaches.

[0054] In this embodiment, this application considers the operational uncertainties of producers and consumers, the application of the distribution network carbon emission factor model (CEF model), and a privacy-preserving game equilibrium solution algorithm through the collaborative management of electricity carbon between virtual power plant operators and prosumers. By constructing a two-way P2P secondary carbon trading market, the influence of carbon emission assessment and compliance pressure and the supply and demand relationship in the local carbon market is introduced into the prosumer bidding model to more reasonably and accurately simulate the strategic bidding behavior of prosumers.

[0055] like Figure 2As shown, considering that some regions have not yet established competitive electricity markets that allow broad user participation, this application studies the operation problem of virtual power plants under the retail electricity market. In this case, virtual power plants only need to complete external electricity transactions based on a single electricity purchase price curve published by the market. This application models the internal electricity transaction clearing between virtual power plant operators and prosumers as a two-level SG problem and uses a distributed iterative algorithm based on penalty functions to solve the game equilibrium. In each iteration, both virtual power plant operators and prosumers aim to minimize their own operating costs. The virtual power plant operator first considers the overall electricity demand of the system, solves the Optimal Power Flow (OPF) problem to determine the generator output plan and market electricity purchase plan, and then derives the nodal marginal price (LMP) and nodal carbon intensity (NCI), which are then distributed to prosumers. Prosumers formulate optimal scheduling plans based on this information and submit electricity purchase and sale curves to the virtual power plant operator.

[0056] In carbon trading, the carbon allowance (CEP) bids of prosumers are limited by the difference between their initially allocated free carbon allowances and actual carbon emissions (i.e., net carbon emissions). This application also proposes a carbon allowance (CEP) bidding model that comprehensively considers factors such as historical transaction prices of local carbon allowances, local supply and demand, and external carbon emission assessment and compliance pressures. The compliance and market pressures reflected in the model are understood as follows: as the assessment period approaches, prosumers with a carbon allowance (CEP) shortfall will significantly increase their willingness to purchase CEPs to avoid regulatory penalties; if there is an oversupply of carbon allowances in a local market, sellers will also be more willing to sell their surplus CEPs, and vice versa. Finally, the P2P carbon trading of prosumers is cleared by virtual power plant operators with the goal of maximizing social welfare.

[0057] The proposed carbon-electricity co-management scheme is based on the following assumptions: (1) In response to the problem of insufficient and difficult-to-obtain data in my country's secondary carbon market, this application generates local P2P carbon trading historical price data through market simulation. Producers and consumers can predict the future carbon quota (CEP) price trend based on this dataset; (2) Except for the generator units of virtual power plant operators and producers and consumers, all other nodes in the distribution network are pure load nodes. These nodes only participate in the electricity market clearing (accepting the node marginal price (LMP)) and will not affect the node carbon intensity (NCI) of the nodes where producers and consumers are located.

[0058] Carbon emission flow calculations in a distribution network environment must be based on power flow calculations of the power system. These calculations obtain key data such as branch power flow, generator injected power, node injected power, and load size. Subsequently, explicit or implicit constraints on carbon emissions are imposed on the entire system to achieve the goal of low-carbon system operation.

[0059] like Figure 3As shown, in a specific embodiment, the core concept of the distribution network carbon emission factor model (CEF model) is carbon potential, which is the amount of carbon dioxide carried per unit of electricity. Carbon potential is further divided into nodal carbon intensity (NCI) and branch carbon potential. Specifically, nodal carbon potential... It is the weighted average of the carbon potential of the incoming power flow and the power injected by the node generators, i.e., total carbon emissions divided by total power (typically in kgCO2 / kWh). Furthermore, according to the Power Supply Utilization (PSP), the power flowing into a node from multiple branches will flow out of the node in proportion to its incoming power; therefore, the carbon potential of each branch... All are equivalent to the nodal carbon potential of their starting nodes. As shown in the following equation:

[0060] (1)

[0061] (2)

[0062] Among them, e b，t For nodes The branch carbon potential at time t; branch road The loading carbon potential at time t; Number the branch roads; Let i be the active power of the load on branch i at time t. Let i be the load current of branch i at time t. The power distribution line resistance is the load carbon emission intensity of branch i. Let be the active power of the generator in branch i at time t. The carbon emission intensity of generator i in branch i is the amount of carbon emissions generated per unit of electricity produced. Number the nodes. For the set of nodes in the distribution network; and These represent the starting and ending nodes of the branch, respectively.

[0063] In a specific embodiment, based on the carbon emission factor model (CEF) at the distribution network level, the carbon emissions of each entity within the virtual power plant are accurately calculated according to their electricity production and consumption behavior. First, a virtual power plant electricity clearing model based on SG in the distribution network environment is constructed. Then, the operational uncertainties within the prosumers are characterized. Next, a distributed iterative algorithm based on a penalty function is proposed to solve this game problem. Finally, a local carbon market is established, and its settlement mechanism is described in detail. Traditional market clearing processes use a 1-hour time scale for both electricity volume and price, but this time scale is insufficient to meet the refined scheduling needs of virtual power plants and prosumers. Therefore, this application sets the power dispatching time scale to 15 minutes (hereinafter referred to as...). (This indicates that) the timescale for local carbon trading is set at 1 hour (hereinafter referred to as...). express).

[0064] The power clearing problem is modeled as a game between virtual power plant operators (VPS) and prosumers. The dominant player (VPS) solves the AC optimal power flow (ACOPF) problem (ACOPF is mainly used to solve resource allocation optimization in power system dispatching) based on the electricity demand of prosumers, and then distributes the nodal marginal price (LMP) and nodal carbon intensity (NCI) to the prosumers. Meanwhile, the followers (prosumers) formulate their day-ahead optimal dispatch plans based on the LMP and NCI, and submit their own power purchase and sale curves at various times to the VPS. After multiple iterations, this game reaches equilibrium, and the decision variables of each player are finally determined.

[0065] The power distribution network is viewed as a radial connection diagram. ,in and For the branch numbers and sets. and Let represent the node number, set, and balancing node, respectively. The total cost for a virtual power plant operator includes generation cost, cost of purchasing electricity from the external grid, cost of electricity trading with producers and consumers, and carbon cost. This application considers that the virtual power plant operator calculates its own carbon cost according to a tiered carbon pricing system. The optimal power flow (OPF) problem is modeled as follows:

[0066] (3)

[0067] (4)

[0068] (5)

[0069] The constraints are:

[0070] (6)

[0071] in, Represents the set of downstream nodes of node j; Let t be the electricity price in the external market. and These represent the power output purchased by virtual power plant producers and consumers from the external market and from other producers and consumers, respectively. The total operating cost for the distribution network operator; The cost of power generation for the distribution network; The carbon trading cost for the power distribution network; , Let be the power generation cost coefficient of generator set i; Let be the electricity trading price coefficient for node i / line m at time t; Let be the power purchased by node i / line m at time t; The time interval is indicated by h, which represents the carbon trading session period. Let be the active power of the load on branch ij at time t; , These are the resistance and reactance of the branch circuit, respectively. Let be the current in branch ij at time t; Let be the active power output of the generator at node j at time t; Let Jk be the active power of the load at time t. , These represent the total active and reactive power demand in the distribution network, respectively. The reactive power output of the generator in branch jk at time t; Let be the voltage at node j at time t; Let be the voltage at node i at time t; Let be the active power of the load on branch ij at time t; , It is the product of the voltage at node i and the current at branch j, used for quantifying the voltage-current relationship; , These are the lower and upper squared values ​​of the branch current, respectively. , These are the lower and upper squared values ​​of the node voltage, respectively. , These are the upper and lower limits of the active power output of generator set i, respectively. , These are the upper and lower limits of the reactive power output of generator set i, respectively. , These are the upper and lower limits of the active power of the load, respectively. , These are the upper and lower limits of the reactive power of the load, respectively. Adjust the power for the demand-side response of node i at time t; The available adjustable power for the demand response of node i at time t is the active power potential for flexible load adjustment. Let be the normal load power of node i at time t, that is, the active power portion of the load that cannot be adjusted; The carbon cost for virtual power plant operators when using tiered carbon pricing (as shown in Equation (1)); The carbon emission factor of the power grid; The amount of free carbon credits allocated to the generator set; and These are the reactive power of the branch circuit and the generator, respectively; The square of the node voltage; Let LMP be the nodal marginal price, which is the price at which a producer-consumer purchases a unit of electricity from a virtual power plant operator. It is considered as the dual variable of the active power balance constraint and is obtained after solving the ACOPF problem. In Equation (6), the upper and lower horizontal lines of the variable represent the upper and lower limits of the variable, respectively. In Equation (4), the price at which the virtual power plant operator purchases electricity from the producer-consumer in the k-th iteration is set to half of the nodal marginal price LMP in the (k-1)-th iteration. That is, for the producer-consumer, the internal electricity sales price is half of the purchase price, and the virtual power plant operator sets the external market electricity price to the initial nodal marginal price LMP.

[0072] Producers and consumers formulate optimal scheduling plans based on nodal marginal price (LMP) and nodal carbon intensity (NCI) with the goal of minimizing electricity and carbon trading costs. Their mathematical model includes various resources, as shown in the following equation:

[0073] (7)

[0074] (8)

[0075] (9)

[0076] (10)

[0077] Constraints

[0078] (11)

[0079] (12)

[0080] (13)

[0081] (14)

[0082] (15)

[0083] in, and For the producer-consumer numbers and total number; and For EVs, their serial numbers and total number; For scheduling time, The time interval is 15 minutes. These represent the total demand response adjustment cost, energy cost, and carbon trading cost for producer-consumer m, respectively; h is the carbon trading period identifier; and H is the set of carbon trading periods. The carbon trading price coefficient for producer-consumer m during carbon trading period h; , , These represent the photovoltaic power output, wind power output, and micro gas turbine output of producer m at time t, respectively. , , The carbon emission intensity of photovoltaic, wind power, and micro gas turbine outputs per producer and consumer (m) are respectively. , These are the energy storage charging power and discharging power of producer m at time t, respectively; The carbon emission intensity of ESS; The electricity purchased by consumer m at time t; For producer-consumer m associated nodes The carbon potential at time t; The electricity sold by consumer m at time t; The carbon trading price coefficient for producer-consumer m during carbon trading period h; The initial carbon allowance for producer-consumer m during carbon trading session h; These are the operating costs of producer-consumer m (mt), operating costs of ESS (Electric Power Supply), costs of interruptible load dispatching, costs of transferable load dispatching, electricity trading costs, and revenue from electric vehicles (EVs). For mt, the fuel price coefficient; The energy conversion efficiency of mt; , , These are the cost coefficients for energy storage systems (ESS), energy storage inter-channel (IL), and energy storage transfer (TL), respectively. For the producer m, reduce the IL power at time t; , Let TL be the input power and output power of producer-consumer m at time t, respectively. The electricity transaction costs for producer-consumer m; For producer-consumer m associated nodes The electricity trading price coefficient at time t; The profit coefficient is ev; , These represent the purchasing and selling power of producer-consumer m at time t, respectively. This refers to the battery capacity of the EVz. These are the initial state of charge and the target state of charge of EVz, respectively. Let PV, WT, and MT be the outputs of producer-consumer m at time t, respectively. Let be the discharge power and charging power of EVz at time t, respectively. These represent the electricity purchase power, ESS discharge power, and ESS charging power of consumer m at time t, respectively. The base load power of producer m at time t; The base load power of producer m at time t; Let m be the load fluctuation of producer-consumer m at time t; IL is the reduction ratio factor; , The allowed call period for IL; This represents the upper limit of the charging and discharging power of the EVz. Let m be the set of EVs of prosumers; , What are they respectively? , These are the charging efficiency and discharging efficiency of the EV, respectively. , These are the upper and lower limits of MT's output, respectively. Let ESS be the state of charge at time t; , These are the charging efficiency and discharging efficiency of ESS, respectively. Let ESS be the state of charge at time t; and For the reduced load, the load transferred in and out; This represents the upper limit of the MT's climbing range. This represents the maximum output capacity of the MT. This refers to the upper limit of the charging and discharging power of the energy storage system (ESS). For the capacity of the energy storage system (ESS), These represent the lower and upper limits of the SoC for the Energy Storage System (ESS).

[0084] In the above model, nodal carbon intensity (NCI) and nodal marginal electricity price (LMP) are key factors for virtual power plants to achieve coordinated carbon management. This is because the unit electricity cost for a producer-consumer is actually equal to the unit electricity purchase cost (i.e., the nodal marginal electricity price LMP) and the carbon emission cost generated by consuming a unit of electricity (i.e., nodal carbon intensity (NCI) multiplied by the carbon allowance (CEP) unit price). Therefore, a change in either the nodal marginal electricity price LMP or the nodal carbon intensity (NCI) will lead to changes in the producer-consumer's electricity consumption, carbon emissions, optimal dispatch plan, and carbon trading demand.

[0085] In one specific embodiment, this application employs an opportunity constraint method to address the impact of renewable energy DRG output and load forecasting errors on the operation of producers and consumers. Unlike traditional point forecasting methods, opportunity constraints transform the original rigid constraints into flexible constraints, that is, setting the probability of the rigid constraints being satisfied to be no less than a given confidence level. Taking equation (11) as an example, the uncertainty of the source load is simultaneously transformed by chance constraints, as shown in the following equation:

[0086] (16)

[0087] (17)

[0088] (18)

[0089] in, The probability of an event occurring; The confidence level for producer-consumer m; The net power demand of producer-consumer m at time t; and These represent the net load and electricity supply within the producer-consumer group, respectively. The opportunity constraint form of the above equation is further transformed into a deterministic form that is easy to solve. According to the concept of CDF, equation (16) is transformed into the following equation:

[0090] (19)

[0091] in, For random variables The inverse function of CDF at a confidence level The value at that location. Generally speaking, random variables The accurate probability distribution of random variables is difficult to obtain, making it difficult to derive the analytical form of the CDF. This application employs the GBRT quantile prediction method, a data-driven approach, to obtain the CDF without prior knowledge or assumptions about the distribution of random variables. The value of . At this point, the original implicit chance constraint is equivalently transformed into a deterministic constraint that is easy to solve.

[0092] In one specific embodiment, this application constructs a game problem-solving algorithm based on a penalty function to accelerate the solution process, and considers the non-convexity of the game problem to avoid its equilibrium solution falling into local oscillations. Unlike the bisection method, the penalty function-based method adds a penalty term to the objective function of the producers and consumers, as shown in the following equation:

[0093] (20)

[0094] in, This represents the number of iterations. Let m be the electricity purchased by producer-consumer m at time t and the kth iteration. Let m be the electricity sold by producer-consumer at time t and in the kth iteration; For consumer m in the first Penalty term in the next iteration; These are the first-order and second-order penalty factors, where The calculation method is defined as follows:

[0095] (twenty one)

[0096] in, is an exponential decay function used to smoothly adjust the second-order penalty factor as the iterative residual decreases, where e is the natural constant; is the residual scaling factor for the k-th iteration, used to adjust the magnitude of the residual index and control the change range of the penalty factor; As an intermediate variable, The value range of is (0,1), which means for The increase in is very sensitive; its value increases rapidly with the increase in the iterative residual (i.e., the difference between the power purchase and sales curves reported by the producers and consumers in the second iteration), thus accelerating the convergence of the game process. Meanwhile, The definition is shown in the following formula:

[0097] (twenty two)

[0098] in, express In the iterative solution of the two-level game problem, when the first... The second iteration and the first When the residuals at each time step between -1 iterations are less than a set threshold, it is considered to have reached an equilibrium state, as shown in the following formula:

[0099] (twenty three)

[0100] In one specific embodiment, the superscripts b and s of the variables shown in this application represent "buyer" and "seller," respectively, and the assessment period for prosumer carbon emissions is set to one day. In the local carbon market constructed in this application, prosumers choose to conduct carbon trading at any time. To simulate continuous trading behavior over a medium- to long-term timescale, this application assumes that prosumers conduct P2P carbon trading at every moment. Referring to the carbon trading allocation mechanisms implemented in my country and the European Union, the initial free carbon allowances for prosumers and generating units are determined based on their historical average carbon emissions.

[0101] The carbon allowances (CEPs) reported by prosumers at each trading moment are constrained by their cumulative net carbon allowances (CEPs) holdings at the previous moment and their net carbon emissions at the current moment, as shown in the following formula:

[0102] (twenty four)

[0103] in, and This represents the reported quantity and upper limit of carbon allowances (CEPs), where a positive value indicates that carbon allowances (CEPs) have been sold, and a negative value indicates that carbon allowances (CEPs) have been purchased. The net carbon emissions of producers and consumers can be calculated according to the method in formula (5-8). This represents the cumulative net carbon allowance (CEP) held by prosumers after the completion of carbon trading at the current moment, equal to the sum of their declared CEP and the actual CEP awarded in the bid. difference.

[0104] In one specific embodiment, in local peer-to-peer carbon trading, producers and consumers (P2Ps) first determine their desired carbon allowance (CEP) trading price, i.e., based on their own forecasts and the day's transaction prices to determine a benchmark value. On this basis, as trading progresses and the carbon emission compliance assessment period arrives within a day, and as the supply and demand relationship of CEPs in the local market changes (e.g., supply falling short of demand), the trading pressure on CEP buyers gradually increases. Simultaneously, the rate of increase in this pressure also rises. Therefore, this application uses an exponential function-based approach to quantify the impact of external factors on P2P pricing and describes the pricing processes of CEP buyers and sellers separately.

[0105] (1) Carbon allowance CEP buyers

[0106] (25)

[0107] (26)

[0108] in, The upper limit of the bid for producer-consumer m as a buyer during carbon trading session h; For producer-consumer m, the compliance pressure factor is used as the buyer during carbon trading period h; For producer-consumer m, during carbon trading period h, supply and demand sensitive factors are considered as buyers; The carbon price forecast for producer-consumer m during carbon trading session h; This is the market carbon price reference value for carbon trading session h; This is the scaling factor for performance pressure; , H represents the baseline value for producer-consumer m's carbon trading account; H represents the total number of carbon trading sessions; h represents the carbon trading session identifier. This is a scaling factor for the supply and demand relationship; The predicted carbon price for producer-consumer m during carbon trading session h is the day-ahead value. denoted as m, representing the net carbon allowance shortfall for producer-consumer m during carbon trading session h; b represents the buyer coefficient. This represents the average CEP carbon allowance transaction price for that day. As a weight, and taking into account the accumulation of prediction errors, this value will also change with the trading time. The increase in the price increases the impact of raising the daily transaction price on the expected price of producers and consumers. These are the carbon emission assessment compliance pressure coefficient and the local supply and demand relationship coefficient of carbon quotas (CEP), respectively, both given in the form of exponential functions. As the base; This refers to the partial carbon allowance (CEP) supply and demand situation published by virtual power plant operators, which is the carbon allowance CEP demand minus the carbon allowance CEP supply. The threshold for producers and consumers regarding performance evaluation and local supply and demand changes is understood as follows: when the evaluation period is approaching or the demand for carbon quotas (CEP) in a local market exceeds the supply of carbon quotas (CEP), buyers of carbon quotas (CEP) tend to raise their bids to win the bid and meet their own carbon quota (CEP) needs.

[0109] (2) Carbon quota CEP sellers

[0110] (27)

[0111] (28)

[0112] Where s is the seller coefficient; the meaning of the variable is consistent with equations (25)-(26). It can be seen that the bidding strategy of carbon quota CEP sellers is similar to that of buyers. The difference is that when they have unsold carbon quota CEP or when the supply of carbon quota CEP exceeds the demand in a local market, they will choose to lower their carbon quota CEP bids in order to win the bid and sell the carbon quota CEP as soon as possible.

[0113] Similar to the clearing process in the electricity market, after receiving bids from producers and consumers, virtual power plant operators clear carbon market transactions at the current moment with the goal of maximizing social welfare, as shown in the following formula:

[0114] (29)

[0115] The constraints are:

[0116] (30)

[0117] in, A collection of CEP (Consumer-Owned Enterprise) carbon allowance buyers and sellers; and This represents the CEP (Carbon Allowance) volume successfully won by both sellers and buyers. The final transaction price is the average of the bids submitted by the successful bidders.

[0118] Example 2:

[0119] This embodiment provides a collaborative optimization system based on virtual power plant operators and prosumers, used to implement the collaborative optimization method based on virtual power plant operators and prosumers as described in Embodiment 1.

[0120] Example 3:

[0121] like Figure 14As shown, this embodiment provides an electronic device, which may include: at least one processor, at least one network interface, a user interface, a memory, and at least one communication bus.

[0122] The communication bus can be used to enable communication between the various components mentioned above.

[0123] The user interface may include buttons, and optional user interfaces may also include standard wired interfaces and wireless interfaces.

[0124] The network interface may include, but is not limited to, Bluetooth modules, NFC modules, Wi-Fi modules, etc.

[0125] The processor may include one or more processing cores. It connects various parts of the electronic device via various interfaces and lines, executing instructions, programs, code sets, or instruction sets stored in memory, and accessing data stored in memory to perform various functions and process data. Optionally, the processor can be implemented using at least one hardware form of DSP, FPGA, or PLA. The processor may integrate one or more of the following: CPU, GPU, and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also be implemented as a separate chip without being integrated into the processor.

[0126] The memory may include RAM or ROM. Optionally, the memory may include a non-transitory computer-readable medium. The memory can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor. The memory, as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an evaluation application. The processor can be used to call the evaluation application stored in the memory and execute the steps of the collaborative optimization method based on virtual power plant operators and prosumers mentioned in the foregoing embodiments.

[0127] Example 4:

[0128] This embodiment provides a computer-readable storage medium storing instructions that, when executed on a computer or processor, cause the computer or processor to perform the above-described instructions. Figure 1 One or more steps in the illustrated embodiment. If the constituent modules of the above-described electronic device are implemented as software functional units and sold or used as independent products, they can be stored in the computer-readable storage medium.

[0129] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this specification are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., Digital Versatile Discs (DVDs)), or semiconductor media (e.g., Solid State Disks (SSDs)).

[0130] Those skilled in the art will understand that all or part of the processes in the method of Embodiment 1 described above can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks. Unless otherwise specified, the technical features of this embodiment and the implementation scheme can be combined arbitrarily.

[0131] Example 5:

[0132] like Figure 5As shown, this application conducted a series of simulation experiments on the IEEE 33-bus system to verify the effectiveness of the proposed method. The rated voltage level was 10kV, and the peak loads in the system were 3MW and 2.55MVar. The virtual power plant operator manages three gas turbine units located at nodes 2, 4, and 10. Four prosumers with different resource endowments were considered to participate in the virtual power plant aggregation, located at nodes 20, 29, 8, and 27. All four prosumers are equipped with renewable energy generation resources such as PV and WT. The remaining equipment configurations are as follows: Prosumer 1 - TL, IL; Prosumer 2 - MT, ESS energy storage system; Prosumer 3 - MT, TL; Prosumer 4 - ESS energy storage system, EV. Figure 6 The net load curves of typical day-ahead prosumers are presented, along with the external energy market electricity price obtained from the PJM market and the equivalent carbon emission factor on the grid side. Since virtual power plant operators themselves do not participate in local carbon trading, this application uses historical CEP carbon allowance price data from relevant EU datasets as a reference price for calculating carbon costs. Other simulation parameters used in this application are shown in Table 1. The free carbon allowances allocated to the four prosumers are 6 tCO2, 9 tCO2, 2.2 tCO2, and 9 tCO2, respectively. The policy data for EVs are consistent with those given in Table 3. The GBRT-based quantile prediction method was implemented using the scikit-learn third-party library in a Python programming environment. All optimization models and simulation experiments in this application were programmed on the Matlab platform and solved using the Gurobi 9.5 commercial solver. The local market studied in this application is a day-ahead market.

[0133] Table 1 Simulation Parameters

[0134]

[0135] Figure 7 This displays the generation and market power purchase curves for virtual power plant operators. To further illustrate the impact of carbon costs on their generation decisions, Figure 7 (b) also shows the relevant decisions of virtual power plant operators without considering carbon costs. Figure 7 (a) The unit cost of electricity refers to the sum of the external market electricity purchase price and the unit carbon cost of electricity, the latter being equal to the product of the external market carbon allowance (CEP) price and the grid-side carbon emission factor. The above unit cost of electricity measures the total carbon cost that a user must bear when purchasing a unit of electricity from the grid.

[0136] Figure 7(a) clearly shows a negative correlation between the amount of electricity purchased from the grid by virtual power plant operators and the unit cost of electricity. Compared to the scenario where carbon costs are ignored (where the unit cost of electricity equals the market price), the output of all generating units decreases when carbon costs are considered because their own carbon emission intensity is too high (all greater than 0.8547 tCO2 / MWh), leading to an excessively high overall generation cost, which is the sum of the unit generation cost and the carbon emission cost generated per unit of electricity generated. Furthermore, generator 1 experiences the largest output decrease (13.22%), while generators 2 and 3 decrease by 7.92% and 9.04%, respectively, because generator 1 has the highest overall generation cost. Therefore, virtual power plant operators need to comprehensively consider the generation economics and carbon emission intensity of their units when planning generation schedules.

[0137] Figure 8 This paper presents the nodal marginal price (LMP) and nodal carbon intensity (NCI) curves for prosumers at different nodes, with comparisons of LMP and NCI for virtual power plant operators under two scenarios: considering and not considering their own carbon costs. Considering the direction and relationship between power flow and carbon emission flow, the location of nodes in the distribution network leads to different nodal carbon intensity (NCI). For prosumer 1 at node 20, the power injected into this node comes from the grid and generator unit 1. For prosumer 3, its nodal carbon intensity (NCI) is jointly affected by grid-side injected power, generator unit 1, and generator unit 2's output power. Prosumers 2 and 4 are affected by grid-side injected power and the output power of all generator units. Meanwhile, the carbon emission intensity of generator units is generally higher than that of the grid side because the latter is a mixture of various forms of power generation (e.g., photovoltaic, wind power, hydropower, and thermal power). Since the nodal carbon intensity (NCI) is affected by both grid-side injected power and generator unit output power, the trend of NCI variation is similar to the adjustment of generator unit output. Figure 8 The 20:00-0:00 period in (b) is more obvious. Therefore, from Figure 8 (b) shows that the node carbon intensity NCI of prosumers 2 and 3 are similar, while the node carbon intensity NCI of prosumer 1 has a significantly different trend. In both scenarios, the node carbon intensity NCI of prosumer 1 has the largest change compared to other prosumers.

[0138] Figure 8The variation periods of the nodal marginal electricity price (LMP) and the nodal carbon intensity (NCI) are similar because they are both affected by the generation decisions of virtual power plant operators. The nodal marginal electricity price (LMP) represents the marginal generation cost of a marginal unit, which is related to the output power of an individual generator and may be discontinuous due to distribution network power constraints. Therefore, the variation of the nodal marginal electricity price (LMP) is more intuitive than that of the nodal carbon intensity (NCI), as the latter is a weighted average of the carbon emission intensity of multiple generator units (a constant value, independent of generation power).

[0139] In summary, changes in the nodal marginal price (LMP) and nodal carbon intensity (NCI) directly affect the dispatch plans and carbon trading demands of producers and consumers. This conclusion can also be extended to virtual power plant operators. Therefore, considering carbon emission costs makes the electricity carbon management scheme more in line with the operational requirements of the real world.

[0140] The energy dispatch results of prosumers are as follows Figure 9 As shown in the figure, the scheduling results of IL and TL are represented by gradient color and striped bar charts, respectively. In the game-based power clearing process, producers and consumers continuously adjust their own resource and equipment output plans to reduce operating costs, which will be described in detail below.

[0141] Prosumer 1 reduced some of its load during periods of high electricity prices, such as 14:00-19:00, and shifted some load to periods with lower electricity prices, such as 9:00-10:00 and 12:00-13:00. Prosumers 2 and 3 increased the output of their own MT (Metal Transport Unit) equipment to reduce the cost of purchasing electricity from external sources. For prosumer 4, it utilized the energy storage characteristics of energy storage systems (ESS) and electric vehicles (EVs) to achieve a certain degree of arbitrage. Regarding the charging needs of EVs, although different types of EVs entered the charging station at different times, prosumer 4 chose to charge them in a concentrated manner during periods of low electricity prices, ultimately ensuring that all EVs were fully charged to the desired SoC (System of Charge) when leaving the charging station. In addition, EVs can be regarded as generalized energy storage system (ESS) resources, and their discharge can be used to meet the electricity demand of prosumers. In summary, compared with traditional electricity consumers, prosumers make full use of the energy consumption characteristics of the different resources they manage to avoid additional electricity-carbon coupling costs per unit of electricity, i.e., the sum of the nodal marginal price (LMP) and the carbon cost incurred per unit of electricity consumed. Two-way power interaction should also be incorporated into the power clearing problem in the distribution network environment through a reasonable mechanism.

[0142] Figure 10 The results of the partial P2P carbon trading proposed in this application are shown, with positive values ​​in the figure representing the purchase of carbon allowances (CEP). Figure 10In (a), the carbon allowance CEP transaction price first decreases, then fluctuates upwards, and remains at a high level at the end of the day. This is because at the beginning of the day, the carbon allowances of various entities are relatively abundant, and sufficient transaction price data for the day has not yet been accumulated. Therefore, the bids of producers and consumers mainly depend on their forecasts (as shown in equations (26) and (28)). As the trading time progresses, demand for carbon allowance CEP begins to appear in local carbon markets, and carbon allowance CEP buyers gradually raise their bids to win the bids, which leads to an increase in the carbon allowance CEP transaction price. Frequent carbon allowance CEP transactions occur after 12:00 noon on the same day because at this time, the sufficient carbon allowance CEP demand / supply accumulated by producers and consumers exceeds their relevant thresholds, forcing them to participate in the trading to meet their own needs.

[0143] Figure 10 (b) shows the net carbon allowance (CEP) holdings of prosumers at various points in time within a 24-hour period. Generally, each prosumer meets carbon emission assessments from carbon emission regulators after engaging in localized carbon trading. The changing trends of different prosumers' net carbon allowances (CEP) depend on their different thresholds for carbon emission assessment compliance pressure, such as prosumer 4's... The larger net carbon allowance (CEP) holdings of Producers 2 and 3 make them less willing to raise their bids when faced with the same demand for CEP compared to other producers and consumers. Therefore, they complete fewer carbon transactions, reflected in their more volatile net CEP holdings. Producers 2 and 3 are more sensitive to related compliance pressures, showing that they are completing carbon transactions almost constantly, resulting in their net CEP holdings fluctuating around zero.

[0144] Figure 11 The figure illustrates the residual curves during the game-theoretic solution process between virtual power plant operators and prosumers. These residuals are the cumulative differences between the prosumer's electricity purchase and sale curves in two consecutive iterations. The figure also shows the convergence curves of these residuals under conditions of no algorithmic intervention (i.e., only iterative solution) and using a distributed iterative algorithm with a fusion bisection method. The results show that without the bisection method or the penalty function method proposed in this application, the solution to this game problem falls into periodic oscillations. Furthermore, the convergence process of the game equilibrium solution differs after using the two methods. Figure 11 Dashed lines are used to divide the different convergence stages. Under the penalty function-based method, the initially large iterative residual causes the residual to decrease rapidly in subsequent iterations. As the residual gradually decreases, the penalty term also decreases, leading to a less tightening of convergence. With the increase in the number of iterations, the coefficient v also increases rapidly to force producers and consumers to stop adjusting their trading plans. For the bisection method, when oscillations occur, it tightens the upper and lower limits of the relevant price variables, resulting in a reduction in the virtual power plant operator's electricity trading profit from $617.9 (based on the penalty function method proposed in this application) to $591.3.

[0145] This application compares the proposed electricity-carbon co-management method with several existing methods through comparative experiments. The methods used for comparison are described below:

[0146] Method 1: Producers and consumers trade using the tiered carbon pricing system issued by carbon emission regulators, with a benchmark purchase price of $65 / tCO2 and a selling price of $30 / tCO2. Electricity trading prices still use the nodal marginal price (LMP) issued by virtual power plant operators.

[0147] Method 2: Virtual power plant operators derive electricity-carbon coupling prices based on the nodal marginal price (LMP) and the distribution network carbon emission factor (CEF) model, which guide producers and consumers to operate in a low-carbon manner by imposing carbon taxes.

[0148] Method 3: Both electricity clearing and carbon trading are cleared uniformly on an hourly time scale. Electricity trading is cleared first, followed by partial carbon trading involving trading preferences. The carbon trading costs are borne equally by the buyers and sellers.

[0149] Table 2 presents some statistical indicators related to prosumers under four different methods.

[0150] In Method 1, the tiered carbon price increases rapidly with the increase in producer-consumer carbon emissions. Therefore, the producer-consumer's carbon emissions are the lowest among the four methods, at only 26.35 tCO2. However, due to the overemphasis on carbon emissions and their costs, the producer-consumer's own scheduling plan is not economically optimal.

[0151] Method 2 achieves joint pricing under electricity-carbon coupling, but it involves a variable strongly correlated with the carbon emissions of marginal units. Since marginal units typically have higher carbon emission intensity, the carbon credits attributed to prosumers are higher than their actual carbon emission costs. In this case, prosumers at higher carbon emission intensity nodes have a stronger incentive to reduce the amount of electricity purchased from virtual power plant operators to reduce their own carbon emissions. Ultimately, this results in more carbon allowance (CEP) transactions among prosumers under this method (a total of 6.81 tCO2).

[0152] Method 3, which involves partial carbon trading, considers that the carbon cost be shared equally between the buyer and seller of carbon allowances (CEPs), thus reducing the carbon cost for producers and consumers. However, this method lacks flexible CEP trading pricing and ignores the impact of carbon costs on electricity trading, leading to more unnecessary carbon emissions. Furthermore, hourly electricity trading decisions may be too short-term, especially considering the interrelationships and influences of flexible resources such as energy storage systems (ESSs) at various points in time during operation.

[0153] The proposed method comprehensively considers the coupling relationship between electricity and carbon in the decision-making process, meaning that carbon costs are taken into account during power dispatch, and local carbon trading is based on the results of power clearing. Simultaneously, the decision-making model for CEP (Consumer-Exclusive Pricing Program) carbon quota trading considers the preferences of producers and consumers and market supply and demand to gradually form reasonable bids and transaction prices. Therefore, in terms of carbon costs and total costs, the proposed method performs best among the four methods.

[0154] Table 2 Comparison of relevant indicators under different co-management methods of electricity and carbon dioxide

[0155]

[0156] Figure 12 The figure shows the price change curves of partial carbon allowances (CEPs) under different initial free carbon allowance allocations for prosumers and consumers. Scenario 1 is the baseline value, while the initial free carbon allowances in scenarios 2 and 3 increase and decrease by 10%, respectively.

[0157] Overall, an increase in the initial amount of free carbon allowances leads to a decrease in the average transaction price of CEP carbon allowances, and vice versa. The price trends of CEP carbon allowances are similar across the three scenarios: an initial decrease, followed by an increase, and finally fluctuation. This is because at the beginning of the trading day, producers and consumers have relatively abundant carbon allowances and low demand for CEP carbon allowances, resulting in lower transaction prices. Furthermore, past transaction prices also influence current producer and consumer CEP carbon allowance bids. Towards the end of the day, some CEP carbon allowance buyers face significant compliance pressure and carbon allowance shortages, prompting them to continuously raise their bids to win the contracts, thus driving up the transaction price of CEP carbon allowances.

[0158] Between 21:00 and 24:00, the CEP carbon allowance transaction prices in scenarios 2 and 3 showed decreasing and increasing trends, respectively, due to different market supply and demand relationships. In scenario 2, excessive initial carbon allowance allocation led to oversupply, forcing CEP carbon allowance sellers to lower their bids to sell the excess CEP. Conversely, in scenario 3, the local carbon market resembled a scarce market, so CEP carbon allowance sellers, influenced by compliance pressures and market supply and demand, continuously raised their bids to meet their needs. The market supply relationship at this time also encouraged CEP carbon allowance sellers to raise their bids to earn more profit, thereby driving up the transaction price. The CEP carbon allowance transaction price changes in scenarios 1 and 3 were quite similar because both were influenced by parameter settings, resulting in a state of supply shortage.

[0159] To verify the feasibility of localized carbon trading from a medium- to long-term perspective, Figure 13The results of four consecutive days of local carbon trading are presented. On Day 1, the price of CEP (China Expenditure Credits) carbon allowances initially decreased and then increased. On Day 2, given the relatively high CEP price at the end of Day 1, producers and consumers rushed to purchase more CEP carbon allowances in advance, leading to a rapid increase in both price and volume. Subsequently, CEP carbon allowance prices remained at a high level, but trading volume was low. This was because the excessively high price dampened buyers' enthusiasm for trading, prompting them to reduce their own carbon emissions to avoid potential carbon emission penalties. Finally, CEP carbon allowance prices fell due to oversupply. On Day 3, the price of CEP carbon allowances fluctuated significantly, indicating a divergence in expectations among producers and consumers regarding CEP prices, and the distribution of trading volume also showed a fluctuating trend. At the end of Day 3, CEP carbon allowance prices decreased again, indicating that buyers' willingness to trade was also low at this point. On the fourth day, the price trend of carbon allowance CEP was similar to that of the third day, fluctuating around $60 / tCO2. At the end of the day, the increase in demand for carbon allowance CEP once again boosted the transaction price and volume of carbon allowance CEP.

[0160] To assess different carbon emission compliance coefficients Carbon quota CEP supply and demand relationship coefficient Regarding the impact of local carbon quota CEP transaction prices, this application addresses the different benchmark values ​​and coefficients mentioned above. A series of simulation experiments were conducted. Table 3 shows the average transaction price of carbon quota CEP ($ / tCO2) under different benchmark values ​​mentioned above. The relevant analysis and conclusions are as follows:

[0161] (1) Buyer correlation coefficient: The increase from 1.0001 to 1.003 led to a 7.79% increase in the average CEP price of carbon allowances. According to the exponential function model in equation (5-25), a larger base... This means that CEP buyers are more sensitive to their CEP demand, and will be more likely to raise their bids significantly when current CEP quotas exceed the threshold. Changes in these two coefficients will have a similar impact on the average transaction price of carbon allowances (CEP). However, the local CEP price will not increase indefinitely as these two coefficients increase, because its transaction price ceiling is the external market CEP price.

[0162] (2) Seller correlation coefficient: larger and smaller This means that sellers with surplus carbon allowances (CEP) tend to lower their bids to secure sales. When these two factors change, the average transaction price of carbon allowances only changes by 3.31% and 4.62%, respectively. and In comparison, this parameter appears to have a smaller impact on the transaction price of local carbon quota CEPs. This is because, in the simulation of this application, the demand for carbon quota CEPs in the local carbon market is higher than the supply of carbon quota CEPs for most of the time. Therefore, the impact of the seller's price adjustment on the transaction price is different from that of the buyer.

[0163] Table 3. Changes in the average transaction price of producer-consumer carbon allowances (CEP) under different coefficient benchmark values.

[0164]

[0165] Taking Producer-consumer 1 as an example, Table 4 shows the changes in the average transaction price of carbon quotas (CEP) and the carbon cost of Producer-consumer 1 under different carbon emission compliance pressure thresholds and market supply and demand pressure thresholds.

[0166] According to equations (25) and (27), the above threshold is located in the exponential part of the exponential function. Therefore, when faced with the same carbon quota CEP demand or market supply and demand situation, a smaller threshold will lead to a greater increase in bids. This parameter reflects the sensitivity of producers and consumers to external pressures. A smaller threshold reflects a more aggressive bidding strategy, and a higher purchase bid will also increase the chances of producers and consumers winning the bid (as shown in equation (28)).

[0167] Table 4 also shows that, since the CEP carbon quota transaction price is affected by the bids of all participating entities, adjustments to the threshold of individual producers and consumers have little impact on it, especially with... and Compared to the previous period, the average transaction price of carbon allowances for CEP and the carbon cost for producer-consumer 1 are... and The carbon cost increases when the threshold is below 0.02 and 0.052. However, as these parameters continue to increase, producer-consumer 1's carbon cost fluctuates. This is because a larger threshold means a lower willingness to raise bids, thus reducing the chances of winning a bid. Ultimately, producer-consumer 1 may be fined by carbon emission regulators for exceeding the assessment requirements.

[0168] In summary, changes in the benchmark values ​​of the carbon emission assessment compliance coefficient and the carbon quota CEP supply-demand relationship coefficient have a greater impact on the average transaction price of carbon quota CEP than changes in the relevant thresholds for producers and consumers. At the same time, the smaller threshold setting makes it easier for producers and consumers to win bids in carbon trading with other entities, but aggressive bidding strategies may also lead to additional carbon trading costs.

[0169] Table 4. Changes in the average CEP carbon quota transaction price and producer-consumer carbon cost under different thresholds.

[0170]

[0171] (1) The proposed carbon co-management method has a low computational cost and can be applied to real-world medium and low voltage distribution networks. The main obstacle lies in the carbon emission accounting of prosumers, especially when they inject power into the distribution network. At this time, virtual power plant operators cannot know the actual carbon emission intensity inside the network due to privacy protection, so they can only rely on the prosumers to calculate and report it themselves. At the same time, virtual power plant operators may need to deploy some energy storage system (ESS) resources at certain nodes to cope with the large-scale bidirectional power flow caused by prosumers and high-penetration renewable energy DRGs.

[0172] (2) The carbon allowance CEP pricing model proposed in this application is applied in the following scenarios: It is used to simulate the trading decision-making behavior of entities in the carbon market, thereby providing support for the improvement of my country's secondary carbon market mechanism, such as generating a large amount of market trading data as a benchmark value for subsequent reference. Simultaneously, this pricing model is also used for the custody decisions of trading entities, with relevant parameters adjusted to form different bidding preferences. It is worth mentioning that in local carbon markets at medium- to long-term scales, it is necessary to include both carbon allowance CEP buyers and sellers as much as possible; otherwise, the market will become a single scarce market or a buyer's market, leading to unreasonable carbon allowance CEP transaction prices.

[0173] (3) In the proposed method, historical local carbon allowance CEP transaction price data is considered as a reference value in the producer-consumer bidding process, which is similar to the general electricity market clearing process. However, hourly transaction data is insufficient in most secondary carbon markets. For example, secondary carbon trading in most provinces of my country is conducted on a one-day cycle. Therefore, this application chooses to generate relevant usable data through prior market simulation. For producers-consumers, as h increases, the carbon allowance CEP transaction price at past times of the day will become the expected bid of producers-consumers to a greater extent than the predicted value (as shown in equations (26) and (28)). Therefore, the source of historical data or the accuracy of predicting potential carbon allowance CEP prices based on historical data does not affect the feasibility of the proposed method.

[0174] This application proposes a virtual power plant (VFP) co-optimization method for carbon emissions that considers distribution network operation and local carbon education. First, the power clearing of VFP operators and producers / consumers is modeled as a two-level SG problem, and a distributed iterative algorithm based on penalty functions is proposed to accelerate the game equilibrium solution. The nodal marginal price (LMP) and nodal carbon intensity (NCI) of each node are calculated using the ACOPF and the distribution network carbon emission factor model CEF. Second, opportunity constraints are used to handle the uncertainty of renewable energy DRG and load output within producers / consumers. Subsequently, a local carbon trading market is constructed among producers / consumers, in which the proposed carbon trading pricing strategy comprehensively considers historical trading data and external factors (such as market supply and demand and carbon emission assessment and compliance pressure), and can more accurately reflect the strategic pricing behavior of market participants. Numerical simulation results show that: (1) the method proposed in this application reduces both the carbon emissions and total cost of producers / consumers; (2) it verifies the feasibility of medium- and long-term continuous carbon trading and presents the bidirectional pricing and trading results of producers / consumers in detail; (3) sensitivity analysis reveals the impact of key parameters on the carbon trading decision-making process, providing a basis for producers / consumers' pricing strategies.

[0175] The above-described specific embodiments are preferred embodiments of a collaborative optimization method, system, equipment, and medium based on virtual power plant operators and producers and consumers in this application. They are not intended to limit the specific scope of this application. The scope of this application includes but is not limited to these specific embodiments. All equivalent changes made in accordance with the shape and structure of this application are within the protection scope of this application.

Claims

1. A collaborative optimization method based on virtual power plant operators and prosumers, characterized in that, Includes the following steps: S1: Establish a collaborative carbon management framework for virtual power plant operators and producers / consumers; S2: Based on the aforementioned electricity-carbon collaborative management framework, a two-layer game model is constructed; the upper layer consists of operators and the lower layer consists of producers and consumers. The operators calculate the node marginal electricity price and node carbon intensity through optimal power flow and distribute them to the producers and consumers. The producers and consumers optimize scheduling based on the node marginal electricity price and node carbon intensity and submit the electricity purchase and sale curve. S3: Solve the two-layer game model using a distributed iterative algorithm with a penalty function, and achieve adaptive optimization of the model solution by introducing a penalty term to obtain the optimal electric carbon synergy strategy; In step S3, a game problem-solving algorithm based on a penalty function is constructed to accelerate the solution process. The penalty function-based method adds a penalty term to the objective function of the producers and consumers, as shown in the following equation: （20） in, This represents the number of iterations. Let m be the electricity purchased by producer-consumer m at time t and the kth iteration. Let m be the electricity sold by producer-consumer at time t and in the kth iteration; For consumer m in the first Penalty term in the next iteration; These are the first-order and second-order penalty factors, where The calculation method is defined as follows: （21） in, is an exponential decay function used to smoothly adjust the second-order penalty factor as the iterative residual decreases, where e is the natural constant; is the residual scaling factor for the k-th iteration, used to adjust the magnitude of the residual index and control the change range of the penalty factor; As an intermediate variable, The value range of is (0,1), which means for The increase of is very sensitive; its value increases rapidly with the increase of the iterative residual, thus accelerating the convergence of the game process. Meanwhile, The definition is as follows: （22） in, express In the iterative solution of the two-level game problem, when the first... The second iteration and the first When the residuals at each time step between -1 iterations are less than a set threshold, it is considered to have reached an equilibrium state, as shown below: （23）。 2. The collaborative optimization method based on virtual power plant operators and prosumers according to claim 1, characterized in that, In the two-layer game model, the objective function of the virtual power plant operator includes generation cost, external power purchase cost, transaction cost with producers and consumers, and tiered carbon cost. The constraints include node power balance, branch power flow limit, and upper and lower limits of generator output.

3. The collaborative optimization method based on virtual power plant operators and prosumers according to claim 1, characterized in that, The resource scheduling optimization for producers and consumers includes the coordinated scheduling of photovoltaic, wind power, micro gas turbines, energy storage systems, electric vehicles, and flexible loads, with the goal of minimizing the sum of electricity costs and carbon trading costs.

4. The collaborative optimization method based on virtual power plant operators and prosumers according to claim 1, characterized in that, An opportunity constraint approach is adopted to address the impact of renewable energy DRG output and load forecasting errors on the operation of producers and consumers. The opportunity constraint transforms the original rigid constraint into a flexible constraint, that is, it sets the probability of the rigid constraint being met to be no less than a certain given confidence level. The uncertainty of the source load is simultaneously transformed into an opportunity constraint, as shown in the following equation: （16） （17） （18） in, The probability of an event occurring; The confidence level for producer-consumer m; The net power demand of producer-consumer m at time t; The amount of electricity supplied within the producer-consumer group; These represent the photovoltaic power output, wind power output, and micro gas turbine output of producer m at time t, respectively. , These represent the purchasing and selling power of producer-consumer m at time t, respectively. These are the ESS discharge power and ESS charging power of consumer m at time t, respectively. Let be the discharge power and charging power of EVz at time t, respectively. , Let TL be the input power and output power of producer-consumer m at time t, respectively. Let m be the set of EVs of prosumers; For the producer m, reduce the IL power at time t; Let m be the base load power of the producer / consumer at time t.

5. The collaborative optimization method based on virtual power plant operators and prosumers according to claim 1, characterized in that, It also includes the following steps: S4: Constructing a P2P secondary carbon trading market, in which the carbon quota (CEP) reported by producers and consumers is constrained by the net difference between the initial free quota and the actual emissions. The pricing model takes into account historical transaction prices, local supply and demand differences and carbon emission assessment and compliance pressure. The compliance pressure increases exponentially as the assessment period approaches.

6. The collaborative optimization method based on virtual power plant operators and prosumers according to claim 5, characterized in that, The clearing rules of the secondary carbon trading market are as follows: virtual power plant operators match carbon quota buyers and sellers with the goal of maximizing social welfare, and the transaction price is the arithmetic average of the winning bids.

7. A collaborative optimization system based on virtual power plant operators and prosumers, characterized in that, Used to implement the collaborative optimization method based on virtual power plant operators and prosumers as described in any one of claims 1-6.

8. A computer device, the computer device comprising a memory, a processor, and a computer program, characterized in that, When the computer program is executed by the processor, it implements the collaborative optimization method based on virtual power plant operators and producers / consumers as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the collaborative optimization method based on virtual power plant operators and producers / consumers as described in any one of claims 1-6.

Images