Method and device for calculating consumption responsibility weight of power market subject in urban area range

Abstract

The invention discloses a method for calculating the consumption responsibility weight of an electric power market subject in a city domain range, and the method comprises the following steps: constructing an electric power market subject consumption responsibility weight limit model which is a double-layer master-slave game model, the upper layer of the game model is a provincial electric power company and a provincial electric power market transaction center including green certificate transaction, and is responsible for scheduling electric quantity and green certificate distribution of a plurality of electric power system areas of the whole province; the optimization objective of the upper layer of the game is to ensure that the total social remain of the provincial electricity market in the transaction process is maximum in the electricity purchasing and selling process; the optimization objective of the lower layer of the game model is to reduce the power generation cost of each region and minimize the power generation cost; carrying out accelerated solution on the model through a Kriging model algorithm; the invention also discloses a method.

Description

Technical Field

[0001] The present invention belongs to the field of power market subject consumption methods, and in particular to a method and device for calculating the consumption responsibility weights of power market subjects within a city area. Background Art

[0002] Improvements in electricity market mechanisms can play an important role in reducing carbon emissions, promoting low-carbon development, meeting the needs of building new power systems, and increasing the proportion of renewable energy consumption. Among them, the green certificate mechanism based on the responsibility weight of renewable energy consumption has attracted people's attention.

[0003] The renewable energy consumption responsibility weight system plays a very important role in improving the green certificate consumption ratio. The renewable energy consumption responsibility weight is currently only at the provincial market stage, and has not further entered every market entity. The clearing process of the power market is the process of power distribution. In the traditional power market optimization and dispatching, the clearing process of the power market is often solved using game theory. Commonly used game theory models include Nash equilibrium theory, Stackelberg game theory, evolutionary game theory, etc. The Nash equilibrium solution obtained by solving the game theory model can guide the optimal power distribution under fixed conditions.

[0004] The present invention takes into account the market scenario in which the responsibility weight of renewable energy consumption needs to be quickly calculated during the market clearing process, and puts the constraint conditions corresponding to the responsibility weight into the clearing of renewable energy for actual consideration, thereby obtaining the corresponding clearing calculation results. Summary of the invention

[0005] The present invention adopts the following technical solutions:

[0006] A method for calculating the consumption responsibility weight of power market entities within the city area, the method comprising the following steps:

[0007] Construct a power market subject consumption responsibility weight limit model, which is a two-layer master-slave game model. The upper layer of the game model is the provincial power company and the provincial power market trading center including green certificate trading, which is responsible for the dispatch of electricity and green certificate allocation in several power system areas of the province; the optimization goal of the upper layer of the game is to ensure the maximum total social surplus in the provincial power market during the transaction process during the purchase and sale of electricity; the optimization goal of the lower layer of the game model is to reduce the power generation cost of each area and minimize the power generation cost;

[0008] The model is solved quickly using the Kriging model algorithm.

[0009] Furthermore, the upper objective function of the game model is:

[0010]

[0011] in, It is the power that the city's power supply companies and power trading institutions purchase from outside the city. It is the power of electricity sold outside the city. is the power purchased by the region from the power supply company. is the power sold by the region to the power supply company, are their corresponding prices respectively; in addition, regarding green certificates, is the price of green certificates purchased from outside the city. is the price of green certificates sold within the city;

[0012] The corresponding constraints are:

[0013] (1) Meeting power balance conditions:

[0014]

[0015] (2) Restrictions on purchasing or selling electricity from within the city:

[0016] z 01,t +z 02,t ≤1 (26)

[0017] Among them, z 01,t 、z 02,t A 0-1 variable representing the amount of power a region buys and sells to a power company, satisfying the corresponding upper and lower bound constraints:

[0018]

[0019]

[0020] In order to ensure that the electricity market players can trade with the power company instead of bypassing the power company to buy electricity from the power market outside the province, and to ensure the continuous profitability of the power supply bureau, the electricity prices set by the power market area are stipulated to meet the following requirements:

[0021]

[0022] Through such operations, market players within the city area, through the city area power company or agent, conduct corresponding transactions with the outside world when the power supply in the area is insufficient or excessive, and include the corresponding costs in the corresponding total social surplus;

[0023] In addition, in order to meet the city's power consumption responsibility constraints, green certificate requirements are stipulated for power consumption within the city. At this time, the city power company or trading center still needs to dispatch electricity, and the corresponding renewable energy consumption responsibility weight is selected to meet the corresponding regional consumption responsibility weight limit:

[0024]

[0025] Furthermore,

[0026] The objective function of the lower layer of the game model is:

[0027] minC total =C bs +C tr +C comp +C cert (31)

[0028] The internal costs of the region include the cost of purchasing and selling electricity C bp , the cost of traditional energy generation C tr , and the compensation cost of demand-side response C comp 、Green Certificate Purchase Cost C cert The sum of four items;

[0029] Among them, the cost of purchasing and selling electricity C bp :

[0030]

[0031] Traditional energy generation cost C tr :

[0032]

[0033] Compensation costs for demand-side response:

[0034]

[0035] Green certificate purchase cost:

[0036]

[0037] The following constraints are met:

[0038] Constraints for purchasing and selling electricity at different times:

[0039] z 1,t +z 2,t ≤1 (36)

[0040] Upper limit of electricity purchase and sale:

[0041]

[0042] Power balance constraints:

[0043]

[0044] Among them, λ bp,t The electricity purchase and sale price at each moment, The amount of electricity purchased from the power company. P is the amount of electricity sold to the power company. wd,t is the output of wind turbine, P pv,t Output power of photovoltaic unit, P l,t is the load power, z 1,t is the electricity purchase variable for each period, z 2,t is the electricity sales variable for each period, is the upper limit of transactions with the power grid company per unit time; w t is the unit working status, u t is the unit startup action variable, v t is the unit shutdown action variable, P tr is the output of the traditional energy generator set, and R is the ramp rate;

[0045] At time t = 1, it satisfies:

[0046] u1=w1-w T (40)

[0047] v1=w1-w T (41)

[0048] When t≥2,

[0049] u t =w t -w t-1 (t=2,…,T) (42)

[0050] Define load constraints:

[0051] 0≤P cut,i,t ≤r cut,l,max P l,t (43)

[0052] P cut,i,t +P cut,i,t-1 ≤r cut,l,jointmax P l,t (44)

[0053] Among them, P l,t is the load, P cut,i,t For the load that can be reduced, r cut,max is the maximum reduction rate of a single unit, r cut,jointmax is the combined maximum reduction rate of the units;

[0054] At the same time, the renewable energy consumption responsibility weight constraints are met:

[0055]

[0056] q cert,t +P wd,t +P pv,t ≥γL total,t (46)

[0057] Among them, γ is the quota ratio, L total,t is the total load taking into account the demand side response, q cert,t The purchase amount of green certificates.

[0058] Furthermore, the method for accelerating the solution of the model by using the Kriging model algorithm is:

[0059] First, in the master-slave two-level programming problem, the problem can be divided into two independent parts, the upper layer and the lower layer. When solving the problem, the master problem and the slave problem can be separated, and the optimal solution of the slave problem is solved first, that is, the Latin hypercube sampling method is first used to equally extract the variables of the slave problem in the feasible domain of the slave problem, and then the variables of the slave problem are brought into the slave problem for calculation to obtain the values ​​of the slave problem, and then brought into the master problem for calculation optimization, and the independent variable-dependent variable mapping of the entire master-slave problem is formed; next, in order to improve the efficiency of the solution, these mappings are divided into several small areas according to the values ​​of the independent variables. In each small area, for the mapping set of each area, the Kriging proxy function is used to approximate such corresponding mappings, and the particle swarm optimization algorithm is used to find the maximum value; finally, the optimal values ​​obtained by optimizing all local areas are summarized together, and the optimal value is obtained as the optimal solution of the master-slave game method.

[0060] Furthermore, for the two-layer master-slave game model, the steps of solving the algorithm are:

[0061] (1) Initialize the lower-level model variable parameters, including the on-grid electricity price in the power market and the renewable energy consumption responsibility weight; these parameters can be obtained through Latin hypercube sampling, so that the sampling points are distributed as evenly as possible in the region of the independent variable, thereby accurately reflecting the overall situation of the whole system;

[0062] (2) Substitute the unknown parameters of the power market grid price and renewable energy consumption responsibility weight in step (1) into the lower model to solve the corresponding value of the problem, such as the corresponding transaction volume; then, substitute the corresponding value of the problem with the corresponding constraints into the upper model to directly obtain the feasible value of the corresponding upper model;

[0063] (3) Establish a mapping relationship between the power market operation conditions obtained in the main problem in step (2) and the independent variable parameters in step (1), such as the electricity price of each market entity;

[0064] (4) According to the feasible domain of the variables of the lower model, the dependent variables of the sampling points are divided into several parts according to the density clustering algorithm, and a Kriging model is generated for the samples corresponding to each part of the parameter lower model to represent the level of the function value of the part;

[0065] (5) Perform particle swarm optimization on each part generated in (4) to find the maximum value of each area as the maximum value of the overall model;

[0066] (6) Rank the maximum values ​​obtained in each region in (5) to obtain the overall optimal value, which is the optimal solution to the master-slave game model.

[0067] Furthermore, when the Kriging model is used to solve the double-layer master-slave game model, it specifically includes:

[0068] 1) Adding points: When the electricity price of each group and the output power of each unit in the known area are known, select the known points in the area and let these points move according to the rules of the particle swarm algorithm to obtain new points. The new points are brought into the model to obtain approximate values, and the fitting function of the Kriging agent model is trained and updated accordingly to obtain the local optimal solution mapped inside the area;

[0069] 2) Calculation: Calculate the approximation between the Kriging value of the current region and the value actually calculated by the optimization model. If the approximation reaches a certain level, it means that the solution of the model can reflect the changes of the independent variable and the dependent variable within the region, thereby satisfying the corresponding approximation conditions. The maximum value can be selected as the optimal solution for the region.

[0070] 3) Traversal: Traverse the optimal solutions of all regions and select the largest optimal solution as the global optimal solution.

[0071] The device for calculating the consumption responsibility weight of the power market subject in the city area includes:

[0072] The model building module is used to build a power market subject consumption responsibility weight limit model. The model is a two-layer master-slave game model. The upper layer of the game model is the provincial power company and the provincial power market trading center including green certificate trading, which is responsible for the dispatch of electricity and green certificate allocation in several power system areas of the province; the optimization goal of the upper layer of the game is to ensure the maximum total social surplus of the provincial power market in the transaction process during the process of purchasing and selling electricity; the optimization goal of the lower layer of the game model is to reduce the power generation cost of each area and minimize the power generation cost;

[0073] The model solving module is used to accelerate the solution of the model through the Kriging model algorithm.

[0074] A non-transitory computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the method for calculating the consumption responsibility weight of power market entities within a municipal area as described above.

[0075] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for calculating the consumption responsibility weight of the electricity market entities in the municipal area as described above is implemented.

[0076] The beneficial technical effects of the present invention are:

[0077] (1) The present invention combines the existing renewable energy consumption responsibility weight system and the power market clearing system, and proposes a two-level programming mathematical model that can calculate when market entities and power market operators can simultaneously achieve optimal benefits and meet the consumption responsibility weight system. It can basically reflect the actual operation of the power market when clearing, and make corresponding calculations based on the actual situation, which can guide the implementation of government policies;

[0078] (2) The present invention realizes accelerated solution of the model by means of the proxy model Kriging model, so that it meets the need of quickly calculating the optimal consumption responsibility weight in the scheduling process as the scene changes in the future; through this model, a preliminary simulation of the renewable energy consumption responsibility weight system is carried out through double-layer master-slave game optimization, and it can meet the needs of accelerated calculation, so that it is suitable for corresponding application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 A diagram of the electricity market architecture provided by an embodiment of the present invention;

[0080] Figure 2 A model framework diagram of the construction provided by an embodiment of the present invention;

[0081] Figure 3An algorithm model diagram of the Kriging proxy optimization model provided in an embodiment of the present invention for solving the master-slave game;

[0082] Figure 4 A block diagram of solving the optimal electricity price and renewable energy consumption responsibility weight using a particle swarm algorithm provided in an embodiment of the present invention;

[0083] Figure 5 The difference in electricity prices before and after optimization provided by the embodiment of the present invention;

[0084] FIG6( a ) shows the power output of region 1 according to an embodiment of the present invention;

[0085] FIG6( b ) shows the load and response of region 1 according to an embodiment of the present invention;

[0086] FIG6( c ) is a diagram showing the transaction status of green certificates in area 1 according to an embodiment of the present invention;

[0087] FIG. 7( a ) shows the power output of region 2 provided in an embodiment of the present invention;

[0088] FIG7( b ) shows the load and response of region 2 provided by an embodiment of the present invention;

[0089] FIG. 7( c ) shows the transaction status of green certificates in area 2 provided by an embodiment of the present invention;

[0090] FIG8( a ) shows the power output of region 3 provided by an embodiment of the present invention;

[0091] FIG8( b ) shows the load and response of region 3 provided by an embodiment of the present invention;

[0092] FIG8( c ) shows the transaction status of green certificates in area 3 provided by an embodiment of the present invention;

[0093] Fig. 9 The time consumption before and after using the Kriging proxy model algorithm provided in the embodiment of the present invention;

[0094] Fig.10 The accuracy of the calculation results varies with the number of sample points during the calculation process provided by the embodiment of the present invention;

[0095] Fig.11 The impact of different consumption responsibility weights on total revenue in the market provided by the embodiment of the present invention;

[0096] Fig.12 The relationship between the consumption responsibility weight and the minimum cost of each area obtained by calculation provided in the embodiment of the present invention;

[0097] Fig.13The relationship between the consumption responsibility weight ratio and the quantity of green certificates purchased by each region provided in the embodiment of the present invention;

[0098] Fig.14 The embodiment of the present invention provides a relationship between the consumption responsibility weight ratio and the proportion of green certificates purchased in each region. DETAILED DESCRIPTION

[0099] The following is a further clear and complete description of a method and device for clearing a power capacity market taking into account flexibility provided by the present invention in conjunction with the accompanying drawings:

[0100] Example 1

[0101] A method for calculating the consumption responsibility weight of power market entities within the city area, the method comprising the following steps:

[0102] Construct a power market subject consumption responsibility weight limit model, which is a two-layer master-slave game model. The upper layer of the game model is the provincial power company and the provincial power market trading center including green certificate trading, which is responsible for the dispatch of electricity and green certificate allocation in several power system areas of the province; the optimization goal of the upper layer of the game is to ensure the maximum total social surplus in the provincial power market during the transaction process during the purchase and sale of electricity; the optimization goal of the lower layer of the game model is to reduce the power generation cost of each area and minimize the power generation cost;

[0103] The model is solved quickly using the Kriging model algorithm.

[0104] In the present invention, the model is a solution framework with a double-layer master-slave game model, and the double-layer master-slave game model is established based on the working principle of provincial power market clearing transactions.

[0105] The model is divided into a two-layer master-slave game part and a part outside the master-slave game. Among them, the two-layer master-slave game represents the power market within the city area; the part outside the two-slave master-slave game represents the part outside the city area. The first is the two-layer master-slave game part.

[0106] The lower part of the game model represents the provincial power grid area. The area is not large and the number of market players is limited. Therefore, it includes several power system areas, including traditional energy power generation companies, green power generation companies, and loads with demand response. Including several market players, in order to facilitate the analysis of market players, take into account the characteristics of different power market players, based on the regional characteristics of smaller regions than other large provinces, and at the same time help to ensure the convergence of calculations, the lower layer of the game model puts the power market players in the region for unified management and calculation. The optimization goal of the lower layer is to minimize the power generation cost of each region and minimize the power generation cost.

[0107] The upper layer of the game model is the provincial power companies and the provincial power market trading center, including green certificate trading, which is responsible for the dispatch of electricity and green certificate allocation in several power system areas across the province. The optimization goal of the upper layer of the game is to ensure the maximum total social surplus in the provincial power market during the transaction process during the purchase and sale of electricity.

[0108] In addition to the hierarchical calculation and the lower-level model being constructed in a unified form, other assumptions are also applied in the master-slave game part of the model for solving the process of absorbing responsibility weights:

[0109] (1) The lower part of the model represents the provincial power market players, which are cleared through bidding with the upper model by solving the optimization model;

[0110] (2) In the region, one unit of green electricity certificate can be exchanged for one unit of renewable energy electricity (hereinafter referred to as green electricity);

[0111] While paying attention to the clearing effect of the power market within the city, we also pay attention to the transmission of power and green certificates across provinces, which is manifested as the external purchase and sale of power at the upper level of the two-layer game model. Therefore, several buyers and sellers outside the province are regarded as a whole and placed above the master-slave game model to measure the need for out-of-province transactions when the provincial power market cannot meet the transaction balance.

[0112] Specific:

[0113] The upper objective function of the game model is:

[0114]

[0115] in, It is the power that the city's power supply companies and power trading institutions purchase from outside the city. It is the power of electricity sold outside the city. is the power purchased by the region from the power supply company. is the power sold by the region to the power supply company, are their corresponding prices respectively; in addition, regarding green certificates, is the price of green certificates purchased from outside the city. is the price of green certificates sold within the city;

[0116] The corresponding constraints are:

[0117] (1) Meeting power balance conditions:

[0118]

[0119] (2) Restrictions on purchasing or selling electricity from within the city:

[0120] z 01,t +z 02,t ≤1 (49)

[0121] Among them, z 01,t 、z 02,t A 0-1 variable representing the amount of power a region buys and sells to a power company, satisfying the corresponding upper and lower bound constraints:

[0122]

[0123] In order to ensure that the electricity market players can trade with the power company instead of bypassing the power company to buy electricity from the power market outside the province, and to ensure the continuous profitability of the power supply bureau, the electricity prices set by the power market area are stipulated to meet the following requirements:

[0124]

[0125] Through such operations, market players within the city area, through the city area power company or agent, conduct corresponding transactions with the outside world when the power supply in the area is insufficient or excessive, and include the corresponding costs in the corresponding total social surplus;

[0126] In addition, in order to meet the city's power consumption responsibility constraints, green certificate requirements are stipulated for power consumption within the city. At this time, the city power company or trading center still needs to dispatch electricity, and the corresponding renewable energy consumption responsibility weight is selected to meet the corresponding regional consumption responsibility weight limit:

[0127]

[0128] Based on the above description, the model of the lower layer of the urban electricity market is constructed:

[0129] For the power system operating areas in the city, each power system operating area includes traditional energy generating units, green power generating units, and corresponding demand-side response loads. The load response amount of the demand-side response load is measured by the reducible load.

[0130] The optimal dispatching scheme within the region is similar to the safety-constrained unit commitment, and its purpose is to minimize the operating cost of the entire model.

[0131] The objective function of the lower layer of the game model is:

[0132] minC total =C bs +C tr +C comp +C cert (54)

[0133] The internal costs of the region include the cost of purchasing and selling electricity Cbp , the cost of traditional energy generation C tr , and the compensation cost of demand-side response C comp 、Green Certificate Purchase Cost C cert The sum of four items;

[0134] Among them, the cost of purchasing and selling electricity C bp :

[0135]

[0136] Traditional energy generation cost C tr :

[0137]

[0138] Compensation costs for demand-side response:

[0139]

[0140] Green certificate purchase cost:

[0141]

[0142] The following constraints are met:

[0143] Constraints for purchasing and selling electricity at different times:

[0144] z 1,t +z 2,t ≤1 (59)

[0145] Upper limit of electricity purchase and sale:

[0146]

[0147] Power balance constraints:

[0148]

[0149] Among them, λ bp,t The electricity purchase and sale price at each moment, The amount of electricity purchased from the power company. P is the amount of electricity sold to the power company. wd,t is the output of wind turbine, P pv,t Output power of photovoltaic unit, P l,t is the load power, z 1,t is the electricity purchase variable for each period, z 2,t is the electricity sales variable for each period, is the upper limit of transactions with the power grid company per unit time; w t is the unit working status, u t is the unit startup action variable, v tis the unit shutdown action variable, P tr is the output of the traditional energy generator set, and R is the ramp rate;

[0150] At time t = 1, it satisfies:

[0151] u1=w1-w T (63)

[0152] v1=w1-w T (64)

[0153] When t≥2,

[0154] u t =w t -w t-1 (t=2,…,T) (65)

[0155] Define load constraints:

[0156] 0≤P cut,i,t ≤r cut,l,max P l,t (66)

[0157] P cut,i,t +P cut,i,t-1 ≤r cut,l,jointmax P l,t (67)

[0158] Among them, P l,t is the load, P cut,i,t For the load that can be reduced, r cut,max is the maximum reduction rate of a single unit, r cut,jointmax is the combined maximum reduction rate of the units;

[0159] At the same time, the renewable energy consumption responsibility weight constraints are met:

[0160]

[0161] q cert,t +P wd,t +P pv,t ≥γL total,t (69)

[0162] Among them, γ is the quota ratio, L total,t is the total load taking into account the demand side response, q cert,t The purchase amount of green certificates.

[0163] When solving the above model, the agent model is an effective method to accelerate the solution of optimization problems. Through the Kriging metamodel, the problem of solving the power market entity consumption responsibility weight model in the first part can be solved and the process can be accelerated.

[0164] The Kriging metamodel is an interpolation model whose interpolation result is defined as the linear weighting of the known sample function response values, as shown in the following formula.

[0165]

[0166] in, is the estimated value at point x, w (i) is the weighting coefficient, and the formula aims to use the weighted sum of all known points in space to estimate the value of the unknown point. This estimate can satisfy the estimated value at the point (x, y) The optimal set of coefficients that minimizes the difference from the true value y, where the estimate satisfies the unbiased estimation condition. Given the weighted coefficient ω = [w (1) w (2) …w (4) ] T , you can get the function value of a known complex function under any complex input. The learning process of Kriging function is to learn through a large amount of data and train the weighted coefficients so that such a fitting model can reflect the profile of the function.

[0167] At present, the main problem facing the power market is to complete the solution of the renewable energy consumption responsibility weight, and there is no high requirement for the corresponding solution speed. However, in order to facilitate the application of the corresponding solution measures in the day-ahead market or even the real-time market where timeliness needs to be considered, the corresponding solution speed requirement is very important. Clarifying the consumption responsibility weight of the power market players at each moment plays an important role in providing enterprises with clear adaptive signals, reducing related uncertainties, and optimizing power market transactions and management methods.

[0168] Therefore, in addition to solving the electricity market clearing model and obtaining the appropriate consumption responsibility weight, the corresponding solution to accelerate the solution of the model is also studied. To this end, this application proposes an accelerated solution method for the consumption responsibility weight model based on the above market transaction model.

[0169] An overview of the method is:

[0170] First, in the master-slave two-level programming problem, the problem can be divided into two independent parts, the upper layer and the lower layer. When solving the problem, the master problem and the slave problem can be separated, and the optimal solution of the slave problem is solved first, that is, the Latin hypercube sampling method is first used to equally extract the variables of the slave problem in the feasible domain of the slave problem, and then the variables of the slave problem are brought into the slave problem for calculation to obtain the values ​​of the slave problem, and then brought into the master problem for calculation optimization, and the independent variable-dependent variable mapping of the entire master-slave problem is formed; next, in order to improve the efficiency of the solution, these mappings are divided into several small areas according to the values ​​of the independent variables. In each small area, for the mapping set of each area, the Kriging proxy function is used to approximate such corresponding mappings, and the particle swarm optimization algorithm is used to find the maximum value; finally, the optimal values ​​obtained by optimizing all local areas are summarized together, and the optimal value is obtained as the optimal solution of the master-slave game method.

[0171] Specifically, for the two-layer master-slave game model, the steps of solving the algorithm are:

[0172] (1) Initialize the lower-level model variable parameters, including the on-grid electricity price in the power market and the renewable energy consumption responsibility weight; these parameters can be obtained through Latin hypercube sampling, so that the sampling points are distributed as evenly as possible in the region of the independent variable, thereby accurately reflecting the overall situation of the whole system;

[0173] (2) Substitute the unknown parameters of the power market grid price and renewable energy consumption responsibility weight in step (1) into the lower model to solve the corresponding value of the problem, such as the corresponding transaction volume; then, substitute the corresponding value of the problem with the corresponding constraints into the upper model to directly obtain the feasible value of the corresponding upper model;

[0174] (3) Establish a mapping relationship between the power market operation conditions obtained in the main problem in step (2) and the independent variable parameters in step (1), such as the electricity price of each market entity;

[0175] (4) According to the feasible domain of the variables of the lower model, the dependent variables of the sampling points are divided into several parts according to the density clustering algorithm, and a Kriging model is generated for the samples corresponding to each part of the parameter lower model to represent the level of the function value of the part;

[0176] (5) Perform particle swarm optimization on each part generated in (4) to find the maximum value of each area as the maximum value of the overall model;

[0177] (6) Rank the maximum values ​​obtained in each region in (5) to obtain the overall optimal value, which is the optimal solution to the master-slave game model.

[0178] The corresponding flow chart is as attached. Figure 3shown.

[0179] This algorithm uses the Kriging model. In order to ensure the availability of the model, the training and optimization algorithm of the Kriging model are carried out simultaneously with the particle swarm algorithm, so that better training results can be achieved than arranging the model first and then optimizing it. In this process, the known parameters are each group of electricity prices and the output power of each unit within the region obtained by completing the regional classification, and it is assumed that there are multiple points within a region. In it, the operations of adding points, calculating, and traversing are completed to obtain the location of the optimal solution in the region. The details are as follows:

[0180] 1) Adding points: When the electricity price of each group and the output power of each unit in the known area are known, select the known points in the area and let these points move according to the rules of the particle swarm algorithm to obtain new points. The new points are brought into the model to obtain approximate values, and the fitting function of the Kriging agent model is trained and updated accordingly to obtain the local optimal solution mapped inside the area;

[0181] 2) Calculation: Calculate the approximation between the Kriging value of the current region and the value actually calculated by the optimization model. If the approximation reaches a certain level, it means that the solution of the model can reflect the changes of the independent variable and the dependent variable within the region, thereby satisfying the corresponding approximation conditions, and the maximum value can be selected as the optimal solution for the region.

[0182] 3) Traversal: Traverse the optimal solutions of all regions and select the largest optimal solution as the global optimal solution.

[0183] Such an algorithm can approximate the nonlinear optimization model solution process by training a model in each area, so that the corresponding independent variable-dependent variable mapping relationship can be approximated in a linear way in each small part, thereby avoiding a long calculation process. For example, in MATLAB, if Gurobi+Yalmip is used to calculate a set of mixed integer linear programming, it takes much longer than calculating a set of polynomial formulas. Therefore, using the metamodel can help improve computational efficiency.

[0184] The corresponding flow chart is as attached. Figure 4 shown.

[0185] As an example, in this embodiment, the following calculation is performed using the method of the present application:

[0186] Get or determine the values ​​and parameters of relevant parameters: there are 24 time periods, simulating 1 day and 1 hour as a time period; the upper limits of electricity sold to and purchased from outside the region are respectively determined by Figure 1The curve specified in the specification; there is a gas turbine in each part, and the power upper limit of this part is set to 1.3MWh and the lower limit is 3.3MWh; it is assumed that the corresponding ramp rate limit is 2MWh / h; the wind power output is determined and generated by normal distribution random numbers; the output of solar photovoltaic also adopts the property of random generation in the form of random numbers within a certain time range. In addition, the load response compensation amount of the three areas is determined to be 600 yuan / MWh, 700 yuan / MWh, and 800 yuan / MWh, the fixed cost of gas turbine operation is 800 yuan / MWh, the marginal cost is 400 yuan / MWh, and the start-stop cost is 600 yuan / MWh;

[0187] Generating sample points means determining the renewable energy consumption responsibility weights at equal intervals within γ = 0.222 to 0.8 and observing the distribution of response results; the number of generated sample points is determined within the range of N = 10 to 80, and the distribution of solution time, accuracy and stability is observed.

[0188] The result is as follows:

[0189] (1) Model solution:

[0190] According to the above content, from the perspective of model solution, the managed power market is divided into one master and multiple slaves and multiple regions, with the upper layer being the power company (power market manager) and the lower layer being the power market participants in each region. Given the upper and lower limits of the output of traditional energy gas boiler units, the upper and lower limits of electricity prices, and the actual output of renewable energy, the optimization of the renewable energy consumption responsibility weight, the actual output value of the gas boiler unit, etc. is actually modeled as an economic dispatch problem with the responsibility weight of renewable energy consumption.

[0191] (2) Improved content

[0192] 1) The solution speed and accuracy have been improved

[0193] See specifically Fig. 9 , 10.

[0194] Fig. 9 It shows the effect that the algorithm can achieve over time. The N on the horizontal axis represents the number of sample points initially used. It can be seen that the algorithm that uses Gurobi-Yalmip alone to solve the optimization model requires more time, and it increases exponentially. Under the premise of a fixed market participant composition, the solution time varies with N. However, as the number of sample points increases, the calculation of the model becomes more and more complicated. The increase in the number of categories and data points that need to be processed during the classification process leads to an increase in the time of model construction and training as the number of sample points increases.

[0195] Fig.10It shows that when the number of sample points N exceeds a certain value, the solution obtained remains basically stable, indicating that the solution has reached stability.

[0196] 2) Intuitively reflect the relationship between consumption responsibility weight and cost, green certificate, etc.

[0197] See specifically Fig.11 , 12 , 13, 14.

[0198] Fig.11 It shows that K is calculated from 0.222 to 0.922, with a value of 0.05 every time, to obtain the relationship between the market's operating income and the quota ratio. This figure can be used to analyze the situation when different market systems change. Under the fixed electricity price system, the government purchases and sells green electricity in proportion. In other words, the government will purchase and distribute green electricity in a unified manner according to a certain proportion, and sell it to market participants at a fixed price. In this case, in order to achieve the responsibility weight of renewable energy consumption, each market entity mainly consumes renewable energy according to the lowest requirements in the document. The specific situation of its consumption is shown at the left end point of the spit line.

[0199] When the consumption responsibility weight is unified, in order to strive for more profits, the operator will set the consumption responsibility weight slightly higher. At this time, the unified quota ratio will lead to increased costs in some areas, especially in the process of increasing the proportion of green certificates, and the benefits obtained will decrease at the same electricity price.

[0200] However, the growth in regional green certificate trading volume also has its advantages. For example, with the increase in consumption responsibility weights, the reduction in electricity purchases and sales will lead to a decline in the profits of provincial market operators. However, thanks to the growth in green certificate trading volume, provincial power market operators need to act as agents for green certificates purchased from outside the region, which results in a small peak when the consumption responsibility weight ratio is 0.7-0.8, which is also beneficial for the continued stable operation of provincial power market operators.

[0201] Fig.12 It shows that the emergence of the consumption responsibility weight ratio is a burden for every market player, which is reflected in the fact that in addition to buying and selling electricity and achieving power balance, an irrelevant factor must be added: completing the consumption responsibility weight of renewable energy. Completing this task will obviously increase the expenditure of each market participant to purchase the cost of green certificates, thereby pushing up the costs of various regions. Although the consumption of renewable energy is indeed increasing, the issuance of green certificates has led to an increase in the amount of green certificates purchased by each region.

[0202] Fig.13 and Fig.14It shows that as the proportion of green certificates purchased by each region increases, the corresponding consumption responsibility weight ratio also tends to rise. In order to balance the lowest cost and consumption responsibility weight for each region, the corresponding optimal consumption responsibility weight ratio can be further found at this time. The optimal quota corresponding to the experimental results under different quota ratios is calculated:

[0203] Table 1 Optimal quotas under different quota ratios

[0204]

[0205] It can be obtained that the corresponding maximum income is 8376.6224 yuan, and the average weight ratio is 22.712%, which meets the constraints higher than the relevant national standards.

[0206] According to the above experimental results, it can be judged that the model can not only output appropriate results according to the parameters, but also improve the solution speed while intuitively reflecting the advantages and disadvantages of the corresponding multiple solutions.

[0207] Example 2

[0208] This embodiment provides a device for calculating the consumption responsibility weight of a power market subject in a city area, including:

[0209] The model building module is used to build a power market subject consumption responsibility weight limit model. The model is a two-layer master-slave game model. The upper layer of the game model is the provincial power company and the provincial power market trading center including green certificate trading, which is responsible for the dispatch of electricity and green certificate allocation in several power system areas of the province; the optimization goal of the upper layer of the game is to ensure the maximum total social surplus of the provincial power market in the transaction process during the process of purchasing and selling electricity; the optimization goal of the lower layer of the game model is to reduce the power generation cost of each area and minimize the power generation cost;

[0210] The model solving module is used to accelerate the solution of the model through the Kriging model algorithm.

[0211] Furthermore, the present invention adopts the following technical solutions:

[0212] A non-transitory computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the method for calculating the consumption responsibility weight of power market entities within a municipal area as described above.

[0213] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for calculating the consumption responsibility weight of the electricity market entities in the municipal area as described above is implemented.

[0214] Through the description of the above embodiments, it can be clearly understood by those skilled in the art that the facilities of the present invention can be implemented by means of software plus the necessary general hardware platform. The embodiments of the present invention can be implemented using an existing processor, or by a dedicated processor used for this purpose or other purposes for an appropriate system, or by a hard-wired system. The embodiments of the present invention also include a non-transitory computer-readable storage medium, which includes a machine-readable medium for carrying or having a machine-executable instruction or data structure stored thereon; such a machine-readable medium can be any available medium that can be accessed by a general or special-purpose computer or other machine with a processor. For example, such a machine-readable medium can include RAM, ROM, EPROM, EEPROM, CD-ROM or other optical disk storage, disk storage or other magnetic storage device, or any other medium that can be used to carry or store the required program code in the form of machine-executable instructions or data structures, and can be accessed by a general or special-purpose computer or other machine with a processor. When information is transmitted or provided to a machine via a network or other communication connection (hard wiring, or wireless, or a combination of hard wiring and wireless), the connection is also considered a machine-readable medium.

[0215] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. The calculation method of the consumption responsibility weight of the power market players in the city area is characterized by: The method comprises the following steps: Construct a power market subject consumption responsibility weight limit model, which is a two-layer master-slave game model. The upper layer of the game model is the provincial power company and the provincial power market trading center including green certificate trading, which is responsible for the dispatch of electricity and green certificate allocation in several power system areas of the province; the optimization goal of the upper layer of the game is to ensure the maximum total social surplus in the provincial power market during the transaction process during the purchase and sale of electricity; the optimization goal of the lower layer of the game model is to reduce the power generation cost of each area and minimize the power generation cost; The model is solved quickly using the Kriging model algorithm.

2. The method for calculating the consumption responsibility weight of the power market subject in the city area according to claim 1 is characterized in that: The upper objective function of the game model is: in, It is the power that the city's power supply companies and power trading institutions purchase from outside the city. It is the power of electricity sold outside the city. is the power purchased by the region from the power supply company. is the power sold by the region to the power supply company, are their corresponding prices respectively; in addition, regarding green certificates, is the price of green certificates purchased from outside the city. is the price of green certificates sold within the city; The corresponding constraints are: (1) Meeting power balance conditions: (2) Restrictions on purchasing or selling electricity from within the city: With 01,t +with 02,t ≤1 (3) Among them, z 01,t 、z 02,t A 0-1 variable representing the amount of power a region buys and sells to a power company, satisfying the corresponding upper and lower bound constraints: In order to ensure that the electricity market players can trade with the power company instead of bypassing the power company to buy electricity from the power market outside the province, and to ensure the continuous profitability of the power supply bureau, the electricity prices set by the power market area are stipulated to meet the following requirements: Through such operations, market players within the city area, through the city area power company or agent, conduct corresponding transactions with the outside world when the power supply in the area is insufficient or excessive, and include the corresponding costs in the corresponding total social surplus; In addition, in order to meet the city's power consumption responsibility constraints, green certificate requirements are stipulated for power consumption within the city. At this time, the city power company or trading center still needs to dispatch electricity, and the corresponding renewable energy consumption responsibility weight is selected to meet the corresponding regional consumption responsibility weight limit:

3. The method for calculating the consumption responsibility weight of the power market subject in the city area according to claim 1 is characterized in that: The objective function of the lower layer of the game model is: my C total =C bs +C tr +C comp +C cert (8) The internal costs of the region include the cost of purchasing and selling electricity C bp , the cost of traditional energy generation C tr , and the compensation cost of demand-side response C comp 、Green Certificate Purchase Cost C cert The sum of four items; Among them, the cost of purchasing and selling electricity C bp : Traditional energy generation cost C tr : Compensation costs for demand-side response: Green certificate purchase cost: The following constraints are met: Constraints for purchasing and selling electricity at different times: With 1,t +with 2,t ≤1 (13) Upper limit of electricity purchase and sale: Power balance constraints: Among them, λ bp,t The electricity purchase and sale price at each moment, The amount of electricity purchased from the power company. P is the amount of electricity sold to the power company. wd,t is the output of wind turbine, P pv,t Output power of photovoltaic unit, P l,t is the load power, z 1,t is the electricity purchase variable for each period, z 2,t is the electricity sales variable for each period, is the upper limit of transactions with the power grid company per unit time; w t is the unit working status, u t is the unit startup action variable, v t is the unit shutdown action variable, P tr is the output of the traditional energy generator set, and R is the ramp rate; At time t = 1, it satisfies: u1=w1-w T (17) v1=w1-w T (18) When t≥2, u t =w t -w t-1 (t=2,…,T) (19) Define load constraints: 0≤P cut,i,t ≤r cut,l,max P l,t (20) P cut,i,t +P cut,i,t-1 ≤r cut,l,jointmax P l,t (21) Among them, P l,t is the load, P cut,i,t For the load that can be reduced, r cut,max is the maximum reduction rate of a single unit, r cut,jointmax is the combined maximum reduction rate of the units; At the same time, the renewable energy consumption responsibility weight constraints are met: q cert,t +P wd,t +P pv,t ≥γL total,t (23) Among them, γ is the quota ratio, L total,t is the total load taking into account the demand side response, q cert,t The purchase amount of green certificates.

4. The method for calculating the consumption responsibility weight of the power market subject in the city area according to claim 1 is characterized in that: The method for accelerating the solution of the model by the Kriging model algorithm is: First, in the master-slave two-level programming problem, the problem can be divided into two independent parts, the upper layer and the lower layer. When solving the problem, the master problem and the slave problem can be separated, and the optimal solution of the slave problem is solved first, that is, the Latin hypercube sampling method is first used to equally extract the variables of the slave problem in the feasible domain of the slave problem, and then the variables of the slave problem are brought into the slave problem for calculation to obtain the values ​​of the slave problem, and then brought into the master problem for calculation optimization, and the independent variable-dependent variable mapping of the entire master-slave problem is formed; next, in order to improve the efficiency of the solution, these mappings are divided into several small areas according to the values ​​of the independent variables. In each small area, for the mapping set of each area, the Kriging proxy function is used to approximate such corresponding mappings, and the particle swarm optimization algorithm is used to find the maximum value; finally, the optimal values ​​obtained by optimizing all local areas are summarized together, and the optimal value is obtained as the optimal solution of the master-slave game method.

5. The method for calculating the consumption responsibility weight of the power market subject in the city area according to claim 1 is characterized in that: For the two-layer master-slave game model, the steps of solving the algorithm are: (1) Initialize the lower-level model variable parameters, including the on-grid electricity price in the power market and the renewable energy consumption responsibility weight; these parameters can be obtained through Latin hypercube sampling, so that the sampling points are distributed as evenly as possible in the region of the independent variable, thereby accurately reflecting the overall situation of the whole system; (2) Substitute the unknown parameters of the power market grid price and renewable energy consumption responsibility weight in step (1) into the lower model to solve the corresponding value of the problem, such as the corresponding transaction volume; then, substitute the corresponding value of the problem with the corresponding constraints into the upper model to directly obtain the feasible value of the corresponding upper model; (3) Establish a mapping relationship between the power market operation conditions obtained in the main problem in step (2) and the independent variable parameters in step (1), such as the electricity price of each market entity; (4) According to the feasible domain of the variables of the lower model, the dependent variables of the sampling points are divided into several parts according to the density clustering algorithm, and a Kriging model is generated for the samples corresponding to each part of the parameter lower model to represent the level of the function value of the part; (5) Perform particle swarm optimization on each part generated in (4) to find the maximum value of each area as the maximum value of the overall model; (6) Rank the maximum values ​​obtained in each region in (5) to obtain the overall optimal value, which is the optimal solution to the master-slave game model.

6. The method for calculating the consumption responsibility weight of the power market subject in the city area according to claim 1 is characterized in that: When using the Kriging model to solve the two-layer master-slave game model, it specifically includes: 1) Adding points: When the electricity price of each group and the output power of each unit in the known area are known, select the known points in the area and let these points move according to the rules of the particle swarm algorithm to obtain new points. The new points are brought into the model to obtain approximate values, and the fitting function of the Kriging agent model is trained and updated accordingly to obtain the local optimal solution mapped inside the area; 2) Calculation: Calculate the approximation between the Kriging value of the current region and the value actually calculated by the optimization model. If the approximation reaches a certain level, it means that the solution of the model can reflect the changes of the independent variable and the dependent variable within the region, thereby satisfying the corresponding approximation conditions. The maximum value can be selected as the optimal solution for the region. 3) Traversal: Traverse the optimal solutions of all regions and select the largest optimal solution as the global optimal solution.

7. A device for calculating the consumption responsibility weight of power market entities within the city area, characterized in that: include: The model building module is used to build a power market subject consumption responsibility weight limit model. The model is a two-layer master-slave game model. The upper layer of the game model is the provincial power company and the provincial power market trading center including green certificate trading, which is responsible for the dispatch of electricity and green certificate allocation in several power system areas of the province; the optimization goal of the upper layer of the game is to ensure the maximum total social surplus of the provincial power market in the transaction process during the process of purchasing and selling electricity; the optimization goal of the lower layer of the game model is to reduce the power generation cost of each area and minimize the power generation cost; The model solving module is used to accelerate the solution of the model through the Kriging model algorithm.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the method for calculating the consumption responsibility weight of the power market entities in the municipal area as described in any one of claims 1 to 6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, it implements the method for calculating the consumption responsibility weight of the power market entity in the municipal area as described in any one of claims 1 to 6.

Images