Comprehensive energy multi-microgrid transaction scheduling method based on pumped storage power station as public facility

Abstract

The invention discloses an integrated energy multi-microgrid transaction scheduling method based on a pumped storage power station as a public facility, and the method comprises the steps: building a model for an integrated energy part of the pumped storage power station and a microgrid, and constructing a corresponding cost function according to the cost elements of the pumped storage power station and the microgrid; taking the pumped storage power station as a public facility, and creating an electric energy interaction model between the pumped storage power station and the microgrid; establishing a Nash game model of the pumped storage power station and the micro-grid, optimizing the cost and income distribution of each main body by using the Nash game model, and solving the inherent operation cost and the cooperative transaction cost in the Nash game model through a hybrid RM-MEDA algorithm and a distributed alternating sub-method; according to the method, green energy storage is fully utilized, carbon emission and resource loss are reduced, the utilization rate of renewable energy sources is increased, the cost of each micro-grid is reduced, and the stability and economical efficiency of regional micro-grid interconnection are improved.

Description

Technical Field

[0001] The present invention belongs to the field of microgrid and pumped storage power station trading and dispatching, and particularly relates to a comprehensive energy multi-microgrid trading and dispatching method based on a pumped storage power station as a public facility. Background Technique

[0002] With the proposal and implementation of new energy and the dual-carbon goal, renewable energy such as solar energy and wind energy has developed rapidly. However, these new energies have intermittency and instability, posing challenges to the safe and stable operation of the power grid. As a large-scale energy storage method, the pumped storage power station has technical and economic advantages such as large capacity, multiple operating conditions, high speed, high reliability, and good economy compared with other energy storage methods. Therefore, in the context of energy transformation and the development of new energy, the pumped storage power station has become an important part of the development of microgrids. The prosperity of renewable energy has further promoted the mutual integration of various types of energy, gradually forming a comprehensive energy system. The comprehensive energy system can draw on the strengths of all parties, complement and assist each other to make the system more stable and efficient; the diversity of energy gives it better flexibility. In order to give full play to the various functions of the pumped storage power station such as peak shaving and valley filling, frequency modulation and phase modulation, and accident standby, solve the intermittency problem of new energy power generation, promote trading cooperation among microgrids, optimize the output of each entity in the comprehensive energy, and improve the safety and stability of the power grid.

[0003] Therefore, proposing a comprehensive energy multi-microgrid trading and dispatching method based on a pumped storage power station as a public facility has important supporting and guiding significance for the research on the simulation of the pumped storage power station comprehensive energy microgrid system, global dispatching optimization design, etc.

[0004] Classical renewable energy microgrid optimization methods have been widely studied, but they do not construct a comprehensive energy utilization system and a pumped storage power station model. For example: (1) Document 1: Wang Shouxiang, Zhang Qi, Wang Han, etc. Optimization planning method for regional multi-microgrid system under high renewable energy penetration [J]. Electric Power Automation Equipment, 2018, 38(12): 3338, 52. Proposed an optimization planning method for regions with a large amount of renewable energy, but it did not involve cooperation among microgrids and did not use the pumped storage power station as an independent entity for multi-microgrid sharing.

[0005] (2) Document 2: Zhang Q, Zhou A, Jin Y. A regularity model-based multiobjective estimation of distribution algorithm: RM-MEDA. IEEE Trans. on Evolutionary Computation, 2008, 12(1): 41 - 63. [doi: 10.1109 / TEVC.2007.894202]. It proposed the model of the RM-MEDA algorithm. However, the regeneration of offspring in the algorithm is single, and the mutation of offspring is weak, making the solution prone to falling into local optimum and unable to fully search for the optimal values in the entire space.

[0006] (3) Document 3:

[23] Cui Mingyong, Xuan Mingyang, Lu Zhigang, et al. Optimal operation strategy of multi-integrated energy service providers based on cooperative game [J]. Proceedings of the CSEE, 2022, 42(10): 3548 - 3564. It proposed an optimal operation strategy for multi-microgrids considering the interests of service providers. However, it did not consider the application of pumped storage power stations in integrated energy, and the solution method for cooperative games is also relatively traditional.

[0007] In summary, there is currently no integrated energy multi-microgrid model with a pumped storage power station as a public facility. Therefore, it is necessary to propose an integrated energy multi-microgrid trading and scheduling method based on a pumped storage power station as a public facility to solve the above problems. Summary of the Invention

[0008] The technical problem to be solved by the present invention is to provide an integrated energy multi-microgrid trading and scheduling method based on a pumped storage power station as a public facility, aiming to solve the problem that the existing technology lacks an integrated energy multi-microgrid model with a pumped storage power station as a public facility, and having the characteristics of reducing the costs of each microgrid and improving the stability and economy of regional microgrid interconnection.

[0009] To achieve the above technical effects, the technical solution adopted by the present invention is as follows: An integrated energy multi-microgrid trading and scheduling method based on a pumped storage power station as a public facility, comprising the following steps: S1, construct the operation models of the pumped storage power station and each entity in the microgrid; the microgrid includes gas turbines, waste heat boilers, wind power generation, and photovoltaic panels as energy devices; S2, construct an integrated energy trading and scheduling model of the pumped storage power station and the microgrid; S3, optimize the multi-agent allocation of benefits through the Nash game model, and use the hybrid RM-MEDA algorithm and the distributed alternating direction method to solve the model.

[0010] Preferably, in step S1, constructing the operation models of each entity in the pumped-storage power station and the microgrid includes: During the operation of the pumped-storage power station, it is restricted by the change in reservoir capacity and its operating capacity. The capacity change model of the reservoir is as follows: ; (1) In the formula, Vr(t) and Vr(t - 1) are the reservoir capacities after and before energy storage and release respectively, in m3; Pbeng(t) and Pfa(t) are the energy storage power and power generation power of the pumped-storage power station respectively, in kW; ηbeng and ηfa are the pump efficiency and turbine efficiency respectively; ρ is the water density, taken as 1000 kg / m3; g is the acceleration due to gravity, taken as 9.81 m / s2; h is the rated head, taken as 40 m, is the continuous operation time; the constraints are as follows: ; (2) ; (3) In the formula, is the water release and discharge power of the pumped-storage power station at time t; is the pumping absorption power of the pumped-storage power station at time t; and are the 0 - 1 water release state variable and pumping state variable of the pumped-storage power station respectively, and the two do not take 1 simultaneously; and are the minimum reservoir capacity and maximum reservoir capacity respectively.

[0011] Preferably, in step S2, constructing the integrated energy trading and dispatching model of the pumped-storage power station and the microgrid includes: S201, constructing the cost model of the pumped-storage power station, the cost model of the microgrid and the carbon revenue model: Evaluating the energy storage cost is an inevitable key issue when comprehensively evaluating the operation value of energy storage; the operation cost of pumped storage includes the initial investment and construction cost and the operation and maintenance cost. The mathematical model for converting the operation cost over the entire life cycle to a daily basis is: ; (4) In the formula, is the unit power cost of the pumped storage system, (yuan / kW); is the rated power of the energy storage, MW; r is the discount rate of the energy storage project; y is the life of the energy storage; is the percentage coefficient of the operation and maintenance cost of the energy storage system; Constructing the objective function of the microgrid with the minimum total cost: ; (5) In the formula, represents the total cooperation cost of the microgrid group, excluding the interaction costs between microgrids; i represents the i-th microgrid, and t represents the t-th moment; represents the total interaction cost between microgrids; the cooperation cost of microgrid i; the interaction cost between microgrid i and other microgrids; is the interaction cost between the microgrid and the service provider; is the gas cost consumed by the microgrid; is the response cost of the flexible load of the microgrid; is the energy storage cost; a is the unit cost coefficient of gas; b, c, and d are the response cost coefficients of the transferable electric load, the reducible electric load, and the reducible heat load respectively; is the energy storage cost coefficient; θ is the operation and maintenance cost of the pumped storage power station in the formula, is the proportion of microgrid i sharing the operation and maintenance cost of the pumped storage power station; S202, as a public facility, the pumped storage power station does not charge fees when trading electric energy with each microgrid, that is, the price of buying and selling electric energy between the microgrid and the pumped storage power station is 0. However, the operation and maintenance cost of the pumped storage power station is shared by each microgrid entity according to the proportion of the difference between the electric energy sent into and absorbed by the pumped storage power station within a cycle to the total consumption of the pumped storage power station. If there is a microgrid i sending electric energy and a microgrid j absorbing electric energy at the same time period, in order to improve the energy utilization efficiency, the electric energy of microgrid i is directly sent into j microgrid, and finally the remaining or insufficient electric energy is transferred from the pumped storage power station. Photovoltaic power generation can sell electricity to other microgrids and the pumped storage power station when the electric energy is in surplus, and the same is true for wind power generation. Since there is also a gas turbine in each microgrid, each microgrid can also mobilize the gas turbine to generate electricity and sell it to other entities when it is beneficial. Construct a power trading model: Power purchase and sale power constraint between the microgrid and the pumped storage power station: ; (6) In the formula, is the electric power value sent by microgrid i to the pumped storage power station during t time period, is the power absorbed by microgrid i from the pumped storage power station during t time period, taking a negative value, uniformly represents the electric power value traded by microgrid i with the pumped storage power station during t time period, positive value means sending in, negative value means absorbing; Renewable energy constraint: ; (7) In the formula, , are the actual power generation powers of wind power and photovoltaic power, , are the predicted power generation powers of wind power and photovoltaic power; Interactive constraints between microgrids: ; (8) In the formula, is t the interactive power of microgrid i to j at time , and i are the upper and lower limits of the interactive power of microgrids j ; Power balance constraints, including thermal power balance and electric power balance: ; (9) In the formula, is the interactive power between the pumped-storage power station and the microgrid, is the predicted electric load power of the microgrid i at t time, is the actual thermal load power of the microgrid i at t time, is the actual electricity consumption load of the microgrid i at t time, is the reducible electric power of the microgrid i , is the transferable electric power of the microgrid i , is the reducible thermal load.

[0012] Preferably, in step S3, each microgrid is divided into different entities, and the pumped-storage power station is used as an independent entity. Different entities have their own interest goals respectively; the Nash game model is used to make both individual interests and social interests reach the best, and the hybrid RM-MEDA algorithm and the distributed alternating sub-method are used to solve the model, including: S301. As a kind of cooperative game, the Nash game model can take into account both individual and collective interests and has a good distribution effect on the cooperative income among multiple entities. The solution of the Nash negotiation is the Pareto optimal solution; construct the Nash game model: ; (10) In the formula, and are entities iThe benefits before and after cooperation. The constraint indicates that the benefits after cooperation are not lower than those before cooperation, ensuring that the interests of individuals are not damaged. In the cooperative game problem among the main bodies of the multi - micro - grid integrated energy system, the mathematical expression of Nash game is as follows: ; (11) In the formula, is the cost of the micro - grid i before cooperation, and divide the cost after cooperation into two parts, namely the micro - grid cooperative operation cost and the micro - grid - to - micro - grid transaction cost. The micro - grid cooperative operation cost does not include the micro - grid - to - micro - grid transaction cost. By transforming the total cooperative cost into two modules, the cooperative cost minimization problem and the negotiation payment problem, it can be solved conveniently and quickly, and the shared electricity quantity and payment electricity price of each micro - grid can be solved successively. S302. Solve the cooperative cost minimization problem, that is, the objective function C2, through the hybrid RM - MEDA algorithm, obtain the value of the minimized cooperative cost C2 of the multi - micro - grid and the optimal electric energy trading volume among the multi - micro - grids , and substitute its result value into the negotiation payment problem to further solve the payment electricity price. The algorithm model and process are as follows: a. Initialization: Set the population size N, the number of sub - populations K, the mutation probability P, and set the algorithm stop condition. Let the iteration number t = 1, randomly initialize the population Y(t) and calculate the objective function values of the individuals in the initial population. b. Judge the stop condition: If the algorithm stop condition is met, output the population Y(t); otherwise, go to the next step. c. Reproduction: Establish a probability model: Adopt the Local PCA clustering algorithm to cluster the individuals in the current population Y(t) into K classes, respectively forming K sub - populations S1, S2, …, SK, and establish a piece - wise linear probability model for each sub - population , , …, ; Specifically as follows: ; (12) ; (13) ; (14) In the formula, is the mean vector of Sj, is the i - th principal component of Sj, that is, the eigenvector corresponding to the i - th largest eigenvalue of the covariance matrix of all antibodies in Sj; project all the antibodies in Sj onto the (m - 1) - dimensional principal subspace of Sj, and denote the smallest line segment or hyperplane covering all the projection points as ; Establish a piece - wise linear probability model for the current population. Distribution of the number of reproductions: Assume that the number after the i th sub-population is generated is , where i = 1, 2, …, K , and then can be obtained from the following formula: ; (15) In the formula, represents the length of the line segment (for two-objective problems) or the hypervolume (for multi-objective problems) projected by all antibodies in population Si in the first m - 1 principal component directions, represents taking the integer part of the real number x; Generating offspring using the clonal selection operator: For each sub-population Si (i = 1, 2, …, K), generate an offspring population , denoted as and the number of antibodies in is Generating offspring using the model sampling operator: For the probability model established for each sub-population Si (i = 1, 2, …, K) , …, ; perform random sampling to generate an offspring population , and the number of sampled antibodies is ; d. Mutation: Denote , , let ; for each antibody in S, perform a mutation operation using the mutation operator based on ε interval segmentation, with a mutation probability of P, to obtain an offspring population ; e. Selection: Let , use the selection method based on the m-nearest neighbor list to select N sub-individuals from Q to form a new population Y(t + 1), let t = t + 1, and go back to the second step; S303. The costs of each microgrid before cooperation are obtained through the solution method in S302, and their values are ; Since the Nash bargaining model in Equation (11) is non-linear and not easy to solve, take the logarithm to transform Equation (11): ; (16) Considering the cost of sharing electrical energy i between microgrid and other microgrids , then there is a key factor of the trading electricity price among multi-microgrids; introduce an auxiliary variable to decouple the trading electricity price: S304. Based on the core idea of the distributed alternating sub - method algorithm, a distributed optimization model is constructed. By introducing Lagrange multipliers and penalty factors to coordinate the optimization process among multiple micro - grids; The distributed optimization model of the micro - grid i is as follows: ; (18) ; (19) ; (20) Model (18) solves the negotiation payment problem according to the distributed iteration formula (19). When the iteration process of formula (20) meets the preset convergence condition less than or equal to the error , the trading electricity price among multiple micro - grids is obtained .

[0013] The beneficial effects of the present invention are as follows: By bringing the problem of minimizing the cooperation cost that does not include the interactive electric energy cost between micro - grids into the hybrid RM - MEDA algorithm for solution, the present invention obtains the interactive electric quantity between micro - grids and the minimum cooperation cost of multiple micro - grids. The interactive cost between each micro - grid is obtained by solving the negotiation payment problem. Under the solution of the negotiation payment, the trading electricity price between micro - grids can be obtained, and then the interactive cost between micro - grids can be known. The integrated energy multi - micro - grid model with a pumped - storage power station as a public facility proposed by the present invention is a model that does not exist currently. The electric energy trading scheduling method based on this model has high practicability and development prospects. It can realize the cooperation and complementarity of multi - energy systems, enhance the adaptability of micro - grids in non - plain areas, improve the development momentum of the micro - grid industry in China, contribute energy to the transformation of the power system, and help to improve the efficiency of clean energy utilization and enhance the benefits of each subject and the whole society. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is the flowchart of the hybrid RM - MEDA algorithm of the present invention; Figure 2 is the power interaction situation between the pumped - storage power station and the micro - grid of the present invention; Figure 3 is the electric energy trading result between the micro - grids of the present invention; Figure 4 is the output diagram of the gas turbine in the micro - grid under the trading scheduling method of the present invention; Figure 5 is the output diagram of the waste heat boiler in the micro - grid under the trading scheduling method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0015] Example 1: As Figure 1As shown, a comprehensive energy multi - microgrid trading and scheduling method based on a pumped - storage power station as a public facility, including the following steps: S1, construct the operation models of each entity in the pumped - storage power station and the microgrid; the microgrid includes gas turbines, waste heat boilers, wind power generation, and photovoltaic panels as energy equipment; S2, construct a comprehensive energy trading and scheduling model for the pumped - storage power station and the microgrid; S3, optimize the multi - entity distribution of benefits through the Nash game model, and use the hybrid RM - MEDA algorithm and the distributed alternating sub - method to solve the model.

[0016] Preferably, in step S1, constructing the operation models of each entity in the pumped - storage power station and the microgrid includes: During the operation of the pumped - storage power station, it is subject to the change of reservoir capacity and its operation capacity constraints. The capacity change model of the reservoir is as follows: ; (1) In the formula, Vr(t) and Vr(t - 1) are the reservoir capacities after and before energy storage and release respectively, in m3; Pbeng(t) and Pfa(t) are the energy storage power and power generation power of the pumped - storage power station respectively, in kW; ηbeng and ηfa are the pump efficiency and turbine efficiency respectively; ρ is the water density, taken as 1000 kg / m3; g is the acceleration due to gravity, taken as 9.81 m / s2; h is the rated head, taken as 40 m, is the continuous operation time; the constraints are as follows: ; (2) ; (3) In the formula, is the water discharge power of the pumped - storage power station at time t; is the water pumping absorption power of the pumped - storage power station at time t; and are the 0 - 1 water discharge state variable and water pumping state variable of the pumped - storage power station respectively, and the two do not take 1 simultaneously; and are the minimum reservoir capacity and maximum reservoir capacity respectively.

[0017] Preferably, in step S2, constructing the comprehensive energy trading and scheduling model for the pumped - storage power station and the microgrid includes: S201, construct the cost model of the pumped - storage power station, the cost model of the microgrid, and the carbon revenue model: Energy storage cost assessment is an inevitable key issue when comprehensively evaluating the operation value of energy storage; the operation cost of pumped - storage includes the initial investment and construction cost and the operation and maintenance cost. The mathematical model of converting the operation cost over the entire life cycle to a daily basis is: ; (4) In the formula, is the unit power cost of the pumped-storage energy storage system, (yuan / kW); is the rated power of the energy storage, MW; r is the discount rate of the energy storage project; y is the life of the energy storage; is the cost percentage coefficient of the operation and maintenance of the energy storage system; Construct the objective function of the microgrid with the minimum total cost: ; (5) In the formula, represents the total cooperation cost of the microgrid group, excluding the interaction cost between microgrids; i represents the i-th microgrid, and t represents the t-th moment; represents the total interaction cost between microgrids; The cooperation cost of microgrid i; The interaction cost between microgrid i and other microgrids; is the interaction cost between the microgrid and the service provider; is the gas cost consumed by the microgrid; is the response cost of the flexible load of the microgrid; is the energy storage cost; a is the unit cost coefficient of gas; b, c, and d are the response cost coefficients of the transferable electric load, the reducible electric load, and the reducible heat load respectively; is the energy storage cost coefficient; θ is the operation and maintenance cost of the pumped-storage power station in the formula, is the proportion of microgrid i sharing the operation and maintenance cost of the pumped-storage power station; S202. As a public facility, the pumped-storage power station does not charge fees when trading electric energy with each microgrid, that is, the price of buying and selling electric energy between the microgrid and the pumped-storage power station is 0. However, the operation and maintenance cost of the pumped-storage power station is shared by each microgrid entity according to the proportion of the difference between the electric energy sent into and absorbed by the pumped-storage power station in a cycle to the total consumption of the pumped-storage power station. If there is both a microgrid i sending electric energy and a microgrid j absorbing electric energy at the same time, in order to improve the energy utilization efficiency, the electric energy of the microgrid i is directly sent into j the microgrid, and finally the remaining or insufficient electric energy is transferred from the pumped-storage power station. Photovoltaic power generation can sell electric energy to other microgrids and the pumped-storage power station when the electric energy is in surplus, and the same is true for wind power generation. Since there is also a gas turbine in each microgrid, each microgrid can also mobilize the gas turbine to generate electricity and sell it to other entities when it is beneficial. Construct the power trading model: Power purchase and sale power constraint of the microgrid to the pumped-storage power station: ; (6) In the formula, is the microgridi The electric power value delivered to the pumped-storage power station during t the time period, for the microgrid i the power absorbed from the pumped-storage power station during t the time period, taking a negative value, is uniformly used to represent the electric power value traded by the microgrid i with the pumped-storage power station during t the time period; a positive value means delivered, and a negative value means absorbed; Renewable energy constraint: ; (7) In the formula, , are the actual power generation powers of wind power and photovoltaic power, , are the predicted power generation powers of wind power and photovoltaic power; Microgrid interaction constraint: ; (8) In the formula, is the t interaction power of the microgrid i towards j at the , are the upper and lower limits of the interaction power of the microgrid i , j ; Power balance constraint, including thermal power balance and electric power balance: ; (9) In the formula, is the interaction power between the pumped-storage power station and the microgrid, is the predicted electric load power of the microgrid i at the t moment, is the actual thermal load power of the microgrid i at the t moment, is the actual electricity consumption load of the microgrid i at the t moment, is the reducible electric power of the microgrid i , is the transferable electric power of the microgrid i , is the reducible thermal load.

[0018] Preferably, in step S3, each microgrid entity is divided into different entities, and the pumped-storage power station is an independent entity. Different entities have their own interest goals respectively. By using the Nash game model, both individual interests and social interests can reach the optimal state, and the hybrid RM-MEDA algorithm and the distributed alternating submethod are used to solve the model, including: S301. As a type of cooperative game, the Nash game model can take into account both individual and collective interests and has a good distribution effect on the cooperative benefits among multiple entities. The solution of the Nash negotiation is the Pareto optimal solution. Construct the Nash game model: ; (10) In the formula, and are the benefits before and after the cooperation of entity i respectively. The constraint indicates that the benefit after cooperation is not lower than that before cooperation, ensuring that the interests of individuals are not damaged; In the cooperative game problem among the entities in the multi-microgrid integrated energy system, the mathematical expression of the Nash game is as follows: ; (11) In the formula, is the cost before the cooperation of microgrid i . and divide the cost after cooperation into two parts, namely the microgrid cooperative operation cost and the microgrid-to-microgrid transaction cost. The microgrid cooperative operation cost does not include the microgrid-to-microgrid transaction cost. By transforming the total cooperative cost into two modules, namely the cooperative cost minimization problem and the negotiation payment problem, it can be solved conveniently and quickly, and the shared power and payment electricity price of each microgrid are solved successively; S302. Solve the cooperative cost minimization problem, that is, the objective function C2, by using the hybrid RM-MEDA algorithm, obtain the value of the minimized cooperative cost C2 of the multi-microgrid and the optimal electric energy trading volume among the multi-microgrids, and substitute the result value into the negotiation payment problem to further solve the payment electricity price. The algorithm model and process are as follows: a. Initialization: Set the population size N, the number of subpopulations K, the mutation probability P, and set the algorithm stop condition. Let the iteration number t = 1, randomly initialize the population Y(t) and calculate the objective function values of the individuals in the initial population; b. Judge the stop condition: If the algorithm stop condition is satisfied, output the population Y(t); otherwise, go to the next step; c. Reproduction: Establish a probability model: Adopt the Local PCA clustering algorithm to cluster the individuals in the current population Y(t) into K classes, respectively forming K subpopulations S1, S2,..., SK, and establish a piecewise linear probability model for each subpopulation , ,…, ; Specifically as follows: ; (12) ; (13) ; (14) In the formula, is the mean vector of Sj, is the i-th principal component of Sj, that is, the eigenvector corresponding to the i-th largest eigenvalue of the covariance matrix of all antibodies in Sj; project all antibodies in Sj onto the (m - 1)-dimensional principal subspace of Sj, and denote the smallest line segment or hyperplane covering all projection points as ; Establish a piecewise linear probability model for the current population; Allocate the number of reproductions: Assume that the number of individuals after the generation of the i -th subpopulation is , where i = 1, 2, …, K , and Then can be obtained from the following formula: ; (15) In the formula, represents the line segment length (for two-objective problems) or hypervolume (for multi-objective problems) of the projection of all antibodies in population Si onto the first m - 1 principal component directions, represents taking the integer part of the real number x; Generate offspring using the clonal selection operator: For each subpopulation Si (i = 1, 2, …, K), generate an offspring population , denote the number of antibodies in as ; Generate offspring using the model sampling operator: For the probability model established for each subpopulation Si (i = 1, 2, …, K), ,…, ; Perform random sampling to generate an offspring population , and the number of sampled antibodies is ; d. Mutation: Denote , , let ; For each antibody in S, perform mutation operations using the mutation operator based on ε interval segmentation, with a mutation probability of P, to obtain an offspring population ; e. Selection: Let , Select N sub - individuals from Q using the selection method based on the m - nearest neighbor list to form a new population Y(t + 1), let t = t + 1, and go to the second step; S303. The costs of each micro - grid before cooperation are obtained through the solution method in S302, and their values are ; Since the Nash bargaining model in Equation (11) is non - linear and not easy to solve, the method of taking logarithms is used to transform Equation (11): ; (16) Considering the i cost of sharing electric energy between the micro - grid and other micro - grids, there is a key factor of trading electricity price among multiple micro - grids; introduce an auxiliary variable to decouple the trading electricity price: ; (17) S304. Based on the core idea of the distributed alternating direction method of multipliers algorithm, construct a distributed optimization model. By introducing Lagrange multipliers and penalty factors to coordinate the optimization process among multiple micro - grids; the distributed optimization model of micro - grid i is as follows: ; (18) ; (19) ; (20) Model (18) solves the bargaining payment problem according to the distributed iteration formula (19). When the iteration process of formula (20) satisfies the preset convergence condition less than or equal to the error , the trading electricity price among multiple micro - grids is obtained .

[0019] Example 2: As Figures 2 to 5 shown, the present invention solves the problem of minimizing the cooperation cost by bringing it into the hybrid RM - MEDA algorithm, and obtains the interactive electricity quantity between micro - grids and the minimum cooperation cost of multiple micro - grids; the interactive cost between each micro - grid is obtained by solving the bargaining payment problem. Under the solution of the bargaining payment, the trading electricity price between micro - grids can be obtained, and then the interactive cost between micro - grids can be known. The following table is the multi - mode comparison cost table 1:

[0020] As can be seen from Table 1, the electricity trading between microgrids realizes the complementary cooperation of different types of energy and can reduce the operating costs of each microgrid. More prominently, under the action of the pumped-storage power station, the cost of the microgrid system is greatly reduced, and its cost is reduced by about 1 / 4, which fully reflects the ability of the pumped-storage power station to cut peaks and fill valleys and absorb surplus energy.

[0021] In summary, the integrated energy multi-microgrid trading and scheduling method of the present invention based on a pumped-storage power station as a public facility uses the pumped-storage power station as a public infrastructure to integrate microgrids of various energy types, improving the ability of the microgrid to resist external risks and internal power fluctuations. The trading and scheduling method of the model proposed in the present invention establishes a new model for an integrated energy microgrid group, greatly increasing the revenue of the microgrid system, effectively exerting the energy of renewable energy, reducing carbon emissions and saving resources.

Claims

1. A comprehensive energy multi-microgrid transaction scheduling method based on a pumped storage power station as a public facility, characterized in that: The following steps are involved: S1, construct the operation model of each subject in the pumped storage power station and the microgrid; the microgrid includes gas turbines, waste heat boilers, wind turbines and photovoltaic panels as energy equipment; S2, construct a comprehensive energy trading and dispatching model for pumped storage power stations and microgrids; S3, optimizes the multi-agent distribution of benefits through the Nash game model, and solves the model using the hybrid RM-MEDA algorithm and the distributed alternating sub-method.

2. The integrated energy multi-microgrid transaction scheduling method based on a pumped storage power station as a public facility according to claim 1 is characterized in that: In step S1, constructing the operation model of each subject in the pumped storage power station and the microgrid includes: During operation, the pumped storage power station is subject to the change of reservoir capacity and its operating capacity constraints. The capacity change model of the reservoir is as follows: ；（1） Where Vr(t) and Vr(t-1) are the reservoir capacities before and after energy storage and release, respectively; Pbeng(t) and Pfa(t) are the storage power and power generation power of the pumped storage power station, respectively; ηbeng and ηfa are the pump efficiency and turbine efficiency, respectively; ρ is the water density; g is the acceleration of gravity; h is the rated water head, is the continuous running time; the constraints are as follows: ；（2） ；（3） In the formula, is the discharge power of the pumped storage power station at time t; is the pumping absorption power of the pumped storage power station at time t; and They are the 0-1 discharge state variable and pumping state variable of the pumped storage power station, and they are not 1 at the same time; and are the minimum storage capacity and the maximum storage capacity respectively.

3. The integrated energy multi-microgrid transaction scheduling method based on pumped storage power stations as public facilities according to claim 2 is characterized in that: In step S2, constructing a comprehensive energy trading dispatch model for a pumped storage power station and a microgrid includes: S201, construct the pumped storage power station cost model, microgrid cost model and carbon benefit model: The operating cost of pumped storage includes the initial investment and construction cost and the operation and maintenance cost. The mathematical model for converting the operating cost of the entire life cycle to the day is: ；（4） In the formula, is the unit power cost of the pumped storage system; is the rated power of energy storage; r is the discount rate of energy storage project; y is the energy storage life; The cost percentage coefficient for the operation and maintenance cost of the energy storage system; The microgrid objective function is constructed with the minimum total cost: ；（5） In the formula, represents the total cooperation cost of the microgrid group, excluding the interaction cost between each microgrid; i represents the i-th microgrid, and t represents the t-th moment; represents the total cost of interaction between microgrids; The cooperation cost of microgrid i; The interaction cost between microgrid i and other microgrids; It is the interaction book between microgrid and service provider; Cost of gas consumed for the microgrid; is the response cost of the microgrid flexible load; is the energy storage cost; a is the unit cost coefficient of gas; b, c, d are the response cost coefficients of the transferable electric load, the abatable electric load and the abatable thermal load respectively; is the energy storage cost coefficient; θ is the operation and maintenance cost of the pumped storage power station. Share the operation and maintenance costs of the pumped storage power station for the proportion of microgrid i; S202, constructing a power trading model: Constraints on power purchase and sale from microgrid to pumped storage power station: ；（6） In the formula, For Microgrid i exist t The electric power value transmitted to the pumped storage power station during the period, For Microgrid i exist t The power drawn from the pumped storage power station during the period takes a negative value. Unified Representation Microgrid i exist t The electric power value traded in the pumped storage power station during the period, with positive value being input and negative value being absorbed; Renewable Energy Constraints: ；（7） In the formula, , is the actual power generation of wind power and photovoltaic power, , Predict power generation for wind power and photovoltaic power; Interaction constraints between microgrids: ；（8） In the formula, for t Moment Micro Network i Towards j The interaction power, , For Microgrid i , j Upper and lower limits of interaction power; Power balance constraints, including thermal power balance and electrical power balance: ；（9） In the formula, is the interactive power between the pumped storage power station and the microgrid, For Microgrid i exist t The predicted electric load power at the time, For Microgrid i exist t The actual heat load power at the moment, For Microgrid i exist t The actual power load at any time, For Microgrid i The electrical power that can be reduced, For Microgrid i The transferable electrical power, To reduce heat load.

4. The integrated energy multi-microgrid transaction scheduling method based on a pumped storage power station as a public facility according to claim 3 is characterized in that: In step S3, each microgrid is divided into different entities, and the pumped storage power station is an independent entity. Different entities have their own interest goals. The Nash game model is used to optimize both individual interests and social interests, and the model is solved using the hybrid RM-MEDA algorithm and the distributed alternating sub-method, including: S301, build Nash game model: ；（10） In the formula, and Subject i The benefits before and after cooperation, the constraint means that the benefits after cooperation are not lower than before cooperation, ensuring that the interests of individuals are not harmed; In the cooperative game problem among the subjects of the multi-microgrid integrated energy system, the mathematical expression of the Nash game is as follows: ；（11） In the formula, For Microgrid i Cost before cooperation, and The cost after cooperation is divided into two parts, namely the cost of microgrid cooperation operation and the transaction cost between microgrids. The cost of microgrid cooperation operation does not include the transaction cost between microgrids. The shared power and paid electricity price of each microgrid are solved in turn. S302, solve the problem of minimum cooperation cost, i.e., objective function C2, by using the hybrid RM-MEDA algorithm, and obtain the value of the minimized cooperation cost C2 of the multi-microgrid and the optimal electricity transaction volume between the multi-microgrids , and substitute its result value into the negotiation payment problem to further solve the payment electricity price; the algorithm model and process are as follows: a. Initialization: Set the population size N, the number of subpopulations K, the mutation probability P, and the algorithm stop condition. Set the number of iterations t=1, randomly initialize the population Y(t) and calculate the objective function value of the individuals in the initial population; b. Determine the stopping condition: If the algorithm stops, the population Y(t) is output; otherwise, proceed to the next step; c. Reproduction: Establish a probability model: Use the Local PCA clustering algorithm to cluster the individuals in the current population Y(t) into K categories, forming K sub-populations S1, S2, ..., SK, and establish a piecewise linear probability model for each sub-population , ,…, ; The details are as follows: ；（12） ；（13） ；（14） In the formula, is the mean vector of Sj, is the i-th principal component of Sj, that is, the eigenvector corresponding to the i-th maximum eigenvalue of the covariance matrix of all antibodies in Sj; all antibodies in Sj are projected onto the (m-1)-dimensional principal subspace of Sj, and the minimum line segment or hyperplane covering all projection points is recorded as ;Establish a piecewise linear probability model for the current population; Assigning the number of reproduction: Assume that i The number of subpopulations generated is ,in, i =1,2,…, K ,and but It can be obtained from the following formula: ；（15） In the formula, represents the line segment length or hypervolume of the projection of all antibodies in population Si in the direction of the first m-1 principal components, It means taking the integer of real number x; Use the clonal selection operator to generate offspring: For each subpopulation Si (i=1,2,…,K), the offspring population is generated by the clonal selection operator ,remember The number of antibodies in ; The model sampling operator is used to generate offspring: for each subpopulation Si (i=1,2,…,K) the probability model is established , ,…, ; Random sampling to generate offspring population , the number of sampled antibodies is ; d. Variation: , ,make ; For each antibody in S, a mutation operator based on ε interval segmentation is used to perform mutation operation with a mutation probability of P, and the offspring population is obtained ; e. Select: , use the selection method based on the m-nearest neighbor list to select N sub-individuals from Q to form a new population Y(t+1), let t=t+1, and go to the second step; S303, the cost of each microgrid before cooperation is obtained by the solution method in S302, and its value is ; Use the logarithm method to transform equation (11): ；（16） Consider microgrids i Share electricity with other microgrids to pay for costs , then there is a key factor of transaction electricity price between multiple microgrids; introducing auxiliary variables Decoupling of transaction electricity prices: ；（17） S304, based on the core idea of ​​the distributed alternating sub-method algorithm, constructs a distributed optimization model by introducing Lagrange multipliers , Penalty Factor To coordinate the optimization process between multiple microgrids; i The distributed optimization model is as follows: ；（18） ；（19） ；（20） Model (18) solves the negotiation payment problem according to the distributed iterative formula (19). When the iterative process of formula (20) satisfies the preset convergence condition, the error is less than or equal to When the transaction electricity price between multiple microgrids is obtained .

Images