Multi-agent optimal carbon emission permit trading method and system based on distributed iterative algorithm

Abstract

The invention relates to a multi-subject optimal carbon emission permit transaction method and system based on a distributed iterative algorithm, and the method comprises the following steps: 1, calculating the initial carbon quota and actual carbon emission of each subject, and providing a reward and punishment continuous carbon emission permit transaction mode; step 2, based on a reward and punishment continuous carbon emission permit transaction mode, establishing an optimal carbon emission permit transaction model taking a comprehensive energy service provider, an energy concentrator operator and a load aggregator as a main body; and 3, solving and evaluating the optimal carbon emission permit transaction model established in the step 2 by adopting a distributed optimization algorithm based on a KKT condition and convex optimization, and outputting an optimized result. According to the invention, the carbon market can be helped to establish a perfect multi-subject carbon emission permit trading model, the benefit maximization of each trading subject is ensured, and the carbon emission is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of carbon emission rights trading, and relates to a multi-subject optimal carbon emission rights trading method and system, in particular to a multi-subject optimal carbon emission rights trading method and system based on a distributed iterative algorithm. Background Art

[0002] The energy market is changing from a vertically integrated structure to an interactively competitive structure. This change means that the energy market is evolving from a model dominated by a few large enterprises to a model of multi-party participation and market competition. In the traditional centralized optimization method, the goal of energy dispatch is to achieve the optimal overall system. However, this method cannot fully consider the independent interests and behaviors of each market participant.

[0003] Since different entities (such as power generation companies, users, energy storage systems, etc.) have different interests and decision-making methods, centralized methods are difficult to reveal the complex interactions between these entities. In the centralized dispatch model, each entity needs to provide detailed parameter information to the superior dispatch center, such as cost, power generation capacity, electricity demand, etc. This information often involves commercial secrets and sensitive data. Disclosure of this information may lead to the leakage of sensitive information and damage the privacy security of each entity.

[0004] In the existing low-carbon scheduling, since the carbon trading price is fixed, it is easy to cause arbitrage among trading entities, resulting in poor carbon reduction effects. The design of the carbon trading mechanism needs to consider dynamic pricing and market supervision to prevent speculative behavior and ensure the actual effect of carbon emission reduction. In summary, the existing carbon emission rights trading has the following problems:

[0005] (1) Traditional centralized optimization methods cannot simultaneously guarantee the optimal interests of multiple entities in carbon emission trading scenarios. Moreover, centralized scheduling models cannot guarantee the privacy and security of information during the interaction between entities.

[0006] (2) In the existing tiered carbon price model, the transaction price is fixed during the carbon emission rights trading process, which can easily lead to arbitrage behavior among trading entities and is not conducive to reducing carbon emissions.

[0007] After searching, no public documents of the prior art identical or similar to the present invention were found. Summary of the invention

[0008] The purpose of the present invention is to overcome the shortcomings of the prior art and propose a multi-subject optimal carbon emission rights trading system and method based on a distributed iterative algorithm, which can help the carbon market establish a complete multi-subject carbon emission rights trading model, ensure the maximization of the interests of each trading subject, and reduce carbon emissions.

[0009] The present invention solves the practical problem by adopting the following technical solutions:

[0010] A multi-agent optimal carbon emission rights trading method based on a distributed iterative algorithm comprises the following steps:

[0011] Collect equipment parameters of coal-fired units and cogeneration systems, calculate the initial carbon quota and actual carbon emissions of each equipment, and establish a continuous carbon emission trading model with rewards and punishments based on the initial carbon quota and actual carbon emissions.

[0012] Based on the continuous reward and punishment carbon emission trading model, an optimal carbon emission trading model is established with integrated energy service providers, energy hub operators and load aggregators as the main bodies.

[0013] The optimal carbon emission trading model is solved by a distributed optimization algorithm based on Karush-Kuhn-Tucker (KKT) conditions and convex optimization, and the optimal trading strategy is output with the goal of maximizing the total revenue of integrated energy service providers, energy hub operators and load aggregators and minimizing system carbon emissions.

[0014] Furthermore, the specific steps of collecting equipment parameters of coal-fired units and combined heat and power systems, calculating the initial carbon quota and actual carbon emissions of each device, and establishing a reward-and-punishment continuous carbon emission trading model based on the initial carbon quota and actual carbon emissions include:

[0015] Calculate the initial carbon quotas for coal-fired units, combined cooling, heating and power units and external power grid purchases;

[0016] The initial carbon quota is freely allocated according to the rated capacity of the unit, including three parts: coal-fired unit CFPU, cold and hot combined production system CCHP and external power grid purchase, which is determined by the following formulas (1) to (2):

[0017]

[0018] E q =E CFPU +E BUY +E CCHP (2)

[0019] Among them, E CFPU 、E BUY and E CCHP They represent coal-fired units, power purchase from the grid, and CCHP initial free carbon emission allocation quotas; μ e,q and μ h,q They represent the initial carbon emission quota per unit of electricity and heat, which are 0.728t / MWh and 0.102t / GJ respectively; P t BUY represents the amount of electricity purchased by IESP from the power grid; ε represents the conversion coefficient of electrical power into thermal energy; E qA quota for the total carbon emissions of the integrated energy system;

[0020] Calculate the actual carbon emissions of coal-fired units, combined heat and power units, power-to-gas equipment and carbon capture equipment;

[0021] RIES actual carbon emissions include: coal-fired units Gas plants (gas boilers and gas turbines) Carbon emissions from power-to-gas plants and the amount of carbon captured by carbon capture equipment It can be expressed as:

[0022]

[0023] Where P t CFPU Indicates the output power value of the coal-fired unit at time t; P represents the thermal power output of the gas boiler GB at time t; t GT Indicates the power value of the gas turbine GT output at time t. ε GT represents the gas turbine electrical / heat conversion coefficient, η GB Thermal efficiency of gas boiler. η PCU represents the carbon capture rate of the PCU unit; and They represent the actual carbon emission coefficients of coal-fired units, gas-fired devices, P2G equipment and PCU devices, respectively, and are expressed by the following equations (4)-(5):

[0024]

[0025] Among them, in the energy cycle process, μ ue It represents the carbon emission intensity per unit of electricity generated by coal-fired units, K ue Indicates the unit coal consumption value of the generator, L ue Represents the carbon emission intensity of greenhouse gases from coal-fired units, μ ug Indicates the carbon emission intensity of gas-fired units per unit of power generation, L GBT It represents the carbon emission intensity of gas turbines and gas boilers per unit of electricity generation, L GB represents the carbon emission intensity per unit heat output of the gas boiler, α he Indicates the conversion coefficient between thermal energy and electrical energy; in the power plant cycle, μ ce and μ cg They represent the carbon emission coefficient per unit of power generation of coal-fired units and gas-fired units respectively;

[0026] The actual carbon emissions of the integrated energy system are obtained from equations (3) to (5): for:

[0027]

[0028] Establish a continuous carbon emission trading model with rewards and penalties based on the initial carbon quota obtained and the actual carbon emissions obtained;

[0029] The continuous reward and punishment carbon emission trading model is:

[0030]

[0031] Among them, E TC represents the total cost of carbon trading, represents the benchmark carbon price, and s represents the range of carbon emissions. j represents the step carbon price (j = 0, 1, 2, 3), and the calculation method is as shown in formula (8), where δ p Represents the reward and punishment factor.

[0032] Moreover, the optimal carbon emission rights trading model based on the continuous reward and punishment carbon emission rights trading model with integrated energy service providers, energy hub operators and load aggregators as the main body specifically includes:

[0033] Taking integrated energy service providers as leaders, a mathematical model for maximizing the profits of integrated energy service providers is established;

[0034] Taking energy hub operators as followers, a mathematical model for maximizing the benefits of energy hub operators is established;

[0035] Taking load aggregators as followers, a mathematical model for load aggregators to maximize consumer surplus is established.

[0036] Moreover, the specific steps of solving the optimal carbon emission rights trading model by using a distributed optimization algorithm based on Karush-Kuhn-Tucker (KKT) and convex optimization and outputting the optimal trading strategy with the goal of maximizing the total revenue of integrated energy service providers, energy hub operators and load aggregators and minimizing system carbon emissions include:

[0037] Using KKT conditions, the mathematical model of load aggregators is transformed into the conditional constraints of the mathematical model of integrated energy service providers, and the multi-agent optimal carbon emission trading model is transformed into a dual-agent nonlinear programming model;

[0038] The two-agent nonlinear programming is transformed into a two-agent mixed integer quadratic programming model through convex optimization.

[0039] Genetic algorithm is used to solve the optimal carbon emission rights trading strategy between integrated energy service providers and energy hub operators.

[0040] Moreover, the specific method of using KKT conditions to transform the mathematical model of the load aggregator into the conditional constraints of the mathematical model of the comprehensive energy service provider, and transforming the multi-agent optimal carbon emission trading model into a dual-agent nonlinear programming model is as follows:

[0041] The optimization of LA is converted into the constraints of IESP optimization through KKT conditions. The conversion process is as follows:

[0042] Construct LA objective function F LA The Lagrangian function L LA ,Right now

[0043]

[0044] Then the KKT condition is:

[0045]

[0046] The above equations (34), (35) and (36) are the Lagrangian functions L LA Ask about P tra,t , H tra,t and G tra,t The result of the first-order partial derivative being equal to zero. and μ e is the dual variable (i = e, h, g). Equations (35) to (43) are KKT transformations of inequality constraints (31) and (33). It means that at least one of x and y is equal to 0. Equation (44) is the equality constraint (32)

[0047] Moreover, the specific method of converting the dual-agent nonlinear programming into a dual-agent mixed integer quadratic programming model through convex optimization is:

[0048] The constraint (38) is transformed into a mixed integer linear constraint using FA transformation, which is expressed as:

[0049]

[0050] In the formula, represents an integer variable with a value of 0 or 1, and M represents a sufficiently large positive number. Constraints (45) and (46) are equivalent to constraint (38). Similarly, constraints (39) to (43) are converted into mixed integer linear constraints.

[0051] and Using the KKT optimal conditions (35) to (37), the dual variables are converted into quadratic terms, expressed as:

[0052]

[0053] At this time, the nonlinear term in the IESP objective function is only quadratic, and the multi-agent optimal carbon emission trading model is equivalent to the BMIQP problem.

[0054] Moreover, the specific steps of using the genetic algorithm to solve the optimal carbon emission rights trading strategy between the integrated energy service provider and the energy hub operator include:

[0055] Input IESP, EHO and RU information, including parameters such as energy price, carbon quota, energy output / consumption power, etc.

[0056] Establish a low-carbon distributed optimization scheduling model for integrated energy systems based on multi-agent two-level game;

[0057] Model transformation: transforming LA optimization into IESP constraints, and transforming multi-agent master-slave game optimization into dual-agent BNLP based on KKT conditions. Through convex optimization, BNLP is transformed into a dual-agent BMIQP model.

[0058] (4) Using a distributed iterative algorithm to find the equilibrium solution between IESP and EHO;

[0059] First, the EHO population is initialized and the energy price is randomly set;

[0060] Secondly, the equilibrium solution between IESP and LA is solved by IBM CPLEX Optimizer;

[0061] Next, the optimized energy purchase plan of IESP is transmitted to EHO, which calculates the transaction profit based on the genetic algorithm and performs a series of selection, crossover, mutation and other operations until the model equilibrium converges;

[0062] Finally, the optimized results are output.

[0063] A multi-agent optimal carbon emission rights trading system based on a distributed iterative algorithm, comprising:

[0064] The module for establishing a continuous carbon emission trading model with rewards and penalties collects equipment parameters of coal-fired units and combined heat and power systems, calculates the initial carbon quota and actual carbon emissions of each device, and establishes a continuous carbon emission trading model with rewards and penalties based on the initial carbon quota and actual carbon emissions;

[0065] The optimal carbon emission rights trading model establishment module, based on the reward and punishment continuous carbon emission rights trading model, establishes an optimal carbon emission rights trading model with integrated energy service providers, energy hub operators and load aggregators as the main body;

[0066] The solution module uses a distributed optimization algorithm based on Karush-Kuhn-Tucker (KKT) and convex optimization to solve the optimal carbon emission trading model, and outputs the optimal trading strategy with the goal of maximizing the total revenue of integrated energy service providers, energy hub operators and load aggregators and minimizing system carbon emissions.

[0067] Moreover, the optimal carbon emission rights trading model establishment module also includes:

[0068] A mathematical model for maximizing the profits of integrated energy service providers is established with integrated energy service providers as the leaders;

[0069] A mathematical model for maximizing the benefits of energy hub operators is established, with energy hub operators as followers;

[0070] The mathematical model of load aggregator maximizing consumer surplus is established with load aggregators as followers.

[0071] A computer storage medium stores one or more instructions, which when executed by one or more computers cause the one or more computers to implement the method.

[0072] Advantages and beneficial effects of the present invention:

[0073] The present invention proposes a multi-subject optimal carbon emission rights trading method based on a distributed iterative algorithm. First, a reward-and-punishment continuous carbon emission rights trading model is proposed to solve the arbitrage problem of trading entities caused by the fixed trading price of traditional interval carbon emission rights, thereby reducing carbon emissions. Secondly, by constructing a multi-subject distributed optimal carbon emission rights trading model, the interests of each trading entity are maximized. Finally, a distributed algorithm based on KKT conditions and convex optimization is proposed to solve the optimal carbon emission rights trading model, which reduces the need for centralized data sharing, protects sensitive information, and improves the solution speed. The present invention is helpful for the carbon market to establish a complete multi-subject carbon emission rights trading model. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 This is a schematic diagram of the reward and punishment ladder continuous carbon trading model of the present invention;

[0075] Figure 2 The present invention Figure 2 It is a flow chart of the solution process of the multi-agent optimal carbon emission trading model. DETAILED DESCRIPTION

[0076] The embodiments of the present invention are further described in detail below with reference to the accompanying drawings:

[0077] A multi-agent optimal carbon emission rights trading method based on a distributed iterative algorithm comprises the following steps:

[0078] Step 1. Input the initial quota of carbon emissions per unit of electricity / heat, the amount of electricity purchased from the power grid, the conversion coefficient of electric power to heat, the electric power value of coal-fired units, the thermal power value of gas boilers, the electric power value of gas turbines, the electric / heat conversion coefficient of gas turbines, the thermal efficiency of gas boilers, the carbon capture rate, the actual carbon emission coefficient, the carbon emission intensity per unit of power generation of coal-fired units, the unit coal consumption value of generators, the carbon emission intensity of greenhouse gases of coal-fired units, the carbon emission intensity per unit of power generation of gas units, the carbon emission intensity per unit of power generation of gas turbines and gas boilers, the carbon emission intensity per unit of heat output of gas boilers, the conversion coefficient between heat and electricity, the carbon emission coefficient per unit of power generation of coal-fired units and gas units, and other parameters, calculate the initial carbon quota and actual carbon emissions of each entity, and establish a continuous reward and punishment carbon emission rights trading model.

[0079] The specific steps of step 1 include:

[0080] (1) Calculate the initial carbon quotas for coal-fired units, combined cooling, heating and power units, and electricity purchased from external power grids;

[0081] In the optimal carbon emission trading model, the initial carbon quota is freely allocated according to the rated capacity of the unit, including three parts: coal-fired unit CFPU, cold and hot spot combined production system CCHP and external power grid purchase, which is determined by the following formulas (1) to (2):

[0082]

[0083] E q =E CFPU +E BUY +E CCHP (2)

[0084] Among them, E CFPU 、E BUY and E CCHP They represent coal-fired units, power purchase from the grid, and CCHP initial free carbon emission allocation quotas; μ e,q and μ h,q They represent the initial carbon emission quota per unit of electricity and heat, which are 0.728t / MWh and 0.102t / GJ respectively; P t BUY represents the amount of electricity purchased by IESP from the power grid; ε represents the conversion coefficient of electrical power into thermal energy; E q It is the total carbon emission quota of the integrated energy system.

[0085] (2) Calculate the actual carbon emissions from coal-fired units, combined heat and power units, power-to-gas equipment, and carbon capture equipment;

[0086] RIES actual carbon emissions include: coal-fired units Gas plants (gas boilers and gas turbines) Carbon emissions from power-to-gas plants and the amount of carbon captured by carbon capture equipment It can be expressed as:

[0087]

[0088] Where P t CFPU Indicates the output power value of the coal-fired unit at time t; P represents the thermal power output of the gas boiler GB at time t; t GT Indicates the power value of the gas turbine GT output at time t. ε GT represents the gas turbine electrical / heat conversion coefficient, η GB Thermal efficiency of gas boiler. η PCU Represents the carbon capture rate of the PCU device. and They represent the actual carbon emission coefficients of coal-fired units, gas-fired devices, P2G equipment and PCU devices, respectively, and are expressed by the following equations (4)-(5):

[0089]

[0090] Among them, in the energy cycle process, μ ue It represents the carbon emission intensity per unit of electricity generated by coal-fired units, K ue Indicates the unit coal consumption value of the generator, L ue Represents the carbon emission intensity of greenhouse gases from coal-fired units, μ ug Indicates the carbon emission intensity of gas-fired units per unit of power generation, L GBT It represents the carbon emission intensity of gas turbines and gas boilers per unit of electricity generation, L GB represents the carbon emission intensity per unit heat output of the gas boiler, α he Indicates the conversion coefficient between thermal energy and electrical energy; in the power plant cycle, μ ce and μ cg They represent the carbon emission coefficients per unit of power generation of coal-fired units and gas-fired units respectively.

[0091] The actual carbon emissions of the integrated energy system can be obtained from equations (3) to (5): for:

[0092]

[0093] (3) Based on the initial carbon quota obtained in step (1) and the actual carbon emissions obtained in step (2), a reward-and-punishment continuous carbon emission rights trading model is established;

[0094] Based on the traditional carbon trading model, a carbon trading model with continuous reward and punishment ladder is proposed, such as Figure 1 shown.

[0095] The continuous reward and punishment carbon emission trading model is:

[0096]

[0097] Among them, E TC represents the total cost of carbon trading, represents the benchmark carbon price, and s represents the range of carbon emissions. j represents the step carbon price (j = 0, 1, 2, 3), and the calculation method is as shown in formula (8), where δ p Represents the reward and punishment factor; E tc Represents the carbon emission rights trading price.

[0098] The continuous carbon trading model with reward and punishment tiers can further guide various entities to reduce carbon emissions through price factors, ensure the rationality of carbon emissions from each entity's equipment, and avoid the problem of increased operating costs due to blind carbon reduction.

[0099] Step 2: Based on the reward-and-punishment continuous carbon emission trading model obtained in step 1, an optimal carbon emission trading model with integrated energy service providers, energy hub operators and load aggregators as the main body is established;

[0100] The specific steps of step 2 include:

[0101] (1) Taking integrated energy service providers as leaders, a mathematical model for maximizing the profits of integrated energy service providers is established;

[0102] The integrated energy service provider IESP comprehensively evaluates the output plan of the energy hub operator EHO and the demand of the load aggregator RU and formulates a pricing strategy. The optimization goal is to maximize the profit, that is,

[0103]

[0104] Among them, in the period t, and They represent the revenue from energy sales to LA, the cost of purchasing energy from EHO, the cost of purchasing electricity from the grid, and the penalty cost of energy interruption. The specific calculation method of each variable in formula (9) is as follows:

[0105]

[0106] In the formula, T represents the optimization scheduling period (24 hours), P t ue , and They represent the LA electric and thermal load power and natural gas consumption at time t respectively; P t oe , and They represent the EHO electrical and thermal output power and natural gas output at time t respectively; and They represent the prices of electricity, heat and natural gas sold to LA at time t respectively; and They represent the energy prices of electricity, heat and natural gas purchased from EHO at time t respectively; and Respectively represent the power sales price and on-grid power price of the power grid; δ p Represents the penalty cost coefficient for interrupting heating and natural gas supply.

[0107] Constraints: To ensure the coordinated operation of multiple entities and avoid direct transactions between EHO and LA, the energy price purchased (sold) by IESP must be higher (lower) than the market price, and the electricity and heat sales prices during the transaction period must not be greater than the maximum average electricity and heat sales prices, that is, constraints (14) to (16) must be met.

[0108]

[0109] in, and Indicates the minimum and maximum limit prices for thermal energy transactions. and Indicates the maximum average electricity and heat selling prices.

[0110] (2) Taking the energy hub operator as a follower, a mathematical model for maximizing the revenue of the energy hub operator is established;

[0111] Energy hub operators EHO include: coal-fired power units (CFPU), wind turbines (WT), photovoltaics (PV) and combined cooling, heating and power (CCHP) systems with carbon capture;

[0112] Among them, CCHP includes: gas turbine, boiler, purification compression unit (PCU), power to gas (P2G), absorption refrigerator AR, energy storage device;

[0113] EHO sets the purchase price of electricity, heat and gas according to IESP to optimize the energy output of equipment, taking into account the carbon trading cost, and its optimization goal is to obtain the maximum benefit, that is,

[0114]

[0115] in, Represents the revenue from electricity sales to IESP. represents the fuel cost, and represents the fuel cost of CCHP and CFPU;

[0116] According to the variable operating efficiency characteristics of coal-fired units, gas turbines and gas boilers, it is expressed as a linear or quadratic function of output power and fuel price:

[0117]

[0118] In the formula, are the cost coefficients of coal-fired units, gas turbines and gas boilers, and the corresponding output power and thermal power at time t are P t CFPU , P t GT and

[0119] In EHO, the fuel cost of clean energy (wind power, photovoltaic power) and auxiliary devices such as power-to-gas equipment and heat storage tanks is considered to be zero; photovoltaic power and wind power are greatly affected by environmental factors, and their output can be predicted based on historical data. In order to reduce carbon emissions and improve energy efficiency, it is assumed that all the electricity generated by wind power and photovoltaic power is consumed, and there is no abandonment of light or wind; at time t, the EHO electrical power output, thermal power output and natural gas output can be expressed as:

[0120]

[0121] Among them, the "+" in front of the battery (BT), heat storage tank (HST) and (gas tank) indicates output energy, and the "-" indicates absorption energy. t WT , P t PV is the wind power and photovoltaic output value, P t PCU , P t P2G , P t EB Respectively represent PCU, P2G, EB input power, P t BT Indicates the battery charge / discharge power. is the output thermal power of the electric boiler, is the output thermal power of the gas boiler, represents the AR input thermal power, Indicates the heat storage / release power of the heat storage tank. and Indicates the output natural gas volume of the power-to-gas equipment and gas source, Indicates the amount of natural gas stored / released in the gas tank.

[0122] Constraints:

[0123] Coal-fired unit P t CFPU and gas boiler The following constraints must be met:

[0124]

[0125] in, and are the rated power generation / heat capacity of coal-fired units and gas boilers, respectively, GB is the input natural gas volume, η GB It is the gas / heat conversion efficiency of gas boiler.

[0126] The gas turbine operating constraints are as follows:

[0127]

[0128] Where D GT It represents the relationship between gas turbine natural gas consumption and output power, W GT It indicates the on / off status of the gas turbine, and μ1, μ2 and μ3 indicate the gas combustion coefficient. and Indicates the maximum and minimum values ​​of the gas turbine output power, and Indicates the gas turbine climbing power limit.

[0129] The operating constraints of electric boilers are:

[0130]

[0131] In the formula, and P t EB Respectively represent the thermal power and electrical power output of the electric boiler, η EB is its electrical / thermal conversion coefficient, Indicates its rated electrical power.

[0132] The operating constraints of the absorption chiller are:

[0133]

[0134] In the formula, and Respectively represent the output cooling power and input heating power of the absorption refrigerator, η AR is its heat / cold conversion coefficient, Indicates the maximum value of its input thermal power.

[0135] The operating conditions of the power-to-gas equipment are:

[0136]

[0137] In the formula, Output of natural gas for P2G equipment, η P2G is the gas / electricity conversion coefficient, P t P2G Input power to it.

[0138] The battery operation constraints are as follows:

[0139]

[0140] in, The charging (discharging) power of the battery; and It is the minimum and maximum value of battery charging (discharging) power; It is the mark of battery charging (discharging) status; S BT is the battery charge state; and It is the upper and lower limits of charge and discharge power; is the charge (discharge) increment; and Indicates the upper and lower limits of the climbing power in the charging (discharging) state.

[0141] The operating constraints of the heat storage tank are as follows:

[0142]

[0143] in, The heat storage (release) power of the heat storage tank; and The minimum and maximum values ​​of the heat storage (discharge) power of the heat storage tank; It is the mark position of heat storage tank heat storage (release) state; S HST Store heat in the heat storage tank; and The upper and lower limits of the heat storage capacity of the heat storage tank; is the heat storage (release) increment; and Indicates the upper and lower limits of the ramp power in the heat storage (discharge) state. The operating constraints of the gas storage equipment are similar to those of the heat storage tank.

[0144] The operating constraints of clean energy units are:

[0145]

[0146] In the formula, and They are the maximum outputs of wind power and photovoltaic power respectively.

[0147] (3) Taking load aggregators as followers, a mathematical model for maximizing consumer surplus by load aggregators is established.

[0148] Under the energy price conditions given by IESP, load aggregators achieve the optimization goal by adjusting their own shiftable load and reducible load to maximize consumer surplus.

[0149] Maximizing consumer surplus is represented by the user utility function, energy cost, and production reduction incentive, that is,

[0150]

[0151] Among them, f u,t It indicates the user's satisfaction with the performance, expressed as a quadratic function, that is,

[0152]

[0153] Among them, α e and β e , α h and β h , α g and β g They respectively represent the user's preference coefficients for the use of different energy sources, and determine the user's demand for electricity, heat and gas energy.

[0154] LA electric load power P t ue By fixed load P fix,t and the translatable load P tra,t Decision, that is

[0155] P t ue =P fix,t +P tra,t (30)

[0156] Among them, P fix,t The electricity consumption time is relatively fixed, and the reliability of electricity is required to be high. It is generally used for life, production and other activities; tra,t The electricity consumption time is relatively flexible and is determined according to the change of electricity price, such as charging load. Its constraints are:

[0157]

[0158] In the formula, is the maximum value of the load power that can be translated; Z tra It represents the total amount of shifted load power in time period T, that is, the total shifted load power value is constant before and after the demand side response.

[0159] LA heat load power (natural gas load) It also contains two parts, namely

[0160]

[0161] Among them, H fix,t (G fix,t ) is the minimum thermal power (natural gas load) requirement; H tra,t (G tra,t ) represents the heat power that can be reduced (natural gas load), which is used to adjust the required heat power reduction according to the satisfaction of heat use, and the constraints that need to be met are

[0162]

[0163] in, It indicates the maximum value of thermal power (natural gas load) that can be reduced at time t.

[0164] Step 3: Solve and evaluate the optimal carbon emission trading model established in step 2 based on the distributed optimization algorithm based on KKT conditions and convex optimization.

[0165] The present invention proposes a method for solving the multi-agent optimal carbon emission trading model. First, the optimization of LA is converted into the constraints of IESP, and the multi-agent master-slave game optimization is converted into a dual-agent nonlinear programming (BNLP) based on KKT (Karush-Kuhn-Tucker) conditions. Secondly, BNLP is converted into a bi-level mixed-integer quadratic programming (BMIQP) model through convex optimization. Finally, a distributed iterative algorithm is used to solve the equilibrium solution between IESP and EHO.

[0166] The specific steps of step 3 include:

[0167] (1) Using KKT conditions, the mathematical model of the load aggregator is transformed into the conditional constraints of the mathematical model of the integrated energy service provider, and the multi-agent optimal carbon emission trading model is transformed into a dual-agent nonlinear programming model;

[0168] The optimization of LA is converted into the constraints of IESP optimization through KKT conditions. The conversion process is as follows:

[0169] Construct LA objective function F LA The Lagrangian function L LA ,Right now

[0170]

[0171] Then the KKT condition is:

[0172]

[0173] The above equations (34), (35) and (36) are the Lagrangian functions L LA Ask about P tra,t , H tra,t and G tra,t The result of the first-order partial derivative being equal to zero. and μ e is the dual variable (i = e, h, g). Equations (35) to (43) are KKT transformations of inequality constraints (31) and (33). It means that at least one of x and y is equal to 0. Equation (44) is the equality constraint (32).

[0174] (2) The dual-agent nonlinear programming is transformed into a dual-agent mixed integer quadratic programming model through convex optimization.

[0175] The dual constraints (38) to (43) are nonlinear, and in (10) and It is a bilinear form, which makes it difficult to directly solve the optimization objective. Therefore, the FA (Fortuny-Amat) transformation is used to transform the constraint (38) into a mixed integer linear constraint, which is expressed as:

[0176]

[0177] In the formula, represents an integer variable with a value of 0 or 1, and M represents a sufficiently large positive number. Constraints (45) and (46) are equivalent to constraint (38). Similarly, constraints (39) to (43) can be converted into mixed integer linear constraints.

[0178] and Using the KKT optimal conditions (35) to (37), the dual variables are converted into quadratic terms, expressed as:

[0179]

[0180] At this time, the nonlinear term in the IESP objective function is only quadratic, and the multi-agent optimal carbon emission trading model is equivalent to the BMIQP problem.

[0181] (3) Genetic algorithm is used to solve the optimal carbon emission trading strategy between integrated energy service providers and energy hub operators.

[0182] The detailed solution process of the proposed two-layer dynamic game model is as follows: Figure 2 As shown, the following steps are included:

[0183] (1) Input IESP, EHO and RU information, including parameters such as energy price, carbon quota, energy output / consumption power, etc.

[0184] (2) Establish a low-carbon distributed optimization scheduling model for an integrated energy system based on a multi-agent two-level game;

[0185] (3) Model transformation: transform the optimization of LA into the constraints of IESP, and transform the multi-agent master-slave game optimization into a two-agent BNLP based on KKT conditions. Then, BNLP is transformed into a two-agent BMIQP model through convex optimization.

[0186] (4) Using a distributed iterative algorithm to find the equilibrium solution between IESP and EHO;

[0187] First, the EHO population is initialized and the energy price is randomly set;

[0188] Secondly, the equilibrium solution between IESP and LA is solved by IBM CPLEX Optimizer;

[0189] Next, the optimized energy purchase plan of IESP is transmitted to EHO, which calculates the transaction profit based on the genetic algorithm and performs a series of selection, crossover, mutation and other operations until the model equilibrium converges;

[0190] Finally, the optimized results are output.

[0191] A multi-agent optimal carbon emission rights trading system based on a distributed iterative algorithm, comprising:

[0192] A module for establishing a continuous carbon emission trading model with rewards and penalties, which calculates the initial carbon quota and actual carbon emissions of each entity and establishes a continuous carbon emission trading model with rewards and penalties;

[0193] The optimal carbon emission rights trading model establishment module, based on the reward and punishment continuous carbon emission rights trading model, establishes an optimal carbon emission rights trading model with integrated energy service providers, energy hub operators and load aggregators as the main body;

[0194] The solution module uses a distributed optimization algorithm based on KKT conditions and convex optimization to solve and evaluate the optimal carbon emission trading model established in step 2, and outputs the optimized results.

[0195] The optimal carbon emission rights trading model establishment module also includes:

[0196] A mathematical model for maximizing the profits of integrated energy service providers is established with integrated energy service providers as the leaders;

[0197] A mathematical model for maximizing the benefits of energy hub operators is established, with energy hub operators as followers;

[0198] The mathematical model of load aggregator maximizing consumer surplus is established with load aggregators as followers.

[0199] A computer storage medium stores one or more instructions, which when executed by one or more computers cause the one or more computers to implement the method.

[0200] In this embodiment, in order to analyze and verify the superiority of the proposed multi-agent optimal carbon emission trading model, a low-carbon regional comprehensive energy system model was built based on the energy data of a certain industrial park, and 4 different simulation scenarios were set for analysis and comparison, as shown in Table 1.

[0201] The scheduling results of IESP, EHO and LA of the game players in each scenario are shown in Table 2. It can be seen that the proposed multi-agent optimal carbon emission rights trading model based on distributed iterative algorithm can increase the benefits of each subject and reduce the overall carbon emissions.

[0202] Table 1 Different simulation scheduling scenarios

[0203]

[0204] Table 2 Comparison of solution results in various scenarios

[0205]

[0206] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0207] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0208] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0209] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

Claims

1. A multi-agent optimal carbon emission rights trading method based on a distributed iterative algorithm, characterized by: The following steps are involved: Collect equipment parameters of coal-fired units and combined heat and power systems, calculate the initial carbon quota and actual carbon emissions of each device, and establish a reward-and-punishment continuous carbon emission rights trading model based on the initial carbon quota and actual carbon emissions; Based on the continuous reward and punishment carbon emission trading model, an optimal carbon emission trading model with integrated energy service providers, energy hub operators and load aggregators as the main body is established; The optimal carbon emission trading model is solved by a distributed optimization algorithm based on the Carlo-Kuhn-Tucker condition (KKT) and convex optimization, and the optimal trading strategy is output with the goal of maximizing the total revenue of integrated energy service providers, energy hub operators and load aggregators and minimizing system carbon emissions.

2. According to claim 1, a multi-agent optimal carbon emission rights trading method based on a distributed iterative algorithm is characterized by: The specific steps of collecting equipment parameters of coal-fired units and combined heat and power systems, calculating the initial carbon quota and actual carbon emissions of each device, and establishing a reward-and-punishment continuous carbon emission rights trading model based on the initial carbon quota and actual carbon emissions include: Calculate the initial carbon quotas for coal-fired units, combined cooling, heating and power units and external power grid purchases; The initial carbon quota is freely allocated according to the rated capacity of the unit, including three parts: coal-fired unit CFPU, cold and hot combined production system CCHP and external power grid purchase, which is determined by the following formulas (1) to (2): AND q =And CFPU +E BUY +E CCHP (2) Among them, E CFPU 、E BUY and E CCHP They represent coal-fired units, power purchase from the grid, and CCHP initial free carbon emission allocation quotas; μ e,q and μ h,q They represent the initial carbon emission quota per unit of electricity and heat, which are 0.728t / MWh and 0.102t / GJ respectively; P t BUY represents the amount of electricity purchased by IESP from the power grid; ε represents the conversion coefficient of electrical power into thermal energy; E q A quota for the total carbon emissions of the integrated energy system; Calculate the actual carbon emissions of coal-fired units, combined heat and power units, power-to-gas equipment and carbon capture equipment; RIES actual carbon emissions include: coal-fired units Gas plants (gas boilers and gas turbines) Carbon emissions from power-to-gas plants and the amount of carbon captured by carbon capture equipment It can be expressed as: Where P t CFPU Indicates the output power value of the coal-fired unit at time t; P represents the thermal power output of the gas boiler GB at time t; t GT Represents the power value of the gas turbine GT output at time t; ε GT represents the gas turbine electrical / heat conversion coefficient, η GB Thermal efficiency of gas boiler; η PCU represents the carbon capture rate of the PCU unit; and They represent the actual carbon emission coefficients of coal-fired units, gas-fired devices, P2G equipment and PCU devices, respectively, and are expressed by the following equations (4)-(5): Among them, in the energy cycle process, μ ue It represents the carbon emission intensity per unit of electricity generated by coal-fired units, K ue Indicates the unit coal consumption value of the generator, L ue Represents the carbon emission intensity of greenhouse gases from coal-fired units, μ ug Indicates the carbon emission intensity of gas-fired units per unit of power generation, L GBT It represents the carbon emission intensity of gas turbines and gas boilers per unit of electricity generation, L GB represents the carbon emission intensity per unit heat output of the gas boiler, α he Indicates the conversion coefficient between thermal energy and electrical energy; in the power plant cycle, μ ce and μ cg They represent the carbon emission coefficient per unit of power generation of coal-fired units and gas-fired units respectively; The actual carbon emissions of the integrated energy system are obtained from equations (3) to (5): for: Establish a continuous carbon emission trading model with rewards and penalties based on the initial carbon quota obtained and the actual carbon emissions obtained; The reward and punishment continuous carbon emission trading model is: Among them, E TC represents the total cost of carbon trading, represents the benchmark carbon price, s represents the range of carbon emissions; E j represents the step carbon price (j = 0, 1, 2, 3), and the calculation method is as shown in formula (8), where δ p Represents the reward and punishment factor.

3. According to claim 1, a multi-agent optimal carbon emission rights trading method based on a distributed iterative algorithm is characterized by: The optimal carbon emission rights trading model based on the continuous reward and punishment carbon emission rights trading model, which is mainly composed of integrated energy service providers, energy hub operators and load aggregators, includes the following three models: Taking integrated energy service providers as leaders, a mathematical model for maximizing the profits of integrated energy service providers is established; Taking energy hub operators as followers, a mathematical model for maximizing the benefits of energy hub operators is established; Taking load aggregators as followers, a mathematical model for load aggregators to maximize consumer surplus is established.

4. According to claim 1, a multi-agent optimal carbon emission rights trading method based on a distributed iterative algorithm is characterized by: The specific steps of solving the optimal carbon emission rights trading model by using a distributed optimization algorithm based on Karush-Kuhn-Tucker (KKT) and convex optimization and outputting the optimal trading strategy with the goal of maximizing the total revenue of integrated energy service providers, energy hub operators and load aggregators and minimizing system carbon emissions include: Using the Karush-Kuhn-Tucker (KKT) condition, the mathematical model of the load aggregator is transformed into the conditional constraints of the mathematical model of the integrated energy service provider, and the multi-agent optimal carbon emission trading model is transformed into a dual-agent nonlinear programming model. The two-agent nonlinear programming is transformed into a two-agent mixed integer quadratic programming model through convex optimization. Genetic algorithm is used to solve the optimal carbon emission rights trading strategy between integrated energy service providers and energy hub operators.

5. According to claim 4, a multi-agent optimal carbon emission rights trading method based on a distributed iterative algorithm is characterized by: The specific method of using the Karush-Kuhn-Tucker (KKT) condition to transform the mathematical model of the load aggregator into the conditional constraints of the mathematical model of the comprehensive energy service provider, and transforming the multi-agent optimal carbon emission rights trading model into a dual-agent nonlinear programming model is as follows: The optimization of load aggregators LA is converted into the optimization constraints of integrated energy service providers IESP (integrated energy service provider) through KKT conditions. The conversion process is as follows: Construct LA objective function F LA The Lagrangian function L LA ,Right now Then the KKT condition is: The above equations (34), (35) and (36) are the Lagrangian functions L LA Ask about P tra,t , H tra,t and G tra,t The result of the first-order partial derivative being equal to zero; and μ e is the dual variable (i = e, h, g); Equations (35) to (43) are the KKT transformations of the inequality constraints (31) and (33); the expression It means that at least one of x and y is equal to 0; equation (44) is the equality constraint (32).

6. The multi-agent optimal carbon emission rights trading method based on a distributed iterative algorithm according to claim 4 is characterized by: The specific method of converting the dual-agent nonlinear programming into a dual-agent mixed integer quadratic programming model through convex optimization is: The constraint (38) is transformed into a mixed integer linear constraint using FA transformation, which is expressed as: In the formula, represents an integer variable with a value of 0 or 1, and M represents a sufficiently large positive number; constraints (45) and (46) are equivalent to constraint (38). Similarly, constraints (39) to (43) are converted into mixed integer linear constraints; and Using the KKT optimal conditions (35) to (37), the dual variables are converted into quadratic terms, expressed as: At this time, the nonlinear term in the IESP objective function is only quadratic, and the multi-agent optimal carbon emission trading model is equivalent to the BMIQP problem.

7. The multi-agent optimal carbon emission rights trading method based on a distributed iterative algorithm according to claim 4 is characterized by: The specific steps of using genetic algorithm to solve the optimal carbon emission rights trading strategy between the integrated energy service provider and the energy hub operator include: Input IESP, EHO and RU information, including parameters of energy price, carbon quota, energy output / consumption power; Establish a low-carbon distributed optimization scheduling model for integrated energy systems based on multi-agent two-level game; Model transformation: transforming LA optimization into IESP constraints, and transforming multi-agent master-slave game optimization into dual-agent BNLP based on KKT conditions. Through convex optimization, BNLP is transformed into a dual-agent BMIQP model. A distributed iterative algorithm is used to find the equilibrium solution between IESP and EHO; First, the EHO population is initialized and the energy price is randomly set; Secondly, the equilibrium solution between IESP and LA is solved by IBM CPLEX Optimizer; Next, the optimized energy purchase plan of IESP is transmitted to EHO, which calculates the transaction profit based on the genetic algorithm and performs a series of selection, crossover and mutation operations until the model equilibrium converges; Finally, the optimized results are output.

8. A multi-agent optimal carbon emission trading system based on a distributed iterative algorithm, characterized by: include: The module for establishing a continuous carbon emission trading model with rewards and penalties collects equipment parameters of coal-fired units and combined heat and power systems, calculates the initial carbon quota and actual carbon emissions of each device, and establishes a continuous carbon emission trading model with rewards and penalties based on the initial carbon quota and actual carbon emissions; The optimal carbon emission rights trading model establishment module, based on the reward and punishment continuous carbon emission rights trading model, establishes an optimal carbon emission rights trading model with integrated energy service providers, energy hub operators and load aggregators as the main body; The solution module uses a distributed optimization algorithm based on Karush-Kuhn-Tucker (KKT) and convex optimization to solve the optimal carbon emission trading model, and outputs the optimal trading strategy with the goal of maximizing the total revenue of integrated energy service providers, energy hub operators and load aggregators and minimizing system carbon emissions.

9. The multi-agent optimal carbon emission rights trading system based on a distributed iterative algorithm according to claim 8 is characterized by: The optimal carbon emission rights trading model establishment module also includes: A mathematical model for maximizing the profits of integrated energy service providers is established with integrated energy service providers as the leaders; A mathematical model for maximizing the benefits of energy hub operators is established, with energy hub operators as followers; The mathematical model of load aggregator maximizing consumer surplus is established with load aggregators as followers.

10. A computer storage medium storing one or more instructions which, when executed by one or more computers, cause the one or more computers to implement the method of claims 1 to 7.

Images