Virtual power plant double-layer distribution robust scheduling method considering shared energy storage

Abstract

A virtual power plant double-layer distributed robust scheduling method considering shared energy storage belongs to the field of power system optimization scheduling, and comprises the following steps: step 1, determining a global scheduling plan based on R-vine Copula and distributed robust optimization upper-layer global robust scheduling; and step 2, executing and decomposing the global scheduling plan based on lower-layer distributed cooperative scheduling of an alternating direction multiplier method (ADMM). By means of the method, it is ensured that all participants including independent shared energy storage operators can achieve profit, a feasible business model is constructed for multiple subjects to collaboratively participate in power grid service, and effective economic incentive is provided.

Description

Technical Field

[0001] This invention relates to the field of power system optimization scheduling, and in particular to a two-layer distributed bar scheduling method for virtual power plants that considers shared energy storage. Background Technology

[0002] With the increasing penetration of renewable energy sources such as wind and solar power into the power grid, their inherent volatility and uncertainty pose significant challenges to the safe and stable operation of the power system. Virtual power plants (VPPs), as an effective technological approach integrating distributed energy resources, controllable loads, and energy storage systems, have become crucial for enhancing grid flexibility. Meanwhile, shared energy storage (SES), as an emerging business model, can provide energy time-shifting services to multiple entities, further enhancing system regulation capabilities.

[0003] However, the coordinated scheduling of VPP and SES faces two major challenges. First, the participating entities (such as VPP operators, SES operators, and users) are independent economic entities, and due to concerns about trade secrets, they are unwilling to share their internal privacy information such as costs and pricing, making traditional centralized optimization methods difficult to implement. Second, resources such as wind, solar, and load exhibit complex nonlinear correlations and tail dependencies under extreme events. Traditional stochastic programming or robust optimization methods either assume that uncertain variables are independent or use simple probability distribution models, making it difficult to accurately characterize these complex coupling characteristics. This can lead to scheduling schemes that are either too aggressive or too conservative in practical applications, failing to balance economy and security.

[0004] In existing research, while distributed algorithms such as the Alternating Direction Multiplier Method (ADMM) have been used to solve multi-agent cooperative optimization problems to protect privacy, most are based on deterministic models and fail to fully consider the impact of uncertainty. Traditional robust optimization methods, although capable of handling uncertainty, often lead to overly conservative results due to their "worst-case" assumptions. Distributed robust optimization (DRO) can model the uncertainty of the probability distribution itself under data-driven conditions, but its effectiveness is highly dependent on the accurate characterization of the joint distribution of uncertainty. Therefore, there is an urgent need for a cooperative scheduling method for VPP and SES that can accurately model multivariate uncertainty, protect the privacy of all parties, and achieve a balance between economy and robustness. Summary of the Invention

[0005] This invention aims to overcome the shortcomings of existing technologies and provide a two-layer distributed bar scheduling method for virtual power plants that considers shared energy storage. This method can accurately characterize the complex correlation of source-load uncertainty, protect the commercial privacy of all participants, and achieve a balance between the economy and robustness of system operation.

[0006] This invention is achieved through the following technical solution: a two-layer distributed bar scheduling method for virtual power plants considering shared energy storage, comprising the following steps: Step 1: Use upper-level global robust scheduling based on R-vine Copula and distributed robust optimization to determine the global scheduling plan.

[0007] Step 1.1: Construct a source-load joint probability distribution model: Collect historical data of wind power, photovoltaic power and load, and use R-Teng Copula theory to construct a high-dimensional joint probability distribution model that can capture the nonlinearity, asymmetry and tail correlation characteristics between variables; To describe the complex dependency among wind power, solar power, and load, we first introduce Copula theory. According to Sklar's theorem, any multidimensional joint distribution function is expressed as a combination of its marginal distribution functions and a Copula function. Let the three random variables—wind power, solar power, and load—be denoted as follows: Joint probability density function Decomposed into:

[0008] in, For the first Marginal probability density function of each variable; For the first Marginal cumulative distribution function of variables The result of the conversion is A variable that is uniformly distributed over an interval; It is the Copula density function, used to describe the dependency structure between variables; To capture asymmetry and tail correlation, R-vine Copula is employed. R-vine Copula decomposes the high-dimensional Copula density function into a product of a series of paired two-dimensional Copulas through a multi-layered tree structure. For a three-dimensional variable, its R-vine Copula density function decomposes as follows:

[0009] in, and It is a pairing copula of the first-level tree structure, which describes the pairs of variables. and Direct dependencies between them; It is a pairing copula in a second-level tree structure, describing the variables Given conditions, variables and The conditional dependencies between them; and The tree structure is determined using a data-driven approach, with the conditional distribution function as the basis, and the optimal family of Copula functions and parameters are selected for each pair.

[0010] Step 1.2: Generate and reduce typical scenarios: Based on the constructed R-vine Copula model, Monte Carlo sampling is performed to generate random scenarios that can reflect the true characteristics of the source load. Then, the K-means clustering algorithm is used to reduce the massive number of scenarios into a group of representative typical scenarios and initial probabilities. Based on the constructed R-vine Copula model A second Monte Carlo sampling is performed to generate a combined wind-solar-load scene. To reduce the computational complexity of subsequent optimization, the K-means clustering algorithm is used to cluster the generated scenes. The scenario was reduced to A representative typical scenario ,in For scene indexing, Indexed by time period, for each typical scenario Represents a cluster center with its initial probability It is determined by the proportion of the number of samples included in the cluster to the total number of samples.

[0011] Step 1.3: Construct and solve a two-stage robust DRO model: Establish a DRO model with the goal of minimizing the total upper-level scheduling cost. This model searches for the worst-case scenario probability distribution within a probability fuzzy set defined by the 1-norm and the ∞-norm, and minimizes the expected operating cost based on this distribution to obtain a global scheduling plan that balances economy and robustness. The global scheduling plan includes the globally optimal energy interaction power between the virtual power plant and the shared energy storage. Step 1.3.1, Model Construction: A two-stage DRO model is established, in the following form:

[0012] in, The first-stage decision variable represents the scheduling decision determined previously and independent of the specific scenario; the feasible region is... ; For the cost of the first phase, For the scene The true probability, all The constructed probability vector Constrained in a probabilistic fuzzy set Inside; The cost function for the second stage represents the decision made in the first stage. The scene is already set. When this occurs, the minimum operating cost generated through optimal real-time scheduling is expressed as:

[0013] in, These are the decision variables for the second stage, also known as follow-up variables, and are used in the scenario. Real-time adjustment of decisions; This represents the cost coefficient vector for the second stage. The feasible region for the second-stage decision is defined by operational constraints and includes: Power balance constraints:

[0014] in, For conventional generator sets exist Efforts during a specific time period; and These are the actual grid-connected power of wind power and solar power, respectively; and These are the discharge and charging power of the shared energy storage, respectively. and These represent the power purchased from and sold to the power grid, respectively. for Actual load during the time period; Energy storage operation constraints:

[0015] in, for State of charge of energy storage at the end of the time period; and These are the charging and discharging efficiencies, respectively. For the scheduling time interval, the state of charge and charging / discharging power of energy storage must meet the upper and lower limits of capacity and power. Other operational constraints include the upper and lower limits of output and ramping constraints of conventional generator sets, the power exchange limit with the grid, the constraint that the output of renewable energy should not exceed its predicted value, and the adjustment constraints of demand response load. probabilistic fuzzy sets It depicts the true probability distribution Compared with empirical probability distribution The range of deviations between them is constrained by both the 1-norm and the ∞-norm of the fuzzy set.

[0016] in, and It is a deviation radius determined based on a preset confidence level, used to limit the fluctuation range of the probability distribution.

[0017] Step 1.3.2, Model Solving: The column and constraint generation C&CG algorithm is used to iteratively solve the above two-stage sub-Bruker optimization problem, which is decomposed into a main problem M and a subproblem to be solved alternately until the upper and lower bounds converge.

[0018] Step 2: Lower-level distributed cooperative scheduling based on the Alternating Direction Multiplier Method (ADMM) is used to execute and decompose the global scheduling plan. Step 2.1: Construct local optimization models for each participant: Establish a local optimization model for each independent aggregation unit of the operator with the goal of optimizing its own operating cost, including its internal equipment constraints and energy interaction variables with other units; The entire system is decomposed into multiple independent aggregation units, each unit The goal is to minimize its local cost function The units are interconnected through energy exchange, and are denoted as units. With unit The interaction power between them is And it needs to satisfy consistency constraints: .

[0019] Step 2.2: Establish the augmented Lagrangian function and solve iteratively: Using the global optimal energy interaction power determined in Step 1.3 as the consistency constraint for the lower-level distributed optimization, construct the augmented Lagrangian function of the system, and solve iteratively using the ADMM algorithm to decompose the global scheduling plan into the optimal operating strategy of each unit until the algorithm converges.

[0020] To solve this distributed optimization problem, an augmented Lagrangian function is constructed, and the ADMM algorithm is used iteratively. The solution is obtained at the 6th... q+1 In the next iteration: Local optimization steps, namely Update: Each aggregation unit Given the interaction power of other units and the current Lagrange multipliers, In this case, solve the local optimization problem to update its internal decision variables. and planned interaction power The objective function for optimization is:

[0021] in, This is the coefficient of the second-order penalty term. In this step, each unit only uses its own private information to make decisions.

[0022] Price update steps, namely Update: After all units have completed local optimization, the coordination center collects the planned interaction power of each unit and updates it according to the degree of violation of consistency constraints, i.e., the residual. To update the Lagrange multipliers :

[0023] The updated multipliers will be sent to each unit as the price signal for the next iteration; The above steps are repeated until both the primal residual and dual residual are less than the preset threshold, indicating that the algorithm has converged and all parties have reached a consensus on the energy exchange plan.

[0024] Compared with existing technologies, the beneficial effects of this invention are as follows: By more accurately characterizing the complex nonlinearity and tail correlation between wind power, photovoltaics, and loads, this invention provides more reliable decision support for dealing with extreme uncertainty scenarios, significantly improving the balance between robustness and economy of virtual power plant dispatching schemes. While ensuring the safe operation of the system, this invention effectively reduces the total operating cost of the system. Compared with traditional uncoordinated robust models, it can reduce the total operating cost of virtual power plants by 12.6% and reduce the renewable energy curtailment rate to 2.3%, thereby achieving a renewable energy consumption level of over 95%.

[0025] Furthermore, in achieving collaborative optimization among independent market players, this invention eliminates the need for participants to share commercially sensitive data such as internal cost functions and operational constraints, thus effectively protecting the commercial privacy of all parties. By establishing a mutually beneficial collaborative mechanism, this invention ensures that all participants, including independent shared energy storage operators, can achieve profitability, constructing a feasible business model for multi-stakeholder collaborative participation in grid services and providing effective economic incentives. Attached Figure Description

[0026] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. The scope of protection of the present invention is not limited to the following description.

[0027] Figure 1 This is a schematic diagram of the two-layer distributed bar scheduling framework proposed in this invention.

[0028] Figure 2 A schematic diagram of the tree-like structure for the three-dimensional joint distribution of wind, light, and load in the R-vine Copula model.

[0029] Figure 3 This diagram illustrates the convergence process of the ADMM algorithm in the lower-level distributed scheduling, showing the iterative curves of cost / profit for each aggregation unit.

[0030] Figure 4This is an optimized scheduling result of virtual power plants and shared energy storage obtained through the method of this invention under a typical severe scenario. Detailed Implementation

[0031] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. The scope of protection of the present invention is not limited to the following description. The present invention proposes a two-layer distributed bar scheduling method for virtual power plants considering shared energy storage, the overall technical process of which is as follows: Figure 1 As shown, it is divided into two levels: upper-level global robust optimization and lower-level distributed collaborative optimization.

[0032] Phase 1: Upper-level global robustness optimization

[0033] The goal of this phase is to develop a global day-ahead scheduling plan that balances economy and robustness at the system level, taking into account the worst-case source-load uncertainty scenarios.

[0034] Step 1: Uncertainty Modeling and Scene Generation

[0035] 1.1 Constructing the joint probability distribution

[0036] To accurately describe the complex dependencies between wind power, solar power, and load, we first introduce Copula theory. According to Sklar's theorem, any multidimensional joint distribution function can be expressed as a combination of its marginal distribution functions and a Copula function. Let the three random variables—wind power, solar power, and load—be denoted as... Its joint probability density function It can be broken down into:

[0037] in, For the first Marginal probability density function of each variable; For the first Marginal cumulative distribution function (CDF) of each variable The result of the conversion is A variable that is uniformly distributed over an interval; It is the Copula density function, used to describe the dependency structure between variables.

[0038] To flexibly capture asymmetry and tail correlation, this invention employs R-vine Copula. R-vine Copula utilizes a multi-layered tree-like structure (such as...) Figure 2 (As shown) the high-dimensional Copula density function is decomposed into a product of a series of paired two-dimensional Copulas. For a three-dimensional variable, its R-copula density function can be decomposed as:

[0039] in, and It is a pairing copula of the first-level tree structure, which describes the pairs of variables. and Direct dependencies between them; It is a pairing copula in a second-level tree structure, describing the variables Given conditions, variables and The conditional dependencies between them; and The conditional distribution function is used. The tree structure is determined through a data-driven approach, and the optimal family of Copula functions (such as Frank, Clayton, etc.) and parameters are selected for each pair.

[0040] 1.2 Scene Generation and Reduction

[0041] Based on the constructed R-vine Copula model Next (for example) Monte Carlo sampling was used to generate a large number of high-fidelity wind-solar-load joint scenes. To reduce the computational complexity of subsequent optimization, the K-means clustering algorithm was used to analyze the generated scenes. The scenario was reduced to One (e.g.) Representative typical scenarios ,in For scene indexing, Indexed by time period. Each typical scenario Represents a cluster center with its initial probability It is determined by the proportion of the number of samples included in the cluster to the total number of samples.

[0042] Step 2: Two-stage Distributed Bar Optimization (DRO) Model

[0043] 2.1 Model Construction

[0044] A two-stage DRO model is established, and its compact form is as follows:

[0045] in, The first-stage decision variable represents the scheduling decision determined previously and independent of the specific scenario, such as the start-up and shutdown status of generator units. Its feasible region is... ; This refers to the cost of the first phase. For the scene The true probability, all The constructed probability vector Constrained in a probabilistic fuzzy set Inside.

[0046] The cost function for the second stage represents the decision made in the first stage. The scene is already set. When this occurs, the minimum operating cost generated through optimal real-time scheduling is expressed as: .

[0047] in, These are the decision variables for the second stage, also known as follow-up variables, and are used in the scenario. Real-time adjustment decisions, such as the actual grid-connected power of renewable energy, the charging and discharging power of energy storage, and the power exchanged with the grid; This represents the cost coefficient vector for the second stage. The feasible region for the second-stage decision is defined by a series of operational constraints, mainly including: Power balance constraints:

[0048] in, For conventional generator sets exist Efforts during a specific time period; and These are the actual grid-connected power of wind power and solar power, respectively; and These are the discharge and charging power of the shared energy storage, respectively. and These represent the power purchased from and sold to the power grid, respectively. for Actual load during the time period.

[0049] Energy storage operation constraints:

[0050] in, for State of charge of energy storage at the end of the time period; and These are the charging and discharging efficiencies, respectively. This is the scheduling time interval. Simultaneously, the state of charge and charging / discharging power of the energy storage must meet its capacity and power upper and lower limits.

[0051] Other operational constraints include upper and lower limits of output and ramping constraints for conventional generator units, power exchange limits with the grid, constraints that renewable energy output should not exceed its forecast value, and regulation constraints for demand response loads.

[0052] probabilistic fuzzy sets The definition of probability distribution is key to this invention; it characterizes the true probability distribution. Compared with empirical (initial) probability distribution The range of deviations between them. This fuzzy set is constrained by both the 1-norm and the ∞-norm:

[0053] in, and The deviation radius is determined based on a preset confidence level to limit the fluctuation range of the probability distribution, thereby ensuring the robustness of the scheduling scheme to probability distribution errors.

[0054] 2.2 Model Solving

[0055] The column and constraint generation (C&CG) algorithm is used to iteratively solve the above three-level (min-max-min) optimization problem. It is decomposed into a master problem and a subproblem and solved alternately until the upper and lower bounds converge.

[0056] Phase Two: Lower-Level Distributed Optimization

[0057] The upper-level DRO model provides a global day-ahead scheduling plan, while the lower-level model executes the plan in a privacy-preserving distributed manner using the ADMM algorithm.

[0058] Step 3: ADMM Distributed Collaboration

[0059] 3.1 Problem Decomposition and Consistency Constraints

[0060] The entire system is decomposed into multiple independent aggregation units (participants), such as VPP, SES, etc. Each unit k The goal is to minimize its local cost function The units are interconnected through energy exchange, denoted as units. With unit The interaction power between them is To ensure overall system power balance, consistency constraints must be met: .

[0061] 3.2 ADMM Iteration Process

[0062] To solve this distributed optimization problem, an augmented Lagrangian function is constructed, and the ADMM algorithm is used for iterative processing. In the... q+1 In this iteration, the following steps are performed: Local optimization steps ( -Updated): Each aggregation unit Given the interaction power of other units and the current Lagrange multipliers (i.e., the internal transaction price). In this case, solve a local optimization problem to update its internal decision variables. and planned interaction power The objective function for optimization is:

[0063] in, This is the coefficient of the secondary penalty term. In this step, each unit only uses its own private information to make decisions.

[0064] Price update steps ( -Updated): After all units have completed local optimization, the coordination center collects the planned interaction power of each unit and determines the power based on the degree of violation of consistency constraints (i.e., residuals). To update the Lagrange multipliers :

[0065] The updated multiplier will be sent to each unit as the price signal for the next iteration.

[0066] The above steps are repeated until both the primal residual and dual residual are less than a preset threshold, indicating that the algorithm has converged and all parties have reached a consensus on the energy exchange plan. In this way, the present invention achieves distributed collaborative execution of a robust upper-layer scheduling plan while protecting the commercial privacy of all participating parties.

[0067] Simulation Results and Analysis

[0068] To verify the effectiveness of the proposed method, a simulation model was built using MATLAB R2021a software and the Gurobi solver. The simulation system includes a virtual power plant (VPP) operator and an independent shared energy storage (SES) operator. The VPP aggregates a 20MW wind farm, a 15MW photovoltaic power plant, a 10MW controllable distributed generator, and various loads. The SES operator manages a 5MW / 20MWh battery energy storage system. The scheduling cycle is 24 hours with a resolution of 1 hour. The system's interaction with the external power grid follows a peak-valley time-of-use pricing mechanism.

[0069] like Figure 2 As shown, this invention first uses the R-Teng Copula model to analyze historical data, obtaining a tree-like structure of dependencies between wind power (node ​​1), photovoltaics (node ​​2), and load (node ​​3). This structure can accurately capture the complex nonlinearities and tail correlations between variables, and compared with traditional models, it can more accurately assess the risk of extreme events, providing high-fidelity scenario input for subsequent upper-level sub-Bruker optimization.

[0070] like Figure 3As shown in the figure, the convergence process of the ADMM algorithm in the lower-level distributed scheduling of this invention is illustrated. The figure shows that the operating costs of the VPP operator and the profits of the SES operator both converge rapidly and reach stability within 30 iterations. This result demonstrates that the distributed cooperation mechanism adopted in this invention has high computational efficiency and can, without exchanging their internal privacy information, guide multiple independent entities to quickly reach a globally near-optimal and mutually beneficial scheduling protocol through price signals.

[0071] like Figure 4 As shown in the figure, the system's final optimized scheduling result is illustrated under a typical worst-case scenario identified by the upper-level distributed bar optimization model. During periods of low load, low electricity prices, and high renewable energy output (e.g., 0:00-7:00), such as nighttime and midday, the shared energy storage charges to absorb surplus green electricity. During peak electricity demand periods in the morning and evening (e.g., 8:00-11:00 and 17:00-21:00), the shared energy storage discharges to meet peak load demand, thereby reducing the VPP's need to purchase high-priced electricity from the grid or start high-cost generator units. This charging and discharging behavior clearly demonstrates the synergistic and complementary relationship between the VPP and the SES achieved by the method of this invention. That is, the SES provides key flexibility support for the VPP, while also generating profits through energy time-shift arbitrage.

[0072] Simulation data comparison and analysis show that, compared with traditional robust optimization scheduling schemes that do not coordinate with SES, the method of this invention can reduce the total operating cost of VPP by 12.6%, while significantly reducing the renewable energy curtailment rate from 6.7% to 2.3%, achieving a renewable energy absorption rate of over 95%. This proves that the method proposed in this invention can perfectly balance the system's operational economy and robustness in the face of uncertainty, achieving significantly better results than existing technologies.

[0073] It is understood that the above specific description of the present invention is only for illustrating the present invention and is not limited to the technical solutions described in the embodiments of the present invention. Those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention to achieve the same technical effect; as long as the use needs are met, they are all within the protection scope of the present invention.

Claims

1. A two-layer distributed bar scheduling method for virtual power plants considering shared energy storage, characterized in that, Includes the following steps: Step 1: Determine the global scheduling plan based on upper-level global robust scheduling using R-vine Copula and distributed robust optimization: Step 1.1: Construct a source-load joint probability distribution model: Collect historical data of wind power, photovoltaic power and load, and use R-Teng Copula theory to construct a high-dimensional joint probability distribution model that can capture the nonlinearity, asymmetry and tail correlation characteristics between variables; Step 1.2: Generate and reduce typical scenarios: Based on the constructed R-vine Copula model, Monte Carlo sampling is performed to generate random scenarios that can reflect the true characteristics of the source load. Then, the K-means clustering algorithm is used to reduce the massive number of scenarios into a group of representative typical scenarios and initial probabilities. Step 1.3: Construct and solve a two-stage robust DRO model: Establish a DRO model with the goal of minimizing the total upper-level scheduling cost. This model searches for the worst-case scenario probability distribution within a probability fuzzy set defined by the 1-norm and the ∞-norm, and minimizes the expected operating cost based on this distribution to obtain a global scheduling plan that balances economy and robustness. The global scheduling plan includes the globally optimal energy interaction power between the virtual power plant and the shared energy storage. Step 2: Based on the alternating direction multiplier method (ADMM), the lower-level distributed cooperative scheduling executes and decomposes the global scheduling plan: Step 2.1: Construct local optimization models for each participant: Establish a local optimization model for each independent aggregation unit of the operator with the goal of optimizing its own operating cost, including its internal equipment constraints and energy interaction variables with other units; Step 2.2: Establish the augmented Lagrangian function and solve iteratively: Using the global optimal energy interaction power determined in Step 1.3 as the consistency constraint for the lower-level distributed optimization, construct the augmented Lagrangian function of the system, and solve iteratively using the ADMM algorithm to decompose the global scheduling plan into the optimal operating strategy of each unit until the algorithm converges.

2. The virtual power plant two-layer distributed bar scheduling method considering shared energy storage as described in claim 1, characterized in that: The specific method in step 1.1 is as follows: To describe the complex dependency among wind power, solar power, and load, we first introduce Copula theory. According to Sklar's theorem, any multidimensional joint distribution function is expressed as a combination of its marginal distribution functions and a Copula function. Let the three random variables—wind power, solar power, and load—be denoted as follows: Joint probability density function Decomposed into: in, For the first Marginal probability density function of each variable; For the first Marginal cumulative distribution function of variables The result of the conversion is A variable that is uniformly distributed over an interval; It is the Copula density function, used to describe the dependency structure between variables; To capture asymmetry and tail correlation, R-vine Copula is employed. R-vine Copula decomposes the high-dimensional Copula density function into a product of a series of paired two-dimensional Copulas through a multi-layered tree structure. For a three-dimensional variable, its R-vine Copula density function decomposes as follows: in, and It is a pairing copula of the first-level tree structure, which describes the pairs of variables. and Direct dependencies between them; It is a pairing copula in a second-level tree structure, describing the variables Given conditions, variables and The conditional dependencies between them; and The tree structure is determined using a data-driven approach, with the conditional distribution function as the basis, and the optimal family of Copula functions and parameters are selected for each pair.

3. The virtual power plant two-layer distributed bar scheduling method considering shared energy storage as described in claim 1, characterized in that: The specific method in step 1.2 is as follows: Based on the constructed R-vine Copula model A second Monte Carlo sampling is performed to generate a combined wind-solar-load scene. To reduce the computational complexity of subsequent optimization, the K-means clustering algorithm is used to cluster the generated scenes. The scenario was reduced to A representative typical scenario ,in For scene indexing, Indexed by time period, for each typical scenario Represents a cluster center with its initial probability It is determined by the proportion of the number of samples included in the cluster to the total number of samples.

4. The virtual power plant two-layer distributed bar scheduling method considering shared energy storage as described in claim 1, characterized in that: The specific method in step 1.3 is as follows: Step 1.3.1, Model Construction: A two-stage DRO model is established, in the following form: in, The first-stage decision variable represents the scheduling decision determined previously and independent of the specific scenario; the feasible region is... ; For the cost of the first phase, For the scene The true probability, all The constructed probability vector Constrained in a probabilistic fuzzy set Inside; The cost function for the second stage represents the decision made in the first stage. The scene is already set. When this occurs, the minimum operating cost generated through optimal real-time scheduling is expressed as: in, These are the decision variables for the second stage, also known as follow-up variables, and are used in the scenario. Real-time adjustment of decisions; This is the cost coefficient vector for the second stage; The feasible region for the second-stage decision is defined by operational constraints and includes: Power balance constraints: in, For conventional generator sets exist Efforts during a specific time period; and These are the actual grid-connected power of wind power and solar power, respectively; and These are the discharge and charging power of the shared energy storage, respectively. and These represent the power purchased from and sold to the power grid, respectively. for Actual load during the time period; Energy storage operation constraints: in, for State of charge of energy storage at the end of the period; and These are the charging and discharging efficiencies, respectively. For the scheduling time interval, the state of charge and charging / discharging power of energy storage must meet the upper and lower limits of capacity and power. Other operational constraints include the upper and lower limits of output and ramping constraints of conventional generator sets, the power exchange limit with the grid, the constraint that the output of renewable energy should not exceed its predicted value, and the adjustment constraints of demand response load. probabilistic fuzzy sets It depicts the true probability distribution Compared with empirical probability distribution The range of deviations between them is constrained by both the 1-norm and the ∞-norm of the fuzzy set. in, and It is a deviation radius determined according to a preset confidence level, used to limit the fluctuation range of the probability distribution; Step 1.3.2, Model Solving: The column and constraint generation C&CG algorithm is used to iteratively solve the above two-stage sub-Bruker optimization problem, which is decomposed into a main problem M and a subproblem to be solved alternately until the upper and lower bounds converge.

5. The virtual power plant two-layer distributed bar scheduling method considering shared energy storage according to claim 1, characterized in that: The specific method in step 2.1 is as follows: The entire system is decomposed into multiple independent aggregation units, each unit The goal is to minimize its local cost function The units are interconnected through energy exchange, and are denoted as units. With unit The interaction power between them is And it needs to satisfy consistency constraints: .

6. The virtual power plant two-layer distributed bar scheduling method considering shared energy storage according to claim 1, characterized in that: The specific method in step 2.2 is as follows: To solve this distributed optimization problem, an augmented Lagrangian function is constructed, and the ADMM algorithm is used iteratively. The solution is obtained at the 6th... q+1 In the next iteration: Local optimization steps, namely Update: Each aggregation unit Given the interaction power of other units and the current Lagrange multipliers, In this case, solve the local optimization problem to update its internal decision variables. and planned interaction power The objective function for optimization is: in, It is the coefficient of the second penalty term. In this step, each unit only uses its own private information to make decisions. Price update steps, namely Update: After all units have completed local optimization, the coordination center collects the planned interaction power of each unit and, based on the degree of violation of consistency constraints, i.e., the residual... To update the Lagrange multipliers : The updated multipliers will be sent to each unit as the price signal for the next iteration; The above steps are repeated until both the primal residual and the dual residual are less than the preset threshold, indicating that the algorithm has converged and all parties have reached a consensus on the energy exchange plan.

Images