Shared energy storage fine scheduling method considering multi-scene uncertainty and hierarchical decision

Abstract

The invention discloses a shared energy storage fine scheduling method considering multi-scene uncertainty and hierarchical decision, and the method comprises the steps: building system power-energy constraint modeling based on a typical scene set; modeling is carried out based on a typical scene set and system power-energy constraint, electricity purchasing and selling, energy storage efficiency and life degradation cost are comprehensively counted in multiple scenes, a multi-target scheduling model is formed, and opportunity constraint is introduced to ensure supply and demand reliability; based on a multi-target scheduling model, constructing a hierarchical decision-making model which takes a shared energy storage operator as a main part and takes a plurality of micro-grids / communities as an auxiliary part, and establishing a coupling constraint and power balance relation of energy storage charging and discharging-energy state-capacity leasing price; an energy storage dynamic rental price, capacity distribution and a time-sharing charging and discharging plan are obtained by adopting distributed iteration solution, and income allocation can be carried out in combination with an improved Shapley mechanism. According to the invention, the optimal decision of consumption improvement, energy consumption cost reduction and multi-agent coordinated operation under the condition of high-proportion access of new energy can be realized.

Description

Technical Field

[0001] This invention relates to the field of scheduling optimization technology in distributed photovoltaic / wind power scenarios, specifically a fine scheduling method for shared energy storage that considers uncertainties in multiple scenarios and hierarchical decision-making. Background Technology

[0002] Existing scheduling optimization methods for distributed photovoltaic / wind power scenarios generally suffer from the following shortcomings: First, they lack precise characterization of uncertainties. Some methods only employ simple probability distributions or perform static stochastic optimization on a limited number of scenarios, making it difficult to simultaneously consider intraday and interday temporal correlations. This can easily lead to insufficient response to extreme scenarios, resulting in wind and solar curtailment or significant fluctuations in electricity purchase costs. Second, they neglect efficiency losses and lifespan degradation during the operation of shared energy storage. They often treat energy storage as an ideal charging and discharging device, considering only power balance without factoring in the impact of cycle count and deep discharge on lifespan. This can easily lead to lower-than-expected investment recovery in energy storage over long-term operation, weakening the project's economic viability.

[0003] Patent application CN113255129A discloses a method and system for incremental distribution network two-layer stochastic optimization operation. This method is applicable to situations with multiple participating entities, achieving optimal scheduling and economical operation of incremental distribution networks. However, the optimization object of this patent is the "incremental distribution network operator," scheduling various units and energy storage throughout the entire distribution network. In terms of cost modeling, only traditional operating / power purchase / sale / reservation costs are considered. Regarding risk constraints, multi-scenario stochastic programming is used, primarily aiming for minimum expected cost. Furthermore, the patent's solution is based on a centralized approach: establishing upper and lower level models → KKT → Big-M → single-layer MILP → solution, completed in one step. Regarding the revenue sharing mechanism, only the power and cost of each entity under the optimal solution are given, without considering how alliance revenue is distributed. In terms of decision variables and master-slave linkage, the master-slave relationship is implicitly reflected through two-layer optimization.

[0004] Patent application CN117060438A discloses a method and apparatus for generating frequency regulation schemes that coordinate power generation and carbon emissions. This method constructs an energy flow model for energy flow and a carbon trading model for carbon emissions based on the power grid's energy framework. These two models are then combined to form an optimized operation model, which is then solved to generate a frequency regulation scheme. This can coordinate the power generation of the power system with the electricity consumption of electric vehicles and avoid exceeding carbon emission limits. However, in this patent, all equipment (combined cooling, heating, and power, energy storage, EVs, etc.) and carbon trading are unified into a centralized optimization model, solved by a single "community system planner" / frequency regulation controller, without a clear master-slave role distinction. Furthermore, energy storage equipment is merely an "internal device" within the community energy system, participating in energy balance only through energy storage power and SOC constraints, without the concept of "shared energy storage leasing / capacity leasing prices." Moreover, the patent's demand response defines the amount of load change and load reduction through the "price elasticity coefficient of electricity price to transferable / reducible load," primarily used to couple electricity price and load in the constraints. Summary of the Invention

[0005] The technical problem to be solved by this invention is to solve the problem of integrated energy storage scheduling and pricing under the uncertainty of new energy output and the collaboration of multiple entities.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: A method for fine-grained scheduling of shared energy storage considering uncertainties in multiple scenarios and hierarchical decision-making includes: Based on conditional generative adversarial networks and Markov chains, a typical scenario set of new energy output and load is generated to characterize intraday / interday randomness and temporal correlation and establish a system work-energy constraint model. Based on typical scenario sets and system power-energy constraint modeling, a multi-objective scheduling model is formed by comprehensively considering electricity purchase and sale, energy storage efficiency and lifetime degradation costs in multiple scenarios, and opportunity constraints are introduced to ensure supply and demand reliability. Based on the multi-objective scheduling model, a hierarchical decision-making model is constructed with shared energy storage operators as the main body and multiple microgrids / communities as the followers, and the coupling constraints and power balance relationship of energy storage charging and discharging, energy status and capacity leasing price are established. The distributed iterative solution is used to obtain the dynamic leasing price, capacity allocation and time-of-use charging and discharging plan of energy storage, and the revenue sharing can be combined with the improved Shapley mechanism.

[0007] In this embodiment, a typical set of scenarios for generating new energy output and load based on conditional generative adversarial networks and Markov chains includes: Establish samples and generator output, including: establishing a scenario dataset based on new energy power output and load, and outputting new energy power output according to the condition generator; A state transition matrix is ​​constructed, and based on the state transition matrix and the scenario dataset, a typical scenario set is established. Within the scenario, power balance and power limit constraints, energy storage evolution and boundary constraints are established.

[0008] In this embodiment, establishing power balance and power limit constraints within the scene includes: Power balance constraints: (5); In the formula, , , Scenes Time period Solar, wind, and load, , These represent the energy storage charging / discharging power, , These refer to the power purchased / sold from the power grid, respectively. Charge and discharge limit constraints: (6); In the formula, , These are the limits for energy storage charging and discharging; Energy storage energy evolution and boundary constraints, including: Energy storage state constraints: (7); In the formula, , They are respectively , Energy storage capacity during a given period of time , These represent the energy storage charging / discharging efficiency, Indicates the length of the time period; Energy storage boundary constraints: (8); In the formula, , These represent the upper and lower bounds of the stored energy, , Representing the initial value of energy storage and T The system stores energy values ​​at all times, ensuring consistency between the beginning and end of the day to guarantee a closed loop.

[0009] In this embodiment, a multi-objective scheduling model is formed by comprehensively considering electricity purchase and sale, energy storage efficiency, and lifetime degradation costs across multiple scenarios. Opportunity constraints are introduced to ensure supply and demand reliability, including: Establish a multi-objective scheduling model: (9); In the formula, , , These represent the weights of multiple objectives. , , These represent operating / degradation / carbon costs, respectively. It is a multi-objective function; The operating cost, degradation cost, and carbon cost of the multi-objective scheduling model are expressed using the following relationship: (10); ,(11) (12); In the formula, , , , These represent the unit cost / revenue for purchasing and selling electricity and charging / discharging electricity, respectively. , These represent the energy storage charging / discharging power, , These are the power purchased / sold from the power grid, respectively. Indicates the length of the time period. Indicates the degradation coefficient. Indicates the rated capacity of energy storage. This indicates the unit cost of carbon for purchasing electricity.

[0010] In this embodiment, a hierarchical decision-making model is constructed, with shared energy storage operators as the main body and multiple microgrids / communities as the followers. This model establishes the coupling constraints and power balance relationships between energy storage charging / discharging, energy state, and capacity leasing prices, including: The hierarchical decision-making model includes microgrid-side flexible load modeling and power purchase balance constraints, power purchase balance and cost function, operator-led profit model and price / capacity boundary; Based on a hierarchical decision-making model, a dynamic coupling and master-slave linkage of price-power closed loop are established.

[0011] In this embodiment, the modeling of flexible loads on the microgrid side and the constraints of power purchase balance are established, including: The load configuration is established using equation (13): (13); In the formula, microgrid During the period Total load, Indicates the base load. Indicates load transfer. This indicates a reduction in load; Use equation (14) to establish transferable / reducible constraints: (14); In the formula, This indicates the maximum load reduction. Use equation (15) to establish upper and lower limits of load and upper limit of transfer constraints: (15); In the formula, Indicates the maximum value of the transferred load. , These represent the minimum and maximum total load values, respectively.

[0012] In this embodiment, the power purchase balance and cost function is established, including: The power balance function between the microgrid and the grid / energy storage is established using equation (16): ,(16) The microgrid operating cost function is established using equation (17): (17); In the formula, This indicates net electricity purchases from the grid. Indicates photovoltaic power on the microgrid side. microgrid During the period Total load, , They represent microgrids During the period Charging power and microgrid During the period The discharge power, Indicates time-of-use electricity pricing. , These represent the user utility loss coefficients, Indicates load transfer. This indicates a reduction in load. This represents the microgrid operating cost function. Indicates the length of the time period.

[0013] In this embodiment, a profit model and price / capacity boundary are established primarily based on the operator, including: Establish the operator's profit maximization function using equation (18): (18); In the formula, Indicates operator profits, , These represent the rental prices for parking and charging, respectively. , These represent fixed and variable costs, respectively. , These represent the time periods of shared energy storage. microgrids Charging power and shared energy storage during the period To microgrids The discharge power, Indicates the length of the time period; Use equation (19) to establish the fixed / variable cost function: ,(19) In the formula, , , Indicates the cost coefficient. Indicates rated power. , These represent the time periods of shared energy storage. Charging power and shared energy storage during the period The discharge power, This indicates the rated capacity of the shared energy storage system.

[0014] Establish the price boundary using equation (20): (20); In the formula, Indicates a price range. It can be any symbol; Furthermore, using equation (21), a dynamic coupling and master-slave linkage of the price-power closed loop are established: ,(twenty one); In the formula, Indicates the price update gain. Indicates the target discharge reference. The charging and discharging price of shared energy storage Rate of change over time.

[0015] In this embodiment, a distributed iterative solution is used to obtain the dynamic leasing price, capacity allocation, and time-of-use charging / discharging plan for energy storage. Furthermore, an improved Shapley mechanism can be incorporated for revenue sharing, including: Solve the local response optimization problem independently from the perspective of the system under a given price, and establish the system response and energy storage constraints; On the operator side, the algorithm converges by using a price iteration function based on the microgrid power response, combined with price update gain, convergence threshold, and end-point energy penalty function. At the same time, the improved Shapley value function is used to correct and allocate the total alliance revenue according to the net power waveform similarity and correlation, so as to establish the alliance revenue allocation and iterative solution.

[0016] In this embodiment, establishing slave response and energy storage constraints includes: Establish the response optimization function using equation (22): ,(twenty two); In the formula, , microgrids During the period The charging power to shared energy storage and microgrids During the period The discharge power obtained from shared energy storage, microgrid Total cost This represents the microgrid operating cost function. Indicates the length of the time period. Indicates the price of electric vehicle rental; Energy storage constraints are established using equation (23): ,(twenty three); In the formula, microgrid The upper limit of available energy storage capacity on the side. This represents the energy stored in the microgrid. , These represent the energy storage charging / discharging efficiency, respectively. Use equation (24) to establish the first and last stage and boundary constraints of energy storage: ,(twenty four); In the formula, , These represent the minimum and maximum energy storage values ​​of the microgrid, respectively. , They represent microgrids Energy storage state and microgrid at the start of the scheduling cycle The energy storage status at the end of the scheduling cycle.

[0017] In this embodiment, the establishment of alliance revenue sharing and iterative solution includes: Establish the Shapley value function using equation (25): (25); In the formula, microgrid Total cost For Shapley value functions, This represents a subset of participating microgrids, used as the set of cooperative alliances for Shapley value calculation. To participate in microgrid aggregation, Represents the alliance value function, The union operator for sets; Establish a similarity correction function using equation (26): ,(26; In the formula, Represents the waveform similarity coefficient. Represents the net power correlation coefficient. This is the similarity-corrected Shapley value function; Establish the price iteration function using equation (27): (27); In the formula, Indicates the iteration round, Indicates the price step size. , They are respectively , The discharge rental price for each iteration round, Indicates the first In this iteration, the microgrid During the period The discharge power, For target discharge reference; The power response function is established using equation (28): (28); In the formula, Indicates the first In this iteration, the microgrid During the period The optimal discharge power, microgrid During the period Discharge power obtained from shared energy storage; Establish convergence criterion constraints using equation (29): (29); In the formula, Indicates the convergence threshold; The terminal energy penalty function is established using equation (30): (30); In the formula, Indicates the overall objective including penalties. This represents the terminal energy deviation penalty coefficient. Represents the probability of a scenario. , Representing the initial value of energy storage and TEnergy storage value at all times This represents a subset of participating microgrids, used as the set of cooperative alliances for Shapley value calculation. It is a multi-objective function.

[0018] Compared with the prior art, the beneficial effects of the present invention are: This invention addresses the challenges of simultaneously considering the multi-scenario randomness of renewable energy output and load demand, the efficiency and lifespan degradation of shared energy storage, carbon emission costs, and the hierarchical game relationship among multiple microgrid participants in uncertain scenarios for distributed photovoltaic / wind power. It constructs an integrated method for refined scheduling and pricing of shared energy storage that can be implemented in practice.

[0019] This invention, while ensuring computational feasibility, generates a set of typical scenarios that can characterize intraday / interday temporal correlations and applies them to power-energy constraint modeling with shared energy storage. The invention achieves a refined fusion of multi-scenario randomness and power-energy constraints: by jointly generating a set of typical scenarios of new energy output and load through a conditional generative adversarial network (GAN) and Markov chains, it maintains the statistical characteristics of historical data while characterizing intraday / interday temporal correlations. Within each scenario, power balance, upper and lower limits of energy storage power, and state of charge (SOC) evolution and boundary constraints are established, achieving refined modeling of output uncertainty and energy storage power-energy coupling, thus improving robustness to extreme scenarios.

[0020] This invention integrates electricity purchase and sale costs, energy storage efficiency losses, lifespan degradation costs, and carbon costs into a multi-objective scheduling model under various scenarios, and ensures supply and demand reliability at a certain confidence level through opportunity constraints. Multi-objective scheduling and opportunity constraints integrating operating costs, degradation costs, and carbon costs: Through multi-objective functions (9)–(12), electricity purchase and sale costs, energy storage lifespan degradation costs, and carbon costs are unified into a multi-objective scheduling framework, and weights are set to reflect operators' different preferences for economy, lifespan, and low carbon emissions; at the same time, opportunity constraints are introduced so that supply and demand balance and the state of charge (SOC) safety boundary are met in most scenarios, thereby reducing operating costs while controlling the risk of load shedding and lifespan loss, and improving the full life cycle economy and low carbon emissions of the shared energy storage system. Avoiding one-sided optimization that "only looks at short-term electricity costs and ignores long-term lifespan and carbon costs", achieving a synergistic trade-off between economic, lifespan, and low carbon objectives.

[0021] This invention revolves around the structure of "shared energy storage operator - multiple microgrids / communities" and constructs a hierarchical decision-making model that couples capacity leasing price, energy storage charging and discharging, and flexible load decision-making, so as to achieve a balance between maximizing operator profits and user energy costs and comfort. Hierarchical decision-making and flexible load collaborative modeling from the perspective of shared energy storage: This invention takes the shared energy storage operator as the main body and multiple microgrids / communities as the slaves, and constructs a hierarchical decision-making model that includes load splitting (basic / transfer / reduction), power purchase balance, user utility loss cost, operator profit and capacity leasing price boundary, as shown in equations (13)–(20). Under a given leasing price and capacity, the microgrid participates in demand response by adjusting the transferable and reducible loads; the operator optimizes capacity allocation and price decisions based on the response results of each microgrid. This structure realizes the organic coupling between shared energy storage capacity leasing, microgrid flexible load response and price mechanism, and improves the shortcomings of traditional centralized models that are difficult to reflect master-slave game and business model. By taking into account capacity leasing price, flexible load demand response and microgrid power purchase decision-making in a coordinated manner, it reflects the real market interaction mechanism and improves the sustainability of the shared energy storage business model.

[0022] This invention designs a distributed iterative solution and revenue sharing mechanism to achieve coordinated convergence of dynamic energy storage leasing prices, capacity allocation, and time-of-use charging and discharging plans without exposing the private data of each microgrid, and to fairly share the economic benefits brought by shared energy storage. At the solution level, this invention adopts a distributed iterative framework: the local response optimization problem is solved independently from the body side under a given price, Equations (22)–(24); the operator side updates the leasing price according to the microgrid power response through the price iteration function, Equation (27), and combines the price update gain, convergence threshold, and end-point energy penalty function, Equations (29)–(30) to ensure algorithm convergence; at the same time, the improved Shapley value function is used, Equations (25)–(26), to correct and share the total revenue of the alliance according to the net power waveform similarity and correlation, realizing "revenue sharing according to contribution and degree of coordination". This mechanism significantly improves the feasibility and fairness of the shared energy storage coordinated scheduling scheme while ensuring the privacy of all parties. It does not require each microgrid to disclose complete cost and load preference information, making it easier for projects to be implemented; at the same time, it fairly distributes revenue according to contribution, enhancing the willingness of participating entities to participate in shared energy storage projects in the long term.

[0023] The scheduling objects of this invention are the charging, discharging, and capacity configuration of shared energy storage stations in multiple typical scenarios, and serve the leasing needs of different microgrids / communities. It belongs to the perspective of "shared energy storage operation" rather than "whole network operation". It explicitly introduces the energy storage efficiency loss and lifetime degradation costs, and incorporates carbon costs into a unified multi-objective function, and gives the sub-item modeling and weight combination through equations (9)-(12). In multiple scenarios, it ensures the supply and demand balance and the energy storage SOC safety boundary under a given confidence level through opportunity constraints, and emphasizes "risk control" rather than just "expected optimization".

[0024] This invention is based on a master-slave game framework, achieving bidirectional feedback between the operator (master) and each microgrid (slave) through price-power iteration, and using convergence conditions to control the end of the iteration, which is a typical distributed / iterative coordination strategy. By combining the Shapley value and the similarity and correlation of the microgrid net power waveform, the total revenue from shared energy storage is finely allocated, solving the problem of "how to fairly allocate the revenue from shared energy storage".

[0025] In the shared energy storage business model and capacity leasing price model, this invention introduces "release / charge leasing price" and fixed / variable costs, as shown in equations (18) and (19); the price range is defined by equation (20); the upper-level goal is to maximize operator profits rather than minimize frequency regulation costs / system costs.

[0026] In the characterization of flexible load and user utility, this invention first decomposes the total load of the microgrid into three types of physical load: basic, transfer, and reduction, as shown in equations (13)–(15), emphasizing the scope of physically feasible load migration and reduction; then, in the cost function, equation (17), a "user utility loss coefficient" is introduced to transform the "comfort / satisfaction" of demand response into a monetized penalty, reflecting the meaning of behavioral economics; the load decision is directly affected by the shared energy storage capacity and price, and is more closely embedded in the hierarchical game. Attached Figure Description

[0027] Figure 1 This is a flowchart of a shared energy storage fine scheduling method that considers uncertainties in multiple scenarios and hierarchical decision-making, according to an embodiment of the present invention.

[0028] Figure 2 This is a flowchart of the scheduling method according to an embodiment of the present invention. Detailed Implementation

[0029] To facilitate understanding of the technical solution of the present invention by those skilled in the art, the technical solution of the present invention will now be further described in conjunction with the accompanying drawings.

[0030] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0031] Please see Figure 1 , 2 As shown, this invention provides a method for fine-grained scheduling of shared energy storage that considers uncertainties in multiple scenarios and hierarchical decision-making, including: S10 generates a typical set of new energy output and load based on conditional generative adversarial networks and Markov chains, characterizes intraday / interday randomness and temporal correlation, and establishes a system work-energy constraint model.

[0032] In one embodiment of the present invention, a typical scenario set of renewable energy output and load is generated based on a conditional generative adversarial network and a Markov chain. This includes: establishing samples and generator output, specifically: establishing a scenario dataset based on renewable energy output and load, and outputting renewable energy output according to the conditional generator. A state transition matrix is ​​constructed, and based on the state transition matrix and the scenario dataset, a typical scenario set is established. Within each scenario, power balance and power limit constraints, energy storage evolution, and boundary constraints are established.

[0033] In this embodiment, specifically, establishing the sample and generator output includes: Establish a scene dataset using equation (1) : ,(1); In equation (1), They represent time periods respectively. Photovoltaic power, wind power and load power, This represents the time-domain set of scheduling.

[0034] Using equation (2) to establish conditional generation and : (2); In equation (2), and These represent the generators of the adversarial network. and Under conditions With noise The power generated below.

[0035] In this embodiment, a state transition matrix is ​​constructed, and a typical scenario set is established based on the state transition matrix and the scenario dataset, including: Establish the state transition matrix using equation (3): (3); In equation (3), Represents the state transition probability. Indicates from state arrive Observational transfer counts, This represents the number of cluster states.

[0036] Based on the state transition probability, typical scenarios are formed. Specifically, a set of typical scenarios is established using equation (4): (4); In equation (4), This represents a subset of participants in a microgrid. Indicates the scene number. Indicates the number of scenes. Indicates the first s A typical scenario includes power sequences from photovoltaics, wind power, and loads. , These represent typical scenarios. s Mid-term t Photovoltaic power output and typical scenarios s Mid-term t Wind power output Representing typical scenarios s Mid-term t The load power. State transition probability. It operates in the process of "generating typical scenes". In formula (4), a complete scene sequence is to be generated: starting state → according to Π ij The standard way to generate Markov chain scenarios is to transition to the next state, retrieve the corresponding power sequence, and form the entire "typical scenario".

[0037] In this embodiment, establishing power balance and power limit constraints, energy storage evolution and boundary constraints within the scenario includes: Establish power balance and power limits within the scenario, including: Power balance constraints are established using equation (5): (5); In equation (5), , , Representing the scene respectively Time period Solar, wind, and load, , These represent the energy storage charging / discharging power, , These represent the power purchased / sold from the power grid, respectively.

[0038] Use equation (6) to establish charge / discharge limit constraints: (6); In equation (6), , These represent the energy storage charging / discharging power, , These represent the limits for energy storage charging and discharging.

[0039] Establish energy storage energy evolution and boundary constraints, including: Energy storage state constraints are established using equation (7): (7); In equation (7), , They are respectively , Energy storage capacity during a given period of time , These represent the energy storage charging / discharging efficiency, Indicates the length of the time period.

[0040] Establish energy storage boundary constraints using equation (8): (8); In equation (8), , These represent the upper and lower bounds of the stored energy, , Representing the initial value of energy storage and T The system stores energy values ​​at all times, ensuring consistency between the beginning and end of the day to guarantee a closed loop.

[0041] S20, based on typical scenario sets and system power-energy constraint modeling, comprehensively considers electricity purchase and sale, energy storage efficiency and lifetime degradation costs in multiple scenarios to form a multi-objective scheduling model, and introduces opportunity constraints to ensure supply and demand reliability.

[0042] In one embodiment of the present invention, the following steps are specifically included: Establish a multi-objective scheduling model and use equation (9) to establish the multi-objective function. Relationship: (9); In the formula, , , These represent the weights of multiple objectives. , , These represent operating / degradation / carbon costs, respectively. It is a multi-objective function.

[0043] The operating cost, degradation cost, and carbon cost of the multi-objective scheduling model are expressed using the following relationship.

[0044] Establish operation using formula (10) Cost relationship: (10); In equation (10), , , , These represent the unit cost / revenue for purchasing and selling electricity and charging / discharging, respectively.

[0045] Establish degradation cost using equation (11) Relationship: ,(11) In equation (11), Indicates the degradation coefficient. This indicates the rated capacity of the energy storage.

[0046] Carbon cost is established using equation (12). Relationship: (12); In equation (12), This indicates the unit cost of carbon for purchasing electricity.

[0047] S30, based on a multi-objective scheduling model, constructs a hierarchical decision-making model with shared energy storage operators as the main body and multiple microgrids / communities as the secondary body, and establishes the coupling constraints and power balance relationship between energy storage charging and discharging, energy state, and capacity leasing price.

[0048] In one embodiment of the present invention, the hierarchical decision-making model includes microgrid-side flexible load modeling and power purchase balance constraints, power purchase balance and cost functions, operator-led profit model, and price / capacity boundaries. Based on the hierarchical decision-making model, a dynamic coupling and master-slave linkage of the price-power closed loop is established.

[0049] In this embodiment, the modeling of flexible loads on the microgrid side and the constraints of power purchase balance are established, including: The load configuration is established using equation (13): (13); In the formula, microgrid During the period Total load, Indicates the base load. Indicates load transfer. This indicates a reduction in load.

[0050] Use equation (14) to establish transferable / reducible constraints: (14); In the formula, This indicates the maximum load reduction. Use equation (15) to establish upper and lower limits of load and upper limit of transfer constraints: (15); In the formula, Indicates the maximum value of the transferred load. , These represent the minimum and maximum total load values, respectively.

[0051] In this embodiment, the power purchase balance and cost function is established, including: The power balance function between the microgrid and the grid / energy storage is established using equation (16): ,(16) In equation (16), This indicates net electricity purchases from the grid. Indicates photovoltaic power on the microgrid side. microgrid During the period Total load, , They represent microgrids During the period Charging power and microgrid During the period The discharge power.

[0052] The microgrid operating cost function is established using equation (17): (17); In equation (17), Indicates time-of-use electricity pricing. , These represent the user utility loss coefficients, Indicates load transfer. This indicates a reduction in load. This represents the microgrid operating cost function. Indicates the length of the time period.

[0053] In this embodiment, a profit model and price / capacity boundary are established primarily based on the operator, including: Establish the operator's profit maximization function using equation (18): (18); In the formula, Indicates operator profits, , These represent the rental prices for parking and charging, respectively. , These represent fixed and variable costs, respectively. , These represent the time periods of shared energy storage. microgrids Charging power and shared energy storage during the period To microgrids The discharge power, Indicates the length of the time period; Use equation (19) to establish the fixed / variable cost function: ,(19) In the formula, , , Indicates the cost coefficient. Indicates rated power. , These represent the time periods of shared energy storage. Total charging power and shared energy storage during the period Total discharge power, This indicates the rated energy capacity of the shared energy storage system.

[0054] Establish the price boundary using equation (20): (20); In the formula, Indicates a price range. It can be any symbol.

[0055] In this embodiment, the dynamic coupling and master-slave linkage of the price-power closed loop are established using equation (21): ,(twenty one); In the formula, Indicates the price update gain. Indicates the target discharge reference. The charging and discharging price of shared energy storage Rate of change over time.

[0056] S40 uses distributed iterative solutions to obtain dynamic energy storage leasing prices, capacity allocation, and time-of-use charging and discharging plans, and can be combined with an improved Shapley mechanism for revenue sharing.

[0057] In one embodiment of the invention, the slave side independently solves the local response optimization problem under a given price, establishing slave response and energy storage constraints. The operator side uses a price iteration function based on the microgrid power response, combined with price update gain, convergence threshold, and end-point energy penalty function to ensure algorithm convergence. Simultaneously, an improved Shapley value function is used to correct and allocate the total alliance revenue based on net power waveform similarity and correlation, thus establishing alliance revenue allocation and iterative solution.

[0058] In this embodiment, establishing slave response and energy storage constraints includes: Establish the response optimization function using equation (22): ,(twenty two); In the formula, , microgrids During the period The charging power to shared energy storage and microgrids During the period The discharge power obtained from shared energy storage, microgrid Total cost This represents the microgrid operating cost function. Indicates the length of the time period. Indicates the price of electric vehicle rental; Energy storage constraints are established using equation (23): ,(twenty three); In the formula, microgrid The upper limit of available energy storage capacity on the side. This represents the energy stored in the microgrid. , These represent the energy storage charging / discharging efficiency, respectively. Use equation (24) to establish the first and last stage and boundary constraints of energy storage: ,(twenty four); In the formula, , These represent the minimum and maximum energy storage values ​​of the microgrid, respectively. , They represent microgrids At the start of the scheduling cycle ( t Energy storage energy state and microgrid (=0) At the end of the scheduling cycle ( t =T) energy storage state.

[0059] In this embodiment, the establishment of alliance revenue sharing and iterative solution includes: Establish the Shapley value function using equation (25): (25); In the formula, microgrid Total cost For Shapley value functions, This represents a subset of participating microgrids, used as the set of cooperative alliances for Shapley value calculation. To participate in microgrid aggregation, Represents the alliance value function. The union operator for sets.

[0060] Establish a similarity correction function using equation (26): ,(26; In the formula, Represents the waveform similarity coefficient. Represents the net power correlation coefficient. This is the similarity-corrected Shapley value function; Establish the price iteration function using equation (27): (27); In the formula, Indicates the iteration round, Indicates the price step size. , They are respectively , The discharge rental price for each iteration round, Indicates the first k In this iteration, the microgrid j During the period t The discharge power, For target discharge reference; The power response function is established using equation (28): (28); In the formula, Indicates the first In this iteration, the microgrid During the period The optimal discharge power, microgrid During the period Discharge power obtained from shared energy storage.

[0061] Establish convergence criterion constraints using equation (29): (29); In the formula, Indicates the convergence threshold; The terminal energy penalty function is established using equation (30): (30); In the formula, Indicates the overall objective including penalties. This represents the terminal energy deviation penalty coefficient. Represents the probability of a scenario. , Representing the initial value of energy storage and T Energy storage value at all times This represents a subset of participating microgrids, used as the set of cooperative alliances for Shapley value calculation. It is a multi-objective function.

[0062] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered illustrative and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention, and no reference numerals in the claims should be construed as limiting the scope of the claims.

[0063] The above embodiments are merely examples of implementation methods of the invention. The scope of protection of the present invention is not limited to the above embodiments. For those skilled in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all fall within the scope of protection of the present invention.

Claims

1. A method for fine-grained scheduling of shared energy storage considering uncertainties in multiple scenarios and hierarchical decision-making, characterized in that, include: Based on conditional generative adversarial networks and Markov chains, a typical scenario set of new energy output and load is generated to characterize intraday / interday randomness and temporal correlation and establish a system work-energy constraint model. Based on typical scenario sets and system power-energy constraint modeling, a multi-objective scheduling model is formed by comprehensively considering electricity purchase and sale, energy storage efficiency and lifetime degradation costs in multiple scenarios, and opportunity constraints are introduced to ensure supply and demand reliability. Based on the multi-objective scheduling model, a hierarchical decision-making model is constructed with shared energy storage operators as the main body and multiple microgrids / communities as the followers, and the coupling constraints and power balance relationship of energy storage charging and discharging, energy status and capacity leasing price are established. The distributed iterative solution is used to obtain the dynamic leasing price, capacity allocation and time-of-use charging and discharging plan of energy storage, and the revenue sharing can be combined with the improved Shapley mechanism.

2. The method for fine-grained scheduling of shared energy storage considering multi-scenario uncertainties and hierarchical decision-making as described in claim 1, characterized in that, Typical scenarios for generating new energy output and load based on conditional generative adversarial networks and Markov chains include: Establish samples and generator output, including: establishing a scenario dataset based on new energy power output and load, and outputting new energy power output according to the condition generator; A state transition matrix is ​​constructed, and based on the state transition matrix and the scenario dataset, a typical scenario set is established. Within the scenario, power balance and power limit constraints, energy storage evolution and boundary constraints are established.

3. The method for fine-tuning shared energy storage considering multi-scenario uncertainties and hierarchical decision-making as described in claim 2, characterized in that, Establish scene-specific power balance and power limit constraints, including: Power balance constraints: ，（5）； In the formula, , , Scenes Time period Solar, wind, and load, , These represent the energy storage charging / discharging power, , These refer to the power purchased / sold from the power grid, respectively. Charge and discharge limit constraints: ，（6）； In the formula, , These are the limits for energy storage charging and discharging; Energy storage energy evolution and boundary constraints, including: Energy storage state constraints: ，（7）； In the formula, , They are respectively , Energy storage capacity during a given period of time , These represent the energy storage charging / discharging efficiency, Indicates the length of the time period; Energy storage boundary constraints: ，（8）； In the formula, , These represent the upper and lower bounds of the stored energy, , Representing the initial value of energy storage and T The system stores energy values ​​at all times, ensuring consistency between the beginning and end of the day to guarantee a closed loop.

4. The method for fine-grained scheduling of shared energy storage considering multi-scenario uncertainties and hierarchical decision-making as described in claim 1, characterized in that, By comprehensively considering electricity purchase and sale, energy storage efficiency, and lifespan degradation costs across multiple scenarios, a multi-objective scheduling model is formed. Opportunity constraints are introduced to ensure supply and demand reliability, including: Establish a multi-objective scheduling model: ，（9）； In the formula, , , These represent the weights of multiple objectives. , , These represent operating / degradation / carbon costs, respectively. It is a multi-objective function; The operating cost, degradation cost, and carbon cost of the multi-objective scheduling model are expressed using the following relationship: ，（10）； ，（11）； ，（12）； In the formula, , , , These represent the unit cost / revenue for purchasing and selling electricity and charging / discharging electricity, respectively. , These represent the energy storage charging / discharging power, , These are the power purchased / sold from the power grid, respectively. Indicates the length of the time period. Indicates the degradation coefficient. Indicates the rated capacity of energy storage. This indicates the unit cost of carbon for purchasing electricity.

5. The method for fine-grained scheduling of shared energy storage considering multi-scenario uncertainties and hierarchical decision-making as described in claim 1, characterized in that, A hierarchical decision-making model is constructed, with shared energy storage operators as the main body and multiple microgrids / communities as the followers. This model establishes the coupling constraints and power balance relationships between energy storage charging / discharging, energy state, and capacity leasing prices, including: The hierarchical decision-making model includes microgrid-side flexible load modeling and power purchase balance constraints, power purchase balance and cost function, operator-led profit model and price / capacity boundary; Based on a hierarchical decision-making model, a dynamic coupling and master-slave linkage of price-power closed loop are established.

6. The method for fine-grained scheduling of shared energy storage considering multi-scenario uncertainties and hierarchical decision-making as described in claim 5, characterized in that, Establish flexible load modeling and power purchase balance constraints on the microgrid side, including: The load configuration is established using equation (13): ，（13）； In the formula, microgrid During the period Total load, Indicates the base load. Indicates load transfer. This indicates a reduction in load; Use equation (14) to establish transferable / reducible constraints: ，（14）； In the formula, This indicates the maximum load reduction. Use equation (15) to establish upper and lower limits of load and upper limit of transfer constraints: ，（15）； In the formula, Indicates the maximum value of the transferred load. , These represent the minimum and maximum total load values, respectively. Establish the electricity purchase balance and cost function, including: The power balance function between the microgrid and the grid / energy storage is established using equation (16): ，（16）； The microgrid operating cost function is established using equation (17): ，（17）； In the formula, This indicates net electricity purchases from the grid. Indicates photovoltaic power on the microgrid side. microgrid During the period Total load, , They represent microgrids During the period Charging power and microgrid During the period The discharge power, Indicates time-of-use electricity pricing. , These represent the user utility loss coefficients, Indicates load transfer. This indicates a reduction in load. This represents the microgrid operating cost function. Indicates the length of the time period.

7. A method for fine-grained scheduling of shared energy storage considering multi-scenario uncertainties and hierarchical decision-making as described in claim 5, characterized in that, Establish a profit model and price / capacity boundaries primarily for operators, including: Establish the operator's profit maximization function using equation (18): ，（18）； In the formula, Indicates operator profits, , These represent the rental prices for parking and charging, respectively. , These represent fixed and variable costs, respectively. , These represent the time periods of shared energy storage. microgrids Charging power and shared energy storage during the period To microgrids The discharge power, Indicates the length of the time period; Use equation (19) to establish the fixed / variable cost function: ，（19）； In the formula, , , Indicates the cost coefficient. Indicates rated power. , These represent the time periods of shared energy storage. Total charging power and shared energy storage during the period Total discharge power, Indicates the rated energy capacity of the shared energy storage system; Establish the price boundary using equation (20): ，（20）； In the formula, Indicates a price range. It can be any symbol; Furthermore, using equation (21), a dynamic coupling and master-slave linkage of the price-power closed loop are established: ，（21）； In the formula, Indicates the price update gain. Indicates the target discharge reference. The charging and discharging price of shared energy storage Rate of change over time.

8. The method for fine-grained scheduling of shared energy storage considering multi-scenario uncertainties and hierarchical decision-making as described in claim 1, characterized in that, Distributed iterative solutions are used to obtain dynamic energy storage leasing prices, capacity allocation, and time-of-use charging / discharging plans. Furthermore, an improved Shapley mechanism can be incorporated for revenue sharing, including: Solve the local response optimization problem independently from the perspective of the system under a given price, and establish the system response and energy storage constraints; On the operator side, the algorithm converges by using a price iteration function based on the microgrid power response, combined with price update gain, convergence threshold, and end-point energy penalty function. At the same time, the improved Shapley value function is used to correct and allocate the total alliance revenue according to the net power waveform similarity and correlation, so as to establish the alliance revenue allocation and iterative solution.

9. A method for fine-grained scheduling of shared energy storage considering multi-scenario uncertainties and hierarchical decision-making as described in claim 8, characterized in that, Establish slave response and energy storage constraints, including: Establish the response optimization function using equation (22): ，（22）； In the formula, , microgrids During the period The charging power to shared energy storage and microgrids During the period The discharge power obtained from shared energy storage, microgrid Total cost This represents the microgrid operating cost function. Indicates the length of the time period. Indicates the price of electric vehicle rental; Energy storage constraints are established using equation (23): ，（23）； In the formula, microgrid The upper limit of available energy storage capacity on the side. This represents the energy stored in the microgrid. , These represent the energy storage charging / discharging efficiency, respectively. Use equation (24) to establish the first and last stage and boundary constraints of energy storage: ，（24）； In the formula, , These represent the minimum and maximum energy storage values ​​of the microgrid, respectively. , They represent microgrids Energy storage state and microgrid at the start of the scheduling cycle The energy storage status at the end of the scheduling cycle.

10. A method for fine-grained scheduling of shared energy storage considering multi-scenario uncertainties and hierarchical decision-making as described in claim 8, characterized in that, Establishing alliance revenue sharing and iterative solution, including: Establish the Shapley value function using equation (25): ，（25）； In the formula, microgrid Total cost For Shapley value functions, This represents a subset of participating microgrids, used as the set of cooperative alliances for Shapley value calculation. To participate in microgrid aggregation, Represents the alliance value function. The union operator for sets; Establish a similarity correction function using equation (26): ，（26）； In the formula, Represents the waveform similarity coefficient. Represents the net power correlation coefficient. This is the similarity-corrected Shapley value function; Establish the price iteration function using equation (27): ，（27）； In the formula, Indicates the iteration round, Indicates the price step size. , They are respectively , The discharge rental price for each iteration round, Indicates the first In this iteration, the microgrid During the period The discharge power, For target discharge reference; The power response function is established using equation (28): ，（28）； In the formula, Indicates the first In this iteration, the microgrid During the period The optimal discharge power, microgrid During the period Discharge power obtained from shared energy storage. Establish convergence criterion constraints using equation (29): ，（29）； In the formula, Indicates the convergence threshold; The terminal energy penalty function is established using equation (30): ，（30）； In the formula, Indicates the overall objective including penalties. This represents the terminal energy deviation penalty coefficient. Represents the probability of a scenario. , Representing the initial value of energy storage and T Energy storage value at all times This represents a subset of participating microgrids, used as the set of cooperative alliances for Shapley value calculation. It is a multi-objective function.

Images