A micro energy grid group hierarchical optimization control method and system based on scene construction

Abstract

The application discloses a micro energy network group hierarchical optimization control method and system based on scene construction, and steps are as follows: mathematical modeling is respectively conducted on each type of energy in the micro energy network; probability density distribution functions of wind power, photovoltaic and load are established; five kinds of Copula functions are introduced, and the joint distribution relationship of random variables of wind power and photovoltaic is respectively established; the goodness of fit of each model is judged by using the Euclidean distance method, and the optimal Copula function is determined; based on the optimal Copula function, typical application scenes are generated after sampling and aggregation; considering the loss of abandoned wind and light, a double-layer coordinated optimization model based on two scenes of winter and summer is established, with the minimum micro energy network group operation cost as the target; the target cascade analysis method is used for iterative solution, and an operation plan is generated; the operation plan is sent to a power grid control center for power flow calculation and safety checking; if the requirements and limits of the power grid are met, each member of the micro energy network executes the generated operation plan, and if the requirements and limits of the power grid are not met, optimization solution is performed again.

Description

Technical Field

[0001] This invention relates to power system optimization operation technology, and in particular to a hierarchical optimization control method and system for micro energy grid groups based on scenario construction. Background Technology

[0002] With the development of modern society, environmental protection has become a major concern, and the development and utilization of clean and efficient energy to achieve sustainable development has become a key focus of research in the energy sector. On the one hand, customer demand for energy, in addition to electricity, requires other forms such as cooling, heating, and natural gas, leading to the introduction of various new distributed energy forms, such as combined heat and power (CHP), distributed energy stations, and new energy storage, into regional power distribution networks. On the other hand, the intermittent and fluctuating nature of clean energy sources such as wind power and solar power poses significant challenges to the safe and stable operation of regional power distribution networks and microgrids.

[0003] Faced with this situation, on the one hand, traditional regional distribution networks cannot mobilize new flexible resources within the region, thus failing to fully utilize clean energy and resulting in wind and solar power curtailment; on the other hand, traditional distribution network dispatch and control systems cannot integrate data from heterogeneous distributed energy sources into the system, thus failing to achieve unified, coordinated, and optimized control of the entire regional micro-energy network. Therefore, establishing a hierarchical optimization control method and system for micro-energy networks is of great significance for improving the ability of micro-energy networks to absorb clean energy such as wind and solar power, and for ensuring their economically stable operation in a market environment.

[0004] After various distributed energy sources from different entities enter the power grid, the optimization function of the microgrid coordination and optimization control system aims to minimize the daily operating cost of the microgrid cluster within its region, providing decision-makers with multiple optimized scheduling schemes. However, in microgrids containing various clean energy sources, the source load is affected by numerous uncertainties; ignoring these uncertainties will reduce the accuracy of planning and operation. Forecasts for wind and solar power will introduce prediction errors. Traditional integrated energy control systems generally do not consider the costs arising from wind and solar prediction errors. Summary of the Invention

[0005] Purpose of the Invention: The purpose of this invention is to provide a hierarchical optimization control method and system for micro energy grids based on scenario construction. Based on wind and solar scenario construction technology, considering the uncertainty and correlation of wind and solar power output, and taking into account the cost of wind and solar curtailment penalties and the cost of wind and solar prediction errors, a two-layer optimization model for micro energy grids is established to achieve hierarchical coordinated optimization of micro energy grids, realize the access of energy equipment of various energy forms to the system, improve the economic benefits of micro energy grids, and increase the clean energy consumption rate of micro energy grid clusters.

[0006] Technical Solution: This invention provides a hierarchical optimization control method and system for microgrid clusters based on scenario construction. Based on wind and solar scenario construction technology, considering the uncertainty and correlation of wind and solar power output, and taking into account the penalty costs of wind and solar curtailment and the cost of wind and solar prediction errors, a two-layer optimization model for the microgrid is established to achieve hierarchical coordinated optimization of the microgrid. The microgrid system encompasses various energy supply, storage, and conversion devices. Energy forms include cooling, heating, electricity, and gas. Energy conversion devices include wind power, photovoltaics, combined heat and power systems, ground source heat pumps, electric chillers, etc., and include batteries, gas storage tanks, and thermal and cold storage devices. Figure 2 , Figure 3 .

[0007] S1. First, models for each energy form must be established. In economic analysis, each energy form is categorized into the cost-optimal model in terms of total price. In power grid operation and dispatch, the relationship between each energy form and power output is obtained and then incorporated into the constraints of the optimization model.

[0008] For example, energy storage models:

[0009]

[0010] In the formula, E XS , The energy stored in the energy storage device at the current moment and the energy stored in the past moment; P store , Storage power, storage efficiency; P release , Release power, release efficiency.

[0011] Such as electric refrigeration units:

[0012]

[0013] In the formula, The magnitude of the cooling output generated by the electric chiller. η is the electrical power consumed by the electric chiller. EC This refers to the conversion efficiency of the electric chiller.

[0014] S2. Analyze the uncertainties of wind power, photovoltaics, and loads in the microgrid. Establishing uncertainty models for wind power, photovoltaics, and loads is the foundation of the entire method. For wind power, the output power of the wind turbine is closely related to the wind speed. The Weibull two-parameter distribution curve is used to describe wind speed; it is a unimodal curve with two parameters, and the wind speed probability density function can be expressed as:

[0015]

[0016] In the formula, v is the wind speed, and k and c represent two important parameters in the wind speed distribution, namely the shape parameter and the scale parameter, respectively; their values ​​can be calculated or estimated using historical wind speed data.

[0017] For photovoltaics, the output power is closely related to the light intensity. Using a Beta distribution to describe the distribution of light intensity, the probability density function of light intensity can be expressed as:

[0018]

[0019] In the formula, Γ represents the gamma function; τ and υ represent shape parameters; I is the illumination intensity, and Imax represents the maximum illumination intensity;

[0020] For load, the normal distribution can represent the load distribution over a period of time relatively well. Its distribution parameters are affected by regional climate, and the probability density function of the load distribution can be expressed as:

[0021]

[0022] In the formula, P represents the load power; σ P Indicates the average load power; μ P This represents the standard deviation of load power.

[0023] S3. Use various Copula functions to describe the correlation of random variables in wind power and photovoltaics. According to Sklar's theorem: Assume random variables x1, x2, ... x... N The joint distribution function is H, and the marginal distributions are F1(x1), F2(x2), ..., F N (x N Then there exists a copula function C such that... Note that c(u1,u2,…,u...) n The domain of is [0,1]. n ;

[0024] H(x1,x2,...x N )=C(F1(x1),F2(x2),…,F N (x N (6)

[0025] If F1(x1), F2(x2), ..., F N (x N If F1(x1), F2(x2), ..., F2(x2) are continuous, then the Copula function C is uniquely determined; otherwise, F1(x1), F2(x2), ..., F2(x2) are not continuous. N (x N If is a univariate distribution function, then C is the corresponding Copula function.

[0026] Five Copula functions are introduced: normal Copula, t-Copula, Gumbel-Copula, Clayton-Copula, and Frank-Copula.

[0027] Normal Copula function:

[0028]

[0029] -1≤ρ≤1 represents the correlation coefficient, while the upper limit of integration φ of the Copula function... -1 It represents the standard inverse normal function based on the correlation coefficient.

[0030] t-Copula function:

[0031]

[0032] The coefficient k represents the degrees of freedom, and -1 ≤ ρ ≤ 1 represents the correlation coefficient, while the upper limit of integration... It represents the inverse function of the standard t-distribution with k degrees of freedom.

[0033] Frank-Copula function:

[0034]

[0035] In the formula, α represents the correlation coefficient. If α approaches 0, it means the two random variables are uncorrelated; if α is greater than 0, it indicates a positive correlation; and if α is less than 0, it indicates a negative correlation. Note that α ≠ 0.

[0036] Gumbel-Copula function:

[0037]

[0038] In this Copula function, α ≥ 1, where α = 1 indicates that the two random variables are independent; if α approaches infinity...

[0039] This indicates that the two random variables are perfectly dependent. The upper tail correlation coefficient of the Gumbel-Copula is... The lower tail correlation coefficient is 0, so the Gumbel-Copula is effective in describing random variables with lower tail correlation. The Gumbel-Copula is an asymmetric distribution.

[0040] Clayton-Copula:

[0041]

[0042] In the Clayton-Copula, α ≥ 1, and α ≠ 0. The lower tail correlation coefficient of the Clayton-Copula is... The upper tail correlation coefficient is 0, so the Clayton-Copula is suitable for describing random variables with upper tail correlation. The Clayton-Copula is also an asymmetric distribution.

[0043] Based on the established probability density function, the marginal cumulative distribution functions of wind power and photovoltaic power output are determined respectively. Then, the random variables are transformed into uniformly distributed random numbers on [0,1] using the marginal probability cumulative distribution function transformation. Subsequently, various Copula function forms are substituted, and the unknown parameters of the Copula function are solved using the two-step maximum likelihood method. Thus, the specific forms of various Copula functions are obtained.

[0044] S4. Use the Euclidean distance method to select the optimal Copula function. The Euclidean distance method compares the Euclidean distance of various Copula functions with the empirical Copula function generated from the sample data. The smaller the Euclidean distance, the better the fit of this Copula function.

[0045] Suppose two vectors X = {x1, x2, ..., xn} n}, Y = {y1, y2, ... y n In this context, Euclidean distance represents the straight-line distance between two points. The expression for Euclidean distance is as follows:

[0046]

[0047]

[0048] These are the fit metrics: Euclidean distance d and maximum distance di. max The smaller the two values ​​above, the better the model matches the original actual data distribution. Maximum distance d max This represents the empirical Copula distribution value and the maximum distance between the Copula distribution values.

[0049] By comparing the Euclidean distance d and the maximum distance d of various Copula functions max The optimal Copula function can be determined.

[0050] S5. Based on the optimal Copula function, to obtain representative operating scenarios, it is necessary to reduce the number of sampled scenarios by using the K-means clustering method. This is a commonly used scenario reduction method, the basic idea of ​​which is that each cluster center represents a typical scenario. The appropriate number of clusters is crucial to the scenario construction technique, and the DBI index is an important method for measuring this, its expression is as follows:

[0051]

[0052] In the formula, Let be the average intra-cluster distance of cluster i. Let be the distance between two cluster centers, and k be the number of clusters. Based on actual field data and extensive calculation results, two typical scenarios were identified: a winter scenario and a summer scenario.

[0053] S6. Based on scenario construction, according to the uncertainty analysis of clean energy output and load power in the micro energy network, the optimization model divides the absorption of clean energy by the regional micro energy network into two layers: the upper layer is the coordinated optimization of the micro energy network group, and the lower layer is the absorption optimization within a single micro energy network.

[0054] The upper-level absorption optimization aims to minimize the daily operating cost of the micro-energy grid cluster within the region. The cost function includes energy interaction costs, energy storage costs, equipment operating costs, controllable load adjustment costs, and wind and solar forecast error costs.

[0055]

[0056]

[0057]

[0058]

[0059]

[0060]

[0061] In the formula, N is the number of microgrids, m is the type of energy, and M = {e, g, h, c}; These are the energy interaction costs between microgrids, energy storage costs, equipment operating costs, controllable load regulation costs, and wind and solar forecasting error costs; It is the transaction cost of the m-th energy source; These are the input and output quantities of the m-th energy source, respectively. It is the operation and maintenance coefficient of the m-th type of energy storage equipment; These are the charging and discharging power of the m-th energy source; These are the operating power of the CCHP system, ground source heat pump, and electric chiller equipment, respectively. These are the operating cost coefficients for CCHP systems, ground source heat pumps, and electric chiller equipment, respectively. It is the cost coefficient for the adjustment and compensation of the m-th controllable load; The m-th type of controllable load participates in the regulation; E i,t Indicates the prediction error for wind and solar power generation, γ u To overestimate the cost factor resulting from overestimating power generation capacity, γ d The cost factor for off-peak power generation. s is the scenario marker, with winter and summer discussed separately.

[0062] Constraints: The constraints for energy storage and equipment are as follows:

[0063]

[0064] In the formula, These are the upper and lower limits of energy storage for the m-th energy source, respectively. These are the upper and lower limits of the energy release for the m-th energy source, respectively. The upper and lower limits of the output of the i-th type of equipment are not specified.

[0065] Controllable load constraints

[0066]

[0067] In the formula, These are the minimum and maximum load power of the m-th energy source, respectively.

[0068] The optimization of lower-level energy consumption aims to minimize the daily operating cost within a single microgrid. The cost function includes energy purchase cost, equipment operating cost, wind and solar curtailment penalty cost, and wind and solar forecast error cost.

[0069]

[0070]

[0071]

[0072]

[0073] In the formula, This represents the energy purchase cost in the s-th scenario of the microgrid. It is the equipment operating cost; It's the cost of penalties for abandoning wind and solar power; These are time-of-use electricity pricing and gas pricing; These are the electricity and gas purchases, respectively. These are the operating power of CCHP systems, ground source heat pumps, electric chillers, energy storage, gas storage, thermal storage, and cold storage equipment; These are the operating cost coefficients for CCHP systems, ground source heat pumps, electric chillers, energy storage, gas storage, thermal storage, and cold storage equipment; cp is the penalty coefficient for wind and solar curtailment. These are the amounts of wind and solar power that are being curtailed.

[0074] The constraints are:

[0075] Constraints on wind, solar, and geothermal power output

[0076]

[0077] In the formula, These are the maximum outputs of wind power, solar power, and geothermal power, respectively.

[0078] Unit output upper and lower limit constraints

[0079]

[0080] In the formula, These are the upper and lower limits of CCHP's output operation, respectively; These are the upper and lower limits of the output operation of the ground source heat pump. These are the upper and lower limits of the electric chiller's output operation; These are the upper and lower limits of the battery's energy storage capacity; These are the upper and lower limits of gas storage in the gas storage tank; These are the upper and lower limits of thermal storage for the thermal storage device; These are the upper and lower limits of the cold storage device.

[0081] Power balance constraints

[0082]

[0083] In the formula, These are the electrical, thermal, and cooling conversion coefficients of CCHP, and the electro-cooling coefficient. These are discharge power and energy storage power, respectively. These are the venting and storage capacities, respectively. These are the heat release and heat storage power, respectively. These are cooling output power and cooling storage power, respectively.

[0084] S7. The two-layer optimal model is solved using objective cascade analysis. Considering that the scheduling strategies of the upper and lower layers are formulated by different operating entities and that the scheduling results affect each other, a hierarchical optimization method based on objective cascade analysis is adopted for solving the model. During the solution process, consistency constraints are added to the objective function in the form of penalty functions, and the solution is obtained by iteratively updating the penalty multipliers until consistency is finally achieved.

[0085] The objective function after adding the Lagrange form penalty function is shown below.

[0086] Objective function of the upper-level micro-energy network cluster:

[0087]

[0088] The objective function of the nth microgrid in the upper layer is:

[0089]

[0090] In the formula, λ n,t ω n,t The penalty multipliers at time t are respectively, with symbols... For Hadama accumulation, Let s represent the square of the norm, s = [s (1) ,s (2) ,s (3) ] T Scaling factor These represent the planned interactive power of the nth microgrid and the optimized interactive power transmitted from the nth microgrid to the upper-level system, respectively. n =[s n (1) ,s n (2) ,s n (3) ] T Let n be the scaling factor for the nth microgrid. These represent the planned interactive power of the nth microgrid and the interactive power transferred to the nth microgrid after optimization by the upper-level model, respectively.

[0091] The solution process is as follows: Figure 5 As shown, the steps are as follows.

[0092] S7.1: Initialize system parameters, set scaling factor, initial penalty multiplier, and set iteration count k=1.

[0093] S7.2: If k=1, solve the upper-level micro-energy network group optimization problem and pass the obtained interactive power to the lower level.

[0094] S7.3: Solve the optimization problem within the lower-level microgrid. Due to the existence of the penalty function, the optimized interactive power in the lower level will be close to the value passed from the upper level.

[0095] S7.4: Check if the interaction power between the upper and lower layers meets the conditions. If it does, stop the iteration and output the calculation results; otherwise, continue to execute S7.5.

[0096] S7.5: Let the iteration number k = k + 1, update the penalty multiplier, and return to S7.2 to continue the iteration.

[0097] During the solution process, consistency constraints are added to the objective function as penalty functions. The solution is then iteratively solved by updating the penalty multipliers until consensus is reached.

[0098] S8. The microgrid control center generates an operation plan after calculation and sends it to the power grid control center. The power grid control center performs power flow calculation and safety verification calculation to check whether the various requirements and restrictions of power grid operation are met.

[0099] S9. If the calculation results meet the requirements and limitations of the power grid, each member of the microgrid executes the generated operation plan; if not, return to step S7 for further optimization.

[0100] S10: After each member of the microgrid executes its operation plan, it records the execution process and measures the consumption of various types of energy.

[0101] A scenario-based hierarchical optimization control system for microgrid groups includes the following modules:

[0102] The first modeling module is used to establish the physical framework model of the regional power grid multi-micro energy network system; the physical framework model includes various energy supply devices and energy storage devices.

[0103] The first analysis module is used to analyze the uncertainties of wind power, photovoltaics, and load in microgrids; it establishes a probability model for the uncertainties, using a random variable probability model to establish the probability density functions of three quantities: wind speed, solar irradiance, and load.

[0104] The first calculation module is used to establish the Copula function that connects random variables between wind power and photovoltaic power generation. It establishes specific models of five Copula functions: normal Copula, t-Copula, Gumbel-Copula, Clayton-Copula, and Frank-Copula, and solves for their specific parameters.

[0105] The second calculation module is used to select the most suitable Copula function to describe the correlation of random variables in wind power and photovoltaics using the Euclidean distance method. First, it simulates and samples the distribution values ​​according to the five Copula functions, and then compares them with the actual collected field data values ​​to calculate the Euclidean distance of each. Then, it compares the Euclidean distance of the five types of Copula functions with the Euclidean distance of the actual sampled data, and the one with the smallest Euclidean distance is the optimal Copula function model.

[0106] Scene generation module: Used to generate basic scenarios. After determining the optimal Copula function, it samples and aggregates the joint probability distribution of the optimal Copula function to generate typical application scenarios.

[0107] The second modeling module is used to establish a two-layer coordinated optimization model for regional micro-energy networks. The optimization of the absorption of clean energy by regional micro-energy networks is divided into two-layer optimization models. The optimization objective of the upper-layer model is to minimize the daily operating cost of the micro-energy network group in the region, while the optimization objective of the lower-layer model is to minimize the daily operating cost within a single micro-energy network.

[0108] The third calculation module uses a hierarchical optimization method based on objective cascade analysis to solve the problem. During the solution process, consistency constraints are added to the objective function in the form of penalty functions. The solution is obtained by iteratively updating the penalty multipliers, and finally, consistency is achieved.

[0109] First discrimination module: The microgrid control center generates an operation plan after calculation and sends it to the power grid control center. The power grid control center performs power flow calculation and safety verification calculation to determine whether it meets the various requirements and restrictions of power grid operation. If it meets the requirements, it proceeds to step S7. If it does not meet the requirements, it returns to step S5 to calculate again.

[0110] Second discrimination module: If the calculation results meet the requirements and limitations of the power grid, each member of the microgrid executes the generated operation plan;

[0111] The sixth calculation module is used to record the execution process and measure the consumption of various types of energy after each member of the micro energy network executes the operation plan.

[0112] In the first modeling module, establishing the physical framework model of the regional power grid multi-micro energy network system refers to performing mathematical modeling for each type of energy, establishing the functional relationships of energy, power, initial conditions, and boundary conditions for each energy form, and deriving the energy conversion methods and relationships between each energy form.

[0113] The uncertainties mentioned include uncertainties in renewable energy output and uncertainties in load power.

[0114] A computer storage medium storing a computer program that, when executed by a processor, implements the aforementioned scenario-based hierarchical optimization control method for micro-energy grid clusters.

[0115] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described scenario-based hierarchical optimization control method for micro-energy grid clusters.

[0116] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0117] 1. By using the Copula function to generate typical power output scenarios, the variation patterns of wind power and photovoltaic power output can be reasonably characterized, providing methods and tools for the planning and operation of new energy sources in regional power grids;

[0118] 2. For various devices of different entities in the micro energy network, hierarchical coordinated control has been achieved, and control authority has been delegated, which has effectively activated the enthusiasm of different entities to participate in the operation of the regional energy network and to consume new energy.

[0119] 3. The optimization model of this invention takes into account the cost of wind and solar curtailment penalties and the cost of wind and solar prediction errors. It can plan and schedule under the consideration of the coupling of multiple energy sources such as electricity, heat and gas, further reducing the amount of wind and solar curtailment. This not only increases the absorption of new energy sources, but also improves the economic benefits of micro energy grid clusters.

[0120] 4. Information on equipment using various energy forms, including electricity, gas, and heat, such as combined heat and power systems, ground source heat pump systems, electric chillers, batteries, gas storage tanks, and thermal and cold energy storage devices, is connected to the microgrid coordination and control system through this method and the communication system provided by the system, thereby achieving unified, coordinated, and optimized operation among various energy forms. Attached Figure Description

[0121] Figure 1 This is a flowchart of the method.

[0122] Figure 2 Diagram of the physical architecture of a microgrid;

[0123] Figure 3 A schematic diagram of the system structure between multiple micro-energy grids;

[0124] Figure 4 Diagram of a microgrid control system;

[0125] Figure 5 Flowchart for solving the objective using the cascade method. Detailed Implementation

[0126] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0127] A hierarchical optimization control method for renewable energy consumption based on scenario-based micro-energy grid clusters includes the following steps:

[0128] S1: Establish a physical framework model of the regional micro-energy grid system, perform mathematical modeling for each type of energy, establish functional relationships between energy, power, initial conditions, and boundary conditions for each energy form, and derive the energy conversion methods and relationships between each energy form; the physical framework model includes various energy supply devices, energy storage devices, and energy conversion devices.

[0129] Microgrids integrate energy interconnection, conversion, coupling, and storage functions. In the energy transmission, storage, and distribution stages, they provide energy to energy users or facilitate energy trading with higher-level energy networks at a certain price. The regional microgrid system established in this embodiment encompasses various energy supply devices, energy storage devices, and energy conversion devices. In addition to distributed energy sources such as wind power and photovoltaics, the energy supply devices also include geothermal energy. Energy conversion devices include combined heat and power systems, ground source heat pumps, electric chillers, etc., and include batteries, gas storage tanks, and thermal and cold storage devices. Figure 2 As shown.

[0130] Therefore, the first step is to establish models for each energy form. This involves establishing functional relationships between various quantities related to energy, power, initial conditions, and boundary conditions for each energy form, and deriving the energy conversion methods and relationships between them. In economic analysis, each energy form is categorized into the cost-optimal model in terms of total price. In power grid operation and scheduling, their relationship with electrical power is obtained and then incorporated into the constraints of the optimization model.

[0131] For example, energy storage models:

[0132]

[0133] In the formula, E XS , The energy stored in the energy storage device at the current moment and the energy stored in the past moment; P store , Storage power, storage efficiency; P release , Release power, release efficiency.

[0134] Such as electric refrigeration units:

[0135]

[0136] In the formula, The magnitude of the cooling output generated by the electric chiller. η is the electrical power consumed by the electric chiller. EC This refers to the conversion efficiency of the electric chiller.

[0137] Multi-energy microgrid clusters primarily rely on energy interaction between individual microgrids. Through energy interaction with other multi-energy microgrids, they achieve their own energy balance and ensure the stability of the overall system. The entire energy system features high energy supply reliability and strong fault resistance. The micro-energy grid cluster scheduling method can be simply represented by a framework diagram. Multiple micro-energy grids are interconnected through energy stations. There is no energy or information exchange between individual multi-energy microgrids; they are directly connected to the energy stations of the cluster system. The model discussed in this embodiment mainly focuses on natural gas networks and power networks, deepening the coupling between multiple micro-energy grids to jointly provide users with various forms of energy such as cooling, heating, electricity, and gas. During normal operation, each micro-energy system optimizes its operating strategy based on its own operating goals. When its own operating goals cannot be met, it reports its energy demand to the energy station and directly purchases energy from the backbone network. It has the advantages of convenient control, simple information flow, and good overall economy. Figure 3 As shown.

[0138] S2: Establishing uncertainty models for wind power, photovoltaics, and load is the foundation of the entire method. For wind power, the output power of the wind turbine is closely related to the wind speed. The Weibull two-parameter distribution curve is used to describe the wind speed; it is a unimodal curve with two parameters, and the wind speed probability density function can be expressed as:

[0139]

[0140] In the formula, k and c represent two important parameters in wind speed distribution, namely the shape parameter and the scale parameter; their values ​​can be calculated or estimated using historical wind speed data.

[0141] For photovoltaics, the output power is closely related to the light intensity. Using a Beta distribution to describe the distribution of light intensity, the probability density function of light intensity can be expressed as:

[0142]

[0143] In the formula, Γ represents the gamma function; τ and υ represent shape parameters; and Imax represents the maximum illumination intensity.

[0144] For loads, the normal distribution can represent the load distribution over a period of time relatively well. Its distribution parameters are affected by regional climate, and the probability density function of the load distribution can be expressed as:

[0145]

[0146] In the formula, P represents the load power; σ P Indicates the average load power; μ P This represents the standard deviation of load power.

[0147] S3: Then, various Copula functions are used to describe the correlation of random variables in wind power and photovoltaics. According to Sklar's theorem: Assume random variables x1, x2, ... x N The joint distribution function is H, and the marginal distributions are F1(x1), F2(x2), ..., F N (x N Then there exists a copula function C such that... Note that c(u1,u2,…,u...) n The domain of is [0,1]. n ;

[0148] H(x1,x2,...x N )=C(F1(x1),F2(x2),…,F N (x N (6)

[0149] If F1(x1), F2(x2), ..., F N (x N If F1(x1), F2(x2), ..., F2(x2) are continuous, then the Copula function C is uniquely determined; otherwise, F1(x1), F2(x2), ..., F2(x2) are not continuous. N (x N If is a univariate distribution function, then C is the corresponding Copula function.

[0150] Five Copula functions are introduced: normal Copula, t-Copula, Gumbel-Copula, Clayton-Copula, and Frank-Copula.

[0151] Normal Copula function:

[0152]

[0153] -1≤ρ≤1 represents the correlation coefficient, while the upper limit of integration φ of the Copula function... -1 It represents the standard inverse normal function based on the correlation coefficient.

[0154] t-Copula function:

[0155]

[0156] The coefficient k represents the degrees of freedom, and -1 ≤ ρ ≤ 1 represents the correlation coefficient, while the upper limit of integration... It represents the inverse function of the standard t-distribution with k degrees of freedom.

[0157] Frank-Copula function:

[0158]

[0159] In the formula, α represents the correlation coefficient. If α approaches 0, it means the two random variables are uncorrelated; if α is greater than 0, it indicates a positive correlation; and if α is less than 0, it indicates a negative correlation. Note that α ≠ 0.

[0160] Gumbel-Copula function:

[0161]

[0162] In this Copula function, α ≥ 1, where α = 1 indicates that the two random variables are independent; if α approaches infinity...

[0163] This indicates that the two random variables are perfectly dependent. The upper tail correlation coefficient of the Gumbel-Copula is... The lower tail correlation coefficient is 0, so the Gumbel-Copula is effective in describing random variables with lower tail correlation. The Gumbel-Copula is an asymmetric distribution.

[0164] Clayton-Copula:

[0165]

[0166] In the Clayton-Copula, α ≥ 1, and α ≠ 0. The lower tail correlation coefficient of the Clayton-Copula is... The upper tail correlation coefficient is 0, so the Clayton-Copula is suitable for describing random variables with upper tail correlation. The Clayton-Copula is also an asymmetric distribution.

[0167] Based on the established probability density function, the marginal cumulative distribution functions of wind power and photovoltaic power output are determined respectively. Then, the random variables are transformed into uniformly distributed random numbers on [0,1] using the marginal probability cumulative distribution function transformation. Subsequently, various Copula function forms are substituted, and the unknown parameters of the Copula function are solved using the two-step maximum likelihood method. Thus, the specific forms of various Copula functions are obtained.

[0168] S4: Subsequently, the Euclidean distance method is used to select the optimal Copula function. The Euclidean distance method compares the Euclidean distances of various Copula functions with those of empirical Copula functions generated from the sample data. The smaller the Euclidean distance, the better the fit of the Copula function.

[0169] Suppose two vectors X = {x1, x2, ..., xn} n}, Y = {y1, y2, ... y nIn this context, Euclidean distance represents the straight-line distance between two points. The expression for Euclidean distance is as follows:

[0170]

[0171]

[0172] These are the fit metrics: Euclidean distance d and maximum distance di. max The smaller the two values ​​above, the better the model matches the original actual data distribution. Maximum distance d max This represents the empirical Copula distribution value and the maximum distance between the Copula distribution values.

[0173] By comparing the Euclidean distance d and the maximum distance d of various Copula functions max The optimal Copula function can be determined.

[0174] S5: Based on the optimal Copula function, to obtain representative operating scenarios, it is necessary to reduce the number of sampled scenarios by using the K-means clustering method. This is a commonly used scenario reduction method, the basic idea of ​​which is that each cluster center represents a typical scenario. The appropriate number of clusters is crucial to the scenario construction technique, and the DBI index is an important method for measuring this, its expression is as follows:

[0175]

[0176] In the formula, Let be the average intra-cluster distance of cluster i. Let be the distance between two cluster centers, and k be the number of clusters. Based on actual field data and extensive calculation results, two typical scenarios were identified: a winter scenario and a summer scenario.

[0177] S6: Based on scenario construction, according to the uncertainty analysis of clean energy output and load power in the micro energy network, the optimization model divides the operation optimization of the regional micro energy network group into two layers: the upper layer is the coordination optimization of the micro energy network group, and the lower layer is the absorption optimization within a single micro energy network.

[0178] The upper-level absorption optimization aims to minimize the daily operating cost of the micro-energy grid cluster within the region. The cost function includes energy interaction costs, energy storage costs, equipment operating costs, controllable load adjustment costs, and wind and solar forecast error costs.

[0179]

[0180]

[0181]

[0182]

[0183]

[0184]

[0185] In the formula, N is the number of microgrids, m is the type of energy, and M = {e, g, h, c}; These are the energy interaction costs between microgrids, energy storage costs, equipment operating costs, controllable load regulation costs, and wind and solar forecasting error costs; It is the transaction cost of the m-th energy source; These are the input and output quantities of the m-th energy source, respectively. It is the operation and maintenance coefficient of the m-th type of energy storage equipment; These are the charging and discharging power of the m-th energy source; These are the operating power of the CCHP system, ground source heat pump, and electric chiller equipment, respectively. These are the operating cost coefficients for CCHP systems, ground source heat pumps, and electric chiller equipment, respectively. It is the cost coefficient for the adjustment and compensation of the m-th controllable load; The m-th type of controllable load participates in the regulation; E i,t Indicates the prediction error for wind and solar power generation, γ u To overestimate the cost factor resulting from overestimating power generation capacity, γ d The cost factor for off-peak power generation. s is the scenario marker, with winter and summer discussed separately.

[0186] Constraints: The constraints for energy storage and equipment are as follows:

[0187]

[0188] In the formula, These are the upper and lower limits of energy storage for the m-th energy source, respectively. These are the upper and lower limits of the energy release for the m-th energy source, respectively. The upper and lower limits of the output of the i-th type of equipment are not specified.

[0189] Controllable load constraints

[0190]

[0191] In the formula, These are the minimum and maximum load power of the m-th energy source, respectively.

[0192] The optimization of lower-level energy consumption aims to minimize the daily operating cost within a single microgrid. The cost function includes energy purchase cost, equipment operating cost, wind and solar curtailment penalty cost, and wind and solar forecast error cost.

[0193]

[0194]

[0195]

[0196]

[0197] In the formula, This represents the energy purchase cost in the s-th scenario of the microgrid. It is the equipment operating cost; It's the cost of penalties for abandoning wind and solar power; These are time-of-use electricity pricing and gas pricing; These are the electricity and gas purchases, respectively. These are the operating power of CCHP systems, ground source heat pumps, electric chillers, energy storage, gas storage, thermal storage, and cold storage equipment; These are the operating cost coefficients for CCHP systems, ground source heat pumps, electric chillers, energy storage, gas storage, thermal storage, and cold storage equipment; cp is the penalty coefficient for wind and solar curtailment. These are the amounts of wind and solar power that are being curtailed.

[0198] The constraints are:

[0199] Constraints on wind, solar, and geothermal power output

[0200]

[0201] In the formula, These are the maximum outputs of wind power, solar power, and geothermal power, respectively.

[0202] Unit output upper and lower limit constraints

[0203]

[0204] In the formula, These are the upper and lower limits of CCHP's output operation, respectively; These are the upper and lower limits of the output operation of the ground source heat pump; These are the upper and lower limits of the electric chiller's output operation; These are the upper and lower limits of the battery's energy storage capacity; These are the upper and lower limits of gas storage in the gas storage tank; These are the upper and lower limits of thermal storage for the thermal storage device; These are the upper and lower limits of the cold storage device.

[0205] Power balance constraints

[0206]

[0207] In the formula, These are the electrical, thermal, and cooling conversion coefficients of CCHP, and the electro-cooling coefficient. These are discharge power and energy storage power, respectively. These are the venting and storage capacities, respectively. These are the heat release and heat storage power, respectively. These are cooling output power and cooling storage power, respectively.

[0208] S7: The objective cascade analysis method is used to solve this two-layer optimal model. Considering that the scheduling strategies of the upper and lower layers are formulated by different operating entities and that the scheduling results affect each other, a hierarchical optimization method based on objective cascade analysis is adopted for solving the model. During the solution process, consistency constraints are added to the objective function in the form of penalty functions, and the solution is obtained by iteratively updating the penalty multipliers until consistency is finally achieved.

[0209] The objective function after adding the Lagrange form penalty function is shown below.

[0210] Objective function of the upper-level micro-energy network cluster:

[0211]

[0212] The objective function of the nth microgrid in the upper layer is:

[0213]

[0214] In the formula, λ n,t ω n,t The penalty multipliers at time t are respectively, with symbols... For Hadama accumulation, Let s represent the square of the norm, s = [s (1) ,s (2) ,s (3) ] T Scaling factor These represent the planned interactive power of the nth microgrid and the optimized interactive power transmitted from the nth microgrid to the upper-level system, respectively. n =[s n (1) ,s n (2) ,s n (3) ] T Let n be the scaling factor for the nth microgrid. These represent the planned interactive power of the nth microgrid and the interactive power transferred to the nth microgrid after optimization by the upper-level model, respectively.

[0215] The solution process is as follows: Figure 5As shown, the steps are as follows.

[0216] S7.1: Initialize system parameters, set scaling factor, initial penalty multiplier, and set iteration count k=1.

[0217] S7.2: If k=1, solve the upper-level micro-energy network group optimization problem and pass the obtained interactive power to the lower level.

[0218] S7.3: Solve the optimization problem within the lower-level microgrid. Due to the existence of the penalty function, the optimized interactive power in the lower level will be close to the value passed from the upper level.

[0219] S7.4: Check if the interaction power between the upper and lower layers meets the conditions. If it does, stop the iteration and output the calculation results; otherwise, continue to execute S7.5.

[0220] S7.5: Let the iteration number k = k + 1, update the penalty multiplier, and return to S7.2 to continue the iteration.

[0221] During the solution process, consistency constraints are added to the objective function in the form of penalty functions. Consistency is then achieved through iterative solutions by updating the penalty multipliers.

[0222] S8: The microgrid control center generates an operation plan after calculation and sends it to the power grid control center. The power grid control center performs power flow calculation and safety verification calculation to check whether it meets the various requirements and restrictions of power grid operation. If it meets the requirements, it proceeds to step S7; if it does not meet the requirements, it returns to step S5 to recalculate.

[0223] S9: If the calculation results meet the requirements and limitations of the power grid, each member of the microgrid will execute the generated operation plan.

[0224] S10: After each member of the micro-energy network executes the operation plan, it records the execution process and measures the consumption of various types of energy.

[0225] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0226] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0227] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0228] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0229] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0230] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A hierarchical optimization control method for microgrid groups based on scenario construction, characterized in that, Includes the following steps: S1: Establish a physical framework model of the regional power grid multi-micro energy network system, perform mathematical modeling for each type of energy, establish the functional relationships of energy, power, initial conditions, and boundary conditions for each energy form, and derive the energy conversion methods and relationships between each energy form; the physical framework model includes various energy supply devices and energy storage devices; S2: Analyze the uncertainties of wind power, photovoltaic power and load in the microgrid. The uncertainties include the uncertainty of renewable energy output and the uncertainty of load power. Establish a probability model for the uncertainties, and use a random variable probability model to establish the probability density functions of three quantities: wind speed, solar irradiance and load. S3: Establish a Copula function to connect random variables between wind power and photovoltaic power generation. The Copula function is essentially a function that connects the joint cumulative distribution function and the marginal cumulative distribution function of random variables. The Copula function can be used to consider the temporal and cross-correlation properties of random variables. Establish specific models of five Copula functions: normal Copula, t-Copula, Gumbel-Copula, Clayton-Copula, and Frank-Copula, and solve for their specific parameters. S4: Using the Euclidean distance method, select the Copula function that best describes the correlation between wind power and photovoltaic random variables; first, simulate and sample the distribution values ​​according to the five Copula functions, and then compare them with the actual collected field data values ​​to calculate the Euclidean distance of each; then compare the Euclidean distance of the five types of Copula functions with the Euclidean distance of the actual sampled data, and the one with the smallest Euclidean distance is the optimal Copula function model; S5: Generate basic scenarios. After determining the optimal Copula function, sample and aggregate based on the joint probability distribution of the optimal Copula function to generate typical application scenarios. S6: Establish a two-layer coordinated optimization model for regional micro-energy networks, dividing the optimization of clean energy consumption by regional micro-energy networks into two-layer optimization models; the optimization objective of the upper-layer model is to minimize the daily operating cost of the micro-energy network group within the region, while the optimization objective of the lower-layer model is to minimize the daily operating cost within a single micro-energy network. S7: A hierarchical optimization method based on objective cascade analysis is used to solve the problem. During the solution process, consistency constraints are added to the objective function as penalty functions. Consistency is then achieved through iterative solutions by updating the penalty multipliers. Specifically, considering that different operating entities formulate scheduling strategies at the upper and lower layers, and that the scheduling results influence each other, a hierarchical optimization method based on objective cascade analysis is used to solve the problem. During the solution process, consistency constraints are added to the objective function as penalty functions. Consistency is then achieved through iterative solutions by updating the penalty multipliers. The objective function with the added Lagrange form penalty function is shown below; Objective function of the upper-level micro-energy network cluster: (30)； The objective function of the nth microgrid in the upper layer is: (31) ； In the formula, , The penalty multipliers at time t are respectively, with symbols... For Hadama accumulation, Represents the square of the norm. Scaling factor , These are the planned interactive power of the nth microgrid and the optimized interactive power of the nth microgrid transmitted to the upper system, respectively. Let n be the scaling factor for the nth microgrid. , These represent the planned interactive power of the nth microgrid and the interactive power transferred to the nth microgrid after optimization by the upper-level model, respectively. This indicates that the upper-level absorption optimization aims to minimize the daily operating cost of the micro energy grid cluster within the region. The cost function includes energy interaction cost, energy storage cost, equipment operating cost, controllable load adjustment cost, and wind and solar forecast error cost. This indicates that the optimization of lower-level energy consumption aims to minimize the daily operating cost within a single microgrid. The cost function includes energy purchase cost, equipment operating cost, wind and solar curtailment penalty cost, and wind and solar forecast error cost. S8: The microgrid control center generates an operation plan after calculation and sends it to the power grid control center. The power grid control center performs power flow calculation and safety verification calculation to check whether it meets the various requirements and restrictions of power grid operation. If it meets the requirements, proceed to step S7; otherwise, return to step S5 to recalculate. S9: If the calculation results meet the requirements and limitations of the power grid, each member of the microgrid will execute the generated operation plan; S10: After each member of the micro-energy network executes the operation plan, it records the execution process and measures the consumption of various types of energy.

2. The hierarchical optimization control method for microgrid groups based on scenario construction according to claim 1, characterized in that, The micro-energy network described in step S1 integrates energy interconnection, conversion, coupling and storage functions. In the energy transmission, storage and distribution links, it provides energy to energy users or conducts energy transactions with the upper-level energy network at a certain price. The established regional power grid multi-micro-energy network system covers a variety of energy supply, energy storage and energy conversion equipment. In addition to wind power generation and photovoltaic distributed energy, the energy supply equipment also includes geothermal energy. The energy conversion equipment includes cogeneration systems, ground source heat pumps, electric chillers, and also incorporates batteries, gas storage tanks and thermal and cold storage energy storage devices. For example, energy storage models: （1）； In the formula, , This refers to the energy stored in the energy storage device at the current moment and the energy stored in the past moment. , Storage power and storage efficiency; , Release power, release efficiency; Such as electric refrigeration units: (2) ； In the formula, The magnitude of the cooling output generated by the electric chiller. The electrical power consumed by the electric chiller. The conversion efficiency of the electric chiller; In economic analysis, all forms of energy are categorized into the cost-optimal model in terms of total price. In power grid operation and dispatch, their relationship with power output is obtained and then incorporated into the constraints of the optimization model.

3. The hierarchical optimization control method for microgrid groups based on scenario construction according to claim 1, characterized in that, The probability density function mentioned in step S2 is as follows: For wind power, the output power of the wind turbine is closely related to the wind speed. The Weibull two-parameter distribution curve is used to describe wind speed; it is a unimodal curve with two parameters, and the wind speed probability density function can be expressed as: （3）； In the formula, v is the wind speed, and k and c represent two important parameters in the wind speed distribution, namely the shape parameter and the scale parameter, respectively. Its value can be calculated or estimated using historical wind speed data; For photovoltaics, the output power is closely related to the light intensity. Using a Beta distribution to describe the distribution of light intensity, the probability density function of light intensity can be expressed as: （4）； In the formula, Represents the gamma function; , Represents shape parameters; I represents light intensity, and Imax represents the maximum light intensity. For loads, the normal distribution can represent the load distribution over a period of time relatively well. Its distribution parameters are affected by regional climate, and the probability density function of the load distribution can be expressed as: （5）； In the formula, Indicates load power; This represents the average load power. This represents the standard deviation of load power.

4. The hierarchical optimization control method for microgrid groups based on scenario construction according to claim 1, characterized in that, Step S3 specifically involves: According to Sklar's theorem, assuming a random variable... The joint distribution function is The edge distribution is Then there exists a Copula function. , so that; pay attention The domain is ; (6) ； like If continuous, then the Copula function Uniquely certain; conversely, If it is a univariate distribution function, then That is, the corresponding Copula function; Five Copula functions are introduced: normal Copula, t-Copula, Gumbel-Copula, Clayton-Copula, and Frank-Copula. Normal Copula function: (7) ； This represents the correlation coefficient, while the upper limit of integration of the Copula function is... It represents the standard normal inverse function based on the correlation coefficient; t-Copula function: (8) ； The coefficient k represents the degrees of freedom, and similarly This represents the correlation coefficient, while the upper limit of integration... It represents the inverse function of the standard t-distribution with k degrees of freedom; Frank-Copula function: (9) ； In the formula, Represents the correlation coefficient, if A value approaching 0 indicates that the two random variables are uncorrelated. A value greater than 0 indicates a positive correlation. A value less than 0 indicates a negative correlation; note ≠0; Gumbel-Copula function: (10) ； In this Copula function , This indicates that two random variables are independent; if If the value approaches infinity, it indicates that the two random variables are completely dependent. The upper tail correlation coefficient of Gumbel-Copula is The lower tail correlation coefficient is 0, so the Gumbel-Copula is effective in describing random variables with lower tail correlation. The Gumbel-Copula is an asymmetric distribution. Clayton-Copula: (11) ； Clayton-Copula ,at the same time The lower tail correlation coefficient of the Clayton-Copula is: The upper tail correlation coefficient is 0, so the Clayton-Copula is suitable for describing random variables with upper tail correlation. The Clayton-Copula is also an asymmetric distribution. Based on the probability density function determined in step S2, the marginal cumulative distribution functions of wind power and photovoltaic power output are determined respectively. Then, the random variables to be solved are transformed into uniformly distributed random numbers on [0,1] using the marginal probability cumulative distribution function transformation. Subsequently, various Copula function forms are substituted, and the unknown parameters of the Copula function are solved using the two-step maximum likelihood method.

5. The hierarchical optimization control method for microgrid groups based on scenario construction according to claim 1, characterized in that, Step S4 specifically involves: The Euclidean distance discriminant method compares the Euclidean distances of various Copula functions with empirical Copula functions generated from sample data. The smaller the Euclidean distance, the better the fit of the Copula function. Suppose two vectors , Here, Euclidean distance represents the straight-line distance between two points, and its expression is as follows: (12) ； (13) ； These are the fit metrics: Euclidean distance d and maximum distance di. max The smaller the two values ​​above, the better the model matches the original actual data distribution; maximum distance d max The maximum distance between the empirical Copula distribution value and the Copula distribution value; By comparing the Euclidean distance d and the maximum distance d of various Copula functions max Determine the optimal Copula function.

6. The hierarchical optimization control method for microgrid groups based on scenario construction according to claim 1, characterized in that, Step S5 specifically involves: Based on the optimal Copula function, in order to obtain representative operating scenarios, it is necessary to reduce the number of sampled scenarios by using the K-means clustering method, the expression of which is as follows: (14) ； In the formula, For clustering The intra-class average distance, The distance between the two cluster centers. The number of clusters is determined by the analysis of actual field data and a large number of calculation results, which identify two typical scenarios: winter scenario and summer scenario.

7. The hierarchical optimization control method for microgrid groups based on scenario construction according to claim 1, characterized in that, Step S6 specifically involves: Based on the uncertainty analysis of clean energy output and load power within the microgrid, the absorption of clean energy by the regional microgrid is divided into two layers: the upper layer is the coordinated optimization of the microgrid group, and the lower layer is the absorption optimization within a single microgrid. The upper-level absorption optimization aims to minimize the daily operating cost of the micro energy network in the region. The cost function includes energy interaction cost, energy storage cost, equipment operation cost, controllable load adjustment cost, and wind and solar forecast error cost. （15）； （16）； （17）； （18）； （19）； （20）； In the formula, N represents the number of microgrids, and m represents the type of energy. ; , , , , These are the energy interaction costs between microgrids, energy storage costs, equipment operating costs, controllable load regulation costs, and wind and solar forecasting error costs; It is the transaction cost of the m-th energy source; , These are the input and output quantities of the m-th energy source, respectively. It is the operation and maintenance coefficient of the m-th type of energy storage equipment; , These are the charging and discharging power of the m-th energy source; , , These are the operating power of the CCHP system, ground source heat pump, and electric chiller equipment, respectively. , , These are the operating cost coefficients for CCHP systems, ground source heat pumps, and electric chiller equipment, respectively. It is the cost coefficient for the adjustment and compensation of the m-th controllable load; The m-th type of controllable load participates in the regulation; This indicates the prediction error for wind and solar power generation. To avoid overestimating the cost factor resulting from overestimating power generation capacity, The cost factor resulting from off-peak power generation; 's' represents the scene marker; winter and summer will be discussed separately. Constraints: The constraints for energy storage and equipment are as follows: （21）； In the formula, , These are the upper and lower limits of energy storage for the m-th energy source, respectively. , These are the upper and lower limits of the energy release for the m-th energy source, respectively. , The upper and lower limits of the output of the i-th type of equipment are not specified. Controllable load constraints （22）； In the formula, , These are the minimum and maximum load power of the m-th energy source, respectively; The optimization of lower-level energy consumption aims to minimize the daily operating cost within a single microgrid. The cost function includes energy purchase cost, equipment operating cost, wind and solar curtailment penalty cost, and wind and solar forecast error cost. （23）； （24）； （25）； （26）； In the formula, This represents the energy purchase cost in the s-th scenario of the microgrid. It is the equipment operating cost; It is the cost of penalties for abandoning wind and solar power; , These are time-of-use electricity pricing and gas pricing; , These are the electricity and gas purchases, respectively. , , , , , , These are the operating power of CCHP systems, ground source heat pumps, electric chillers, energy storage, gas storage, thermal storage, and cold storage equipment; , , , , , , These are the operating cost coefficients for CCHP systems, ground source heat pumps, electric chillers, energy storage, gas storage, thermal storage, and cold storage equipment. CP is the penalty coefficient for abandoning wind and light. , These are respectively the amount of wind and solar power curtailment; The constraints are: Constraints on wind, solar, and geothermal power output （27）； In the formula, , , These are the maximum outputs of wind power, solar power, and geothermal power, respectively. Unit output upper and lower limit constraints （28）； In the formula, , These are the upper and lower limits of CCHP's output operation, respectively; , These are the upper and lower limits of the output operation of the ground source heat pump; , These are the upper and lower limits of the electric chiller's output operation; , These are the upper and lower limits of the battery's energy storage capacity; , These are the upper and lower limits of gas storage in the gas storage tank; , These are the upper and lower limits of thermal storage for the thermal storage device; , These are the upper and lower limits of the cold storage device; Power balance constraints （29）； In the formula, , , , These are the electrical-thermal-cooling conversion coefficients and the electro-cooling coefficient of CCHP, respectively. , These are discharge power and energy storage power, respectively. , These are the venting and storage capacities, respectively. , These are the heat release and heat storage power, respectively. These are cooling output power and cooling storage power, respectively.

8. The hierarchical optimization control method for microgrid groups based on scenario construction according to claim 1, characterized in that, The solution steps for step S7 are as follows: S7.1: Initialize system parameters, set scaling factor, initial penalty multiplier, and set iteration count k=1; S7.2: If k=1, solve the upper-level micro-energy network group optimization problem and pass the obtained interactive power to the lower level; S7.3: Solve the optimization problem within the lower-level microgrid; due to the existence of the penalty function, the optimized interactive power of the lower level will be close to the value passed from the upper level; S7.4: Check if the interaction power between the upper and lower layers meets the conditions. If it does, stop the iteration and output the calculation results; otherwise, continue to execute S7.

5. S7.5: Let the iteration number k = k + 1, update the penalty multiplier, and return to S7.2 to continue the iteration.

9. A scenario-based hierarchical optimization control system for a microgrid group, as described in any one of claims 1 to 8, characterized in that, Includes the following modules: The first modeling module is used to establish the physical framework model of the regional power grid multi-micro energy network system; the physical framework model includes various energy supply devices and energy storage devices. The first analysis module is used to analyze the uncertainties of wind power, photovoltaics, and load in microgrids; it establishes a probability model for the uncertainties, using a random variable probability model to establish the probability density functions of three quantities: wind speed, solar irradiance, and load. The first calculation module is used to establish the Copula function that connects random variables between wind power and photovoltaic power generation. It establishes specific models of five Copula functions: normal Copula, t-Copula, Gumbel-Copula, Clayton-Copula, and Frank-Copula, and solves for their specific parameters. The second calculation module is used to select the most suitable Copula function to describe the correlation of random variables in wind power and photovoltaics using the Euclidean distance method. First, it simulates and samples the distribution values ​​according to the five Copula functions, and then compares them with the actual collected field data values ​​to calculate the Euclidean distance of each. Then, it compares the Euclidean distance of the five types of Copula functions with the Euclidean distance of the actual sampled data, and the one with the smallest Euclidean distance is the optimal Copula function model. Scene generation module: Used to generate basic scenarios. After determining the optimal Copula function, it samples and aggregates the joint probability distribution of the optimal Copula function to generate typical application scenarios. The second modeling module is used to establish a two-layer coordinated optimization model for regional micro-energy networks, which divides the optimization of clean energy consumption in regional micro-energy networks into two-layer optimization models. The optimization objective of the upper-level model is to minimize the daily operating cost of the microgrid cluster within the region, while the optimization objective of the lower-level model is to minimize the daily operating cost within a single microgrid. The third calculation module uses a hierarchical optimization method based on objective cascade analysis to solve the problem. During the solution process, consistency constraints are added to the objective function in the form of penalty functions. The solution is obtained by iteratively updating the penalty multipliers, and finally, consistency is achieved. First discrimination module: The microgrid control center generates an operation plan after calculation and sends it to the power grid control center. The power grid control center performs power flow calculation and safety verification calculation to determine whether it meets the various requirements and restrictions of power grid operation. If it meets the requirements, it proceeds to step S7. If it does not meet the requirements, it returns to step S5 to calculate again. Second discrimination module: If the calculation results meet the requirements and limitations of the power grid, each member of the microgrid executes the generated operation plan; The sixth calculation module is used to record the execution process and measure the consumption of various types of energy after each member of the micro energy network executes the operation plan.

10. A scenario-based hierarchical optimization control system for microgrid clusters according to claim 9, characterized in that, In the first modeling module, establishing the physical framework model of the regional power grid multi-micro energy network system refers to performing mathematical modeling for each type of energy, establishing the functional relationships of energy, power, initial conditions, and boundary conditions for each energy form, and deriving the energy conversion methods and relationships between each energy form.

11. A scenario-based hierarchical optimization control system for microgrid clusters according to claim 9, characterized in that, The uncertainties mentioned include uncertainties in renewable energy output and uncertainties in load power.

12. A computer storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a scenario-based hierarchical optimization control method for microgrid groups as described in any one of claims 1-8.

13. A computer 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 computer program, it implements a scenario-based hierarchical optimization control method for micro-energy grid clusters as described in any one of claims 1-8.

Images