Active distribution network-multi-microgrid collaborative optimization scheduling method and system

Abstract

This invention belongs to the field of power system optimization and scheduling technology for multi-microgrid systems, and particularly relates to a method and system for collaborative optimization scheduling of active distribution networks and multi-microgrids. It includes establishing a multi-entity shared energy storage operation framework; considering the bilateral uncertainties of active distribution networks and multi-microgrids, introducing adjustment costs and default penalties to quantify performance risks, and designing multi-timescale operation strategies for shared energy storage operators; establishing a multi-entity collaborative scheduling model based on Nash game theory, and decomposing the collaborative scheduling problem into two sub-problems: maximizing social benefits and allocating cooperative revenue; and introducing an ADMM distributed solution algorithm based on consensus variables and event-triggered communication to solve the two sub-problems, thereby achieving collaborative optimization scheduling among multiple entities. This invention improves the utilization efficiency of cross-layer energy storage resources and the level of renewable energy consumption, reduces system operating costs, and enhances the performance reliability of leasing plans under uncertain scenarios, while protecting the data privacy of multiple entities.

Description

Technical Field

[0001] This invention belongs to the field of power system optimization and dispatching technology for multi-microgrid power systems, and particularly relates to a method and system for collaborative optimization and dispatching of active distribution networks and multi-microgrids. Background Technology

[0002] Distributed renewable energy sources, with their flexibility, low losses, and clean and low-carbon characteristics, are showing increasingly broad application prospects. However, the uncertainties of renewable energy sources and the challenges of local consumption also pose serious challenges to the power system.

[0003] Existing research on the coordinated dispatch of multi-microgrid power systems still has the following shortcomings: (1) First, in terms of shared energy storage architecture, existing technologies mostly adopt a centralized shared energy storage mode. Users or microgrids only act as renters of energy storage capacity and do not regard the distributed energy storage inside the microgrid as a rentable redundant resource for cross-layer aggregation, thus making it difficult to fully tap the potential of redundant energy storage in multiple microgrids.

[0004] (2) Secondly, in terms of uncertainty handling, some studies only focus on the uncertainty of renewable energy output such as wind power and photovoltaics, and lack a leasing contract modeling and risk management mechanism for the uncertainty of both sides of the energy storage demand of active distribution network (ADN) and redundant energy storage supply of multi-microgrid, making it difficult to ensure the stable performance of cross-layer energy storage leasing plans in actual operation.

[0005] (3) Finally, regarding privacy protection, Active Distribution Networks (ADN), Shared Storage Operators (SESO), and various microgrids (MG) are generally considered as independent economic entities. Existing collaborative scheduling models mostly rely on centralized optimization or require entities to disclose complete operational data. Even when distributed solution algorithms are introduced, their design focus is still on reducing solution complexity, making it difficult to balance privacy protection and efficient resource utilization. Summary of the Invention

[0006] To overcome the shortcomings of the existing technologies, this invention provides a method and system for collaborative optimization scheduling of active distribution networks and multiple microgrids. It proposes a shared energy storage operation framework that allows for the leasing of redundant energy storage in multiple microgrids, constructs a risk-privacy-aware Nash game model, considers intraday uncertainty risk and privacy exposure costs in collaborative scheduling, and proposes an ADMM distributed solution algorithm based on consensus variables and event-triggered communication for solving the problem. This approach can improve the utilization efficiency of cross-layer energy storage resources and the level of renewable energy consumption while protecting the data privacy of multiple stakeholders, reduce system operating costs, and enhance the reliability of lease plan performance in uncertain scenarios.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions: The first aspect of this invention provides a method for coordinated optimization scheduling of active distribution networks and multiple microgrids.

[0008] The active distribution network-multi-microgrid collaborative optimization scheduling method includes the following steps: Establish a shared energy storage operation framework that includes active distribution networks, shared energy storage operators, and multiple microgrids and stakeholders; Considering the bilateral uncertainties of active distribution networks and multiple microgrids, adjustment costs and default penalties are introduced to quantify performance risks, and a multi-timescale operation strategy for shared energy storage operators is designed. Considering the risk perception cost and the privacy exposure risk cost, a multi-agent collaborative scheduling model based on Nash game is established, and the collaborative scheduling problem is decomposed into two sub-problems: maximizing social benefits and distributing cooperative benefits. An ADMM distributed solution algorithm based on consensus variables and event-triggered communication is introduced. The coupling relationship between the subjects is explicitly decoupled through consensus variables, and based on the event-triggered mechanism, the coupling variables are exchanged only when the triggering condition is met, and the two sub-problems are solved, realizing the collaborative optimization scheduling of multiple subjects.

[0009] The second aspect of the present invention provides an active distribution network-multi-microgrid collaborative optimization scheduling system.

[0010] An active distribution network-multi-microgrid collaborative optimization dispatch system includes: The framework building module is configured to: build a shared energy storage operation framework that includes active distribution networks, shared energy storage operators, and multiple microgrids and multiple stakeholders; The operation strategy design module is configured to: consider the bilateral uncertainties of active distribution networks and multiple microgrids, introduce adjustment costs and default penalties to quantify performance risks, and design multi-timescale operation strategies for shared energy storage operators; The collaborative scheduling model building module is configured to: consider the risk perception cost and the privacy exposure risk cost, establish a multi-agent collaborative scheduling model based on Nash game, and decompose the collaborative scheduling problem into two sub-problems: maximizing social benefits and distributing cooperative benefits. The solution module is configured to introduce the ADMM distributed solution algorithm based on consensus variables and event-triggered communication. The consensus variables explicitly decouple the coupling relationships between the subjects, and based on the event-triggered mechanism, the coupling variables are exchanged only when the triggering conditions are met, and the two sub-problems are solved to achieve collaborative optimization scheduling of multiple subjects.

[0011] The above one or more technical solutions have the following beneficial effects: This invention proposes a shared energy storage operation framework for redundant distributed energy storage leasing. In this framework, no energy storage is installed within the active distribution network (ADN). The shared energy storage operator possesses dispatchable centralized shared energy storage and can further aggregate the redundant distributed energy storage capacity within each microgrid through economic incentives, virtualizing it as a unified, dispatchable shared energy storage pool. Based on the energy storage requirements of the ADN and the leasing capacity of each microgrid, SESO jointly formulates a coordinated scheduling and leasing scheme for centralized and distributed energy storage, achieving unified management and efficient utilization of cross-layer energy storage resources.

[0012] This invention constructs a multi-timescale risk management strategy for SESO that considers bilateral uncertainties. Addressing the characteristic that both ADN energy storage demand and microgrid redundant energy storage supply are simultaneously affected by internal source-load uncertainties, this invention models both as bilateral uncertainties in a unified manner. For SESO, a multi-timescale operation strategy combining day-ahead planning and intraday adjustments is designed, introducing adjustment costs and default penalty costs. This improves system economics while enhancing the reliability of energy storage lease contract execution.

[0013] This invention establishes a risk- and privacy-aware Nash game-based collaborative scheduling and revenue distribution model. By decomposing the collaborative scheduling problem into two sub-problems—maximizing social benefits and distributing cooperative revenue—the globally optimal scheduling scheme and the rental prices and revenue distribution for each entity are determined, respectively. Intraday uncertainty risk measurement and privacy exposure costs are introduced to characterize the risk preferences and privacy sensitivities of different entities. Furthermore, bargaining factors are combined to characterize the contribution and bargaining power of ADN, SESO, and multiple microgrids in energy storage leasing, achieving a scheduling and distribution mechanism that balances overall benefits and individual fairness under controllable risk and privacy constraints.

[0014] This invention proposes an ADMM distributed solution algorithm that incorporates consensus variables and event-triggered communication. The consensus variables explicitly decouple the coupling relationships between stakeholders, allowing each staker to exchange only a small amount of boundary information without disclosing its detailed internal operating mechanisms. Simultaneously, the introduction of an event-triggered mechanism ensures that information is published only when the change in the current update relative to the previously published value exceeds a threshold. This reduces communication overhead and mitigates the risk of privacy breaches caused by differential inference of interaction sequences. While ensuring feasibility and convergence, this invention improves the efficiency of distributed solution and the level of privacy protection.

[0015] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0016] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0017] Figure 1 This is a flowchart of the method in Example 1.

[0018] Figure 2 This is a system structure diagram of Example 1.

[0019] Figure 3 This is a diagram showing the iterative convergence of subproblem 1 in Example 1.

[0020] Figure 4 This is a diagram showing the iterative convergence of subproblem 2 in Example 1.

[0021] Figure 5 This is a diagram showing the energy storage leasing situation of microgrid 1 in Example 1.

[0022] Figure 6 This is a diagram showing the energy storage self-use situation of microgrid 1 in Example 1.

[0023] Figure 7 This is a chart showing the rental prices of ADN and SESO in Example 1.

[0024] Figure 8 This is a chart showing the rental prices of SESO and microgrid 1 in Example 1.

[0025] Figure 9 This is a graph showing the changes in self-consumption energy storage of the microgrid within one day in Example 1.

[0026] Figure 10 This is a graph showing the changes in energy storage leasing within one day for the microgrid in Example 1. Detailed Implementation

[0027] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used in these embodiments have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0028] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations of the present invention.

[0029] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0030] Example 1 The inherent contradiction between privacy protection requirements and the efficient cross-layer utilization of distributed energy storage fundamentally challenges the feasibility of multi-microgrids (MGs) participating as independent economic entities in the coordinated optimization scheduling of the power system. Addressing the current situation in distribution-microgrid collaborative systems containing multi-microgrids, where a large amount of idle distributed energy storage capacity exists within each microgrid, this embodiment proposes an active distribution network-shared energy storage operator-multi-microgrid coordinated optimization scheduling method considering shared energy storage (SES) leasing. Overall, it includes: First, a shared economy framework for leasing redundant distributed energy storage in microgrids is constructed. In this framework, the Shared Energy Storage Operator (SESO) further aggregates the redundant distributed energy storage capacity of multiple microgrids on the basis of scheduling its own centralized energy storage, and virtualizes it as a dispatchable shared energy storage pool to provide peak shaving and consumption support for ADN through leasing. Under this operating framework, considering the dual uncertainties of ADN energy storage demand and the leaseable energy storage capacity of multiple microgrids, this embodiment designs a day-ahead-intraday two-stage operating strategy for SESO and introduces adjustment costs and default penalties to quantify performance risks. Furthermore, a Nash game model considering risk and privacy perception is constructed to simultaneously characterize the intraday performance risk of each participating entity and the privacy exposure cost brought about by information exchange during the interactive information decision-making and benefit distribution process, so as to achieve optimal cooperative benefits and fair distribution under the conditions of overall benefits, controllable risks and privacy constraints. To achieve distributed solution among various subjects, we propose the Alternating Directional Multiplier Method (ADMM) with consensus variables and superimposed with an event-triggered communication mechanism. By exchanging a small number of coupled variables only when the triggering condition is met, we reduce the frequency of full communication and suppress the risk of interactive sequence differential inference.

[0031] The calculation results show that the method proposed in this embodiment can improve the utilization efficiency of cross-layer energy storage resources and the level of renewable energy consumption, reduce system operating costs, and improve the performance reliability of leasing plans under uncertain scenarios, while protecting the data privacy of multiple subjects.

[0032] like Figure 1 As shown, the active distribution network-multi-microgrid collaborative optimization scheduling method includes the following steps: Establish a shared energy storage operation framework that includes active distribution networks, shared energy storage operators, and multiple microgrids and stakeholders; Considering the bilateral uncertainties of active distribution networks and multiple microgrids, adjustment costs and default penalties are introduced to quantify performance risks, and a multi-timescale operation strategy for shared energy storage operators is designed. Considering the risk perception cost and the privacy exposure risk cost, a multi-agent collaborative scheduling model based on Nash game is established, and the collaborative scheduling problem is decomposed into two sub-problems: maximizing social benefits and distributing cooperative benefits. An ADMM distributed solution algorithm based on consensus variables and event-triggered communication is introduced. The coupling relationship between the subjects is explicitly decoupled through consensus variables, and based on the event-triggered mechanism, the coupling variables are exchanged only when the triggering condition is met, and the two sub-problems are solved, realizing the collaborative optimization scheduling of multiple subjects.

[0033] This embodiment focuses on scenarios where the ADN itself does not have energy storage, there is redundant distributed energy storage within multiple microgrids, and each entity is subject to privacy protection constraints. It aims to solve the following problems: First, how to construct an economic framework under the existing power grid architecture that allows multiple microgrids to lease redundant distributed energy storage capacity to shared energy storage operators, and for these operators to uniformly provide energy storage services to the ADN, thereby achieving cross-layer aggregation and efficient utilization of centralized and distributed energy storage; Secondly, given the significant uncertainties in both ADN energy storage demand and the leaseable energy storage capacity of multiple microgrids, how can we design a multi-timescale operation strategy for SESO that takes into account both day-ahead planning and intraday adjustments, introduce adjustment costs and default penalties, and effectively manage the performance risks of energy storage lease contracts? Finally, given that ADN, SESO, and multiple microgrids are independent economic entities and are unwilling to disclose their internal operating data, how can the coordinated scheduling of the entire system and the fair distribution of cooperative benefits be achieved through information exchange of a small number of coupled variables?

[0034] The following section will provide a detailed explanation of the solution in this embodiment.

[0035] (i) Shared energy storage operation strategy considering microgrid redundant energy storage leasing and risk management.

[0036] 1.1 Shared energy storage operation strategy.

[0037] The shared energy storage operation strategy proposed in this embodiment considers the leasing of redundant energy storage in microgrids. The system comprises multiple entities, including an Active Distribution Network (ADN), microgrids, and a Shared Energy Storage Operator (SESO). No energy storage is installed within the ADN; the SESO possesses dispatchable centralized energy storage; and each microgrid contains distributed energy storage.

[0038] Each microgrid is equipped with renewable energy sources (wind power, solar power, etc.), gas turbines, and various electrical and thermal loads. Microgrids are connected to the distribution network through multiple nodes. Microgrid energy storage is primarily used to meet its own renewable energy consumption needs and smooth loads. When renewable energy generation within the microgrid is insufficient, it can purchase electricity from the ADN or discharge it through energy storage. Conversely, when renewable energy generation exceeds load demand, it can sell electricity to the ADN or charge its energy storage. After meeting the internal needs of the microgrid, the remaining capacity can be leased to the SESO to generate additional revenue. By aggregating a large amount of redundant energy storage resources within multiple microgrids, the SESO's ability to meet the ADN's energy storage needs is enhanced. By leasing redundant energy storage, the distribution network can also obtain additional energy storage supply, thereby reducing its power dependence on the main grid and reducing interaction costs with the main grid.

[0039] An Active Distribution Network (ADN) connects distributed renewable energy (DG), primarily including wind turbines (WT) and photovoltaic (PV) systems, and also incorporates traditional loads. Multiple microgrids (MGs) connect to the ADN through different nodes, and the ADN is connected to the main grid through a root node. The ADN prioritizes consuming its own renewable energy generation capacity. When renewable energy output is insufficient, it leases shared energy storage devices or purchases electricity from microgrids; when renewable energy generation is excessive, it charges energy storage devices or sells electricity to microgrids. Ultimately, if a source-load imbalance still exists, it balances the load by purchasing and selling electricity from the main grid. The operational goal of the ADN is to minimize its own operating costs.

[0040] As an intermediary between the distribution network and the microgrid, SESO possesses the capability to aggregate massive distributed energy storage resources within the microgrid. It prioritizes dispatching its own centralized energy storage to meet the distribution network's demand for renewable energy absorption and mitigation. Only when its own energy storage is insufficient does it consider leasing microgrid energy storage resources. SESO's core profit model lies in "buying low and selling high," while simultaneously leveraging its integration capabilities to participate in the leasing of ADN and microgrid energy storage resources, thereby profiting from the price difference.

[0041] 1.2 Risk management strategies for shared energy storage operators considering bilateral uncertainties.

[0042] To address the bilateral uncertainties between ADN energy storage demand and microgrid redundant energy storage supply, this embodiment proposes a multi-timescale risk management strategy suitable for SESO.

[0043] During the day-ahead phase, as previously stated, the optimization objective is to minimize costs for each participant. During the intraday phase, the optimization objective is to restore the adjusted energy storage state to the day-ahead plan as much as possible while minimizing adjustment costs and default penalties.

[0044] During intraday dispatch, SESO primarily addresses the uncertainty of ADN energy storage demand by adjusting its own energy storage capacity. If a capacity shortfall persists, SESO will adjust the energy storage capacity leased from the microgrid. The default cost for SESO in intraday optimization is the penalty for failing to meet ADN energy storage requirements.

[0045] (1)

[0046] in, This represents SESO's total intraday cost; This represents the adjustment cost of SESO. This indicates the penalty for SESO's breach of contract. Indicates the intraday time scheduling scale, taken as 15 minutes. This indicates the number of microgrids, where m represents the m-th microgrid. This refers to SESO's adjustment of the unit cost of leasing energy storage for microgrids. , These represent SESO in the time period. The adjustment amount of the leased charging / discharging power of the microgrid m; This is a penalty imposed on SESO units when they fail to meet the energy storage needs of the distribution network. Indicates the distribution network in Energy storage power demand during a given period This indicates the actual energy storage capacity that SESO provides to the distribution network; Indicates charging; This indicates a discharge.

[0047] Similarly, microgrids and ADNs must also bear corresponding intraday costs. Without considering rapid gas turbine regulation, microgrids and ADNs primarily address internal supply and demand uncertainties by adjusting demand response and the charging and discharging power of energy storage. In this context, due to their own energy storage regulation needs, microgrids may be able to fulfill day-ahead leasing plans offered to SESOs.

[0048]

[0049] in, This represents the total daily cost of microgrid m; This represents the intraday adjustment cost of microgrid m; This represents the penalty cost for microgrid m failing to lease out energy storage as scheduled; This represents the unit cost of adjusting the self-use energy storage of microgrid m. This represents the unit cost of microgrid m adjusting its demand response; where, This indicates the charging and adjustment status of the microgrid m's self-used energy storage. This indicates the discharge adjustment status of the microgrid m's self-used energy storage; This indicates the adjustment status of the microgrid m's demand response; and This indicates the day-ahead charge / discharge leasing plan for microgrid m, in which, This indicates the day-ahead charging leasing plan for microgrid m. This represents the day-ahead discharge leasing plan for microgrid m; and This indicates the intraday leasing situation of microgrid m, where, This indicates the daily charging rental situation of microgrid m. This indicates the intraday discharge leasing situation of microgrid m; M m This represents the penalty imposed on the microgrid for failing to lease out energy storage as planned.

[0050]

[0051] in, This represents the intraday cost of the distribution network. For daily adjustment costs of the distribution network; This indicates that ADN has adjusted the unit cost of leasing energy storage from SESO. This represents the unit adjustment cost of demand response; This indicates the adjustment amount for the distribution network leasing energy storage and charging plan. This indicates the adjustment amount of the energy storage discharge plan for the distribution network leasing; This indicates the amount of adjustment to the distribution network demand response plan.

[0052] (II). A collaborative scheduling model for active distribution network, multiple microgrids, and shared energy storage operators based on Nash game theory.

[0053] Allowing microgrids to lease redundant energy storage will affect the energy storage utilization rate of both the microgrid and the SESO, and will also change the energy consumption strategy of the ADN. Therefore, this embodiment proposes a distribution network-multi-microgrid collaborative scheduling strategy based on Nash game theory. This method formulates energy consumption strategies for all stakeholders by maximizing social benefits and achieves a fair distribution of benefits among participants through bargaining factors.

[0054] 2.1 Active distribution network model.

[0055] The costs of an ADN mainly include its own generation costs, demand response costs, transaction costs with microgrids, transaction costs with the main grid, and the cost of leasing shared energy storage from SESO.

[0056]

[0057] in This indicates the operating cost of an active distribution network; , , , , These represent the generation cost of the distribution network, the cost of purchasing and selling electricity with the main grid, the demand response cost, the cost of interacting with the microgrid, and the cost of leasing energy storage from SESO, respectively; m represents the microgrid number that interacts with the distribution network.

[0058] The operational constraints of ADN include power balance constraints, interaction constraints with other entities, and upper and lower limit constraints on parameters.

[0059]

[0060] in, and These represent the charging and discharging power of the energy storage leased by the ADN, respectively. and These represent the maximum values ​​of the rental charging and discharging power, respectively. and This represents a 0-1 state variable representing charging and discharging.

[0061]

[0062] in and Let represent the active power flowing into and out of node i at time t, respectively; This represents the net power exchange between node i and the microgrid at time t; and Let i represent the output power of photovoltaic and wind power at time t, respectively; and These represent the active power that the ADN purchases from the main network and sells to the main network at time t, respectively. This indicates the discharge power of the ADN. This indicates the charging power of the ADN; This represents the net active load of node i at time t. Reactive power balance can be obtained using the same method.

[0063] Not every node in the distribution network is connected to a microgrid. Here, i and j are nodes in the distribution network, and the microgrid is numbered by m.

[0064]

[0065] in, Represents a node j Voltage amplitude at time t; Represents a node i exist tVoltage amplitude at any given moment; Represents a node i and j The resistance between; Indicates the time from node t i Flow to Node j The active power; Represents a node i and j The reactance between them; express t From the node i Flow to Node j reactive power; express t Time Node i and j The current values ​​between.

[0066]

[0067] in, and These represent the upper and lower limits of the node voltage, respectively. This represents the upper limit of the line current. P buy / sell,max and Q buy / sell,max These represent the upper limits for the purchase and sale of active and reactive power, respectively. This indicates the active power purchased / sold from the main grid. This indicates the reactive power purchased / sold from the main grid.

[0068] 2.2 Microgrid Model.

[0069] The objective function of a microgrid is to minimize its operating costs (electricity purchase cost and gas turbine power generation cost), while also considering electricity sales revenue and energy storage leasing revenue.

[0070]

[0071] in: C m The total cost of operating microgrid m. This represents the interaction cost between the microgrid m and the distribution network. The gas turbine cost represents the microgrid m. This represents the revenue generated from leasing energy storage within a microgrid m.

[0072] The operational constraints of microgrids include energy storage operation and leasing constraints, demand response constraints, and gas turbine operation constraints. Among them, the demand response constraints are similar to those of distribution networks and will not be described in detail.

[0073]

[0074] in and This represents a 0-1 variable that indicates the charging and discharging state of energy storage. , These represent the total charging and discharging power of the energy storage in the microgrid m; and These represent the upper limits of the charging and discharging power of energy storage, respectively. and This represents the energy storage charging and discharging power consumed by the microgrid itself. and This indicates the charging and discharging power leased by the microgrid to the SESO.

[0075]

[0076] in, This indicates the state of charge of the energy stored in time period t+1; The state of charge of the stored energy during time period t. This requires that the state of charge of the energy storage device be equal at the beginning and end of the energy storage period. T The total number of scheduling periods. Indicates the state of charge of the energy storage during the first period. This indicates the state of charge at the end of the energy storage period. and These represent the upper and lower limits of the state of charge of the microgrid's energy storage, respectively. and These represent the energy storage charging efficiency and the discharging efficiency, respectively.

[0077] 2.3 SESO model.

[0078] SESO aims to minimize its net operating costs, including revenue from leasing energy storage to the distribution network, costs associated with leasing redundant energy storage from microgrids, and operating costs of its own centralized energy storage. Constraints include operational constraints on its own energy storage and interaction variable constraints.

[0079]

[0080] in: For the operating costs of SESO, The cost for shared energy storage operators to use their own energy storage. This refers to the net income generated during the leasing process.

[0081] The constraints of centralized energy storage in SESO can be referenced from those of microgrid energy storage. In addition, SESO also needs to meet the energy storage requirements of ADN:

[0082] In the formula, and They are respectivelyt The charging and discharging power of SESO's own energy storage during the time period. This represents the charging power leased by SESO to microgrid m. This represents the discharge power leased by SESO to microgrid m; , These represent the energy storage charging and discharging demand of the distribution network during time period t.

[0083] 2.4 Nash game model based on risk-privacy awareness.

[0084]

[0085]

[0086] Where Nn is the number of participants in the game. n Indicates the first n Number of participants and Let these represent the total cost to the nth participant when negotiations break down, and the total cost to the nth participant when negotiations succeed, taking into account risk and privacy perception. n The total cost to each participant. This represents the total cost to participant n after considering risk and privacy perception. This represents the operating cost of the nth participant. This represents the perceived risk cost for the nth participant. For privacy exposure sensitivity coefficient, This indicates the privacy breach situation of the nth participant, which aims to reduce the possibility that the temporal trajectory and change characteristics of the interaction information can be used by external entities for parameter identification and reverse inference. For the trajectory of the interaction variable, , These are the difference characteristic functions of the interaction variables. Represents the weighting coefficients of subject n with respect to the first-order temporal difference features of the interaction variables; Indicates time; This represents the weighting coefficient of subject n for the second-order temporal difference features of the interaction variables. n and n It is used to reflect the degree of temporal fluctuation and abrupt change of the interaction variable.

[0087]

[0088] in, For the scene The probability of occurrence For a set of uncertain scenarios, For the scene The intraday risk cost of subject n; The risk aversion coefficient for subject n. For confidence level of In the case of [condition], the conditional risk value of subject n for intraday costs. For the tail threshold variable, For the scene The excess costs introduced reflect the potential performance deviation losses of energy storage leasing contracts under uncertain scenarios. CVaR is used to strengthen the mitigation of default and high adjustment costs in extreme scenarios, ensuring that the leasing plans formed so far are both economically viable and performance reliable in a probabilistic sense.

[0089] Considering that pricing strategies based on Nash games involve a large number of non-convex and nonlinear problems, this problem is decomposed into two linearized subproblems: maximizing social benefits and allocating cooperative revenues.

[0090] Sub-problem 1: Maximizing social benefits:

[0091] in, , , These represent the total costs of SESO, ADN, and microgrid m, respectively, considering risk and privacy awareness.

[0092] When solving subproblem 1, the costs and benefits related to energy storage leasing for each entity are ignored.

[0093] Sub-problem 2: Profit distribution problem, i.e. pricing problem of each entity.

[0094] The optimization model for subproblem 2 is as follows: To ensure that all entities are willing to cooperate, the cost after cooperation must be lower than the cost before cooperation. The model solution is as follows:

[0095] in, This represents the total cost of ADN before the collaboration. This represents the total cost of ADN after collaboration. This represents SESO's total cost before the collaboration. This represents the total cost of SESO after the collaboration. This represents the total cost of microgrid m before the cooperation. This represents the total cost of microgrid m after the collaboration; , and These represent SESO's revenue from leasing energy storage to ADN, SESO's cost of leasing energy storage from microgrids, and SESO's net profit from leasing energy storage, respectively. , and These represent the bargaining factors for ADN, SESO, and microgrid, respectively.

[0096] It is generally believed that power suppliers have greater bargaining power. For ADN, this relates to the charging power it provides to SESO; for SESO and microgrids, it relates to the discharging power it leases.

[0097]

[0098] in, , , These represent the total supply capacity of ADN, SESO, and microgrid m, respectively. P ADN-SESO,c,t The charging power from ADN to SESO; P dis,SESO,t and P dis,lease,m,t These represent the discharge power of SESO's own energy storage and the discharge power of the microgrid leased out, respectively.

[0099] Similarly, aggregate demand situation E require The same method can also be used for calculation.

[0100] Therefore, the bargaining power of each party can be derived using the following formula:

[0101] in, For the demand situation of the distribution network, , To maximize demand and supply capacity.

[0102] Similarly, the bargaining factors for SESO and microgrids are determined using a similar method.

[0103] (III) Multi-agent distributed solution algorithm based on ADMM.

[0104] The decision-making problem regarding energy storage leasing power involves coupling among multiple stakeholders. To address the complex coupling relationships among these stakeholders, coupling variables are introduced. , , , Describe the charging / discharging leasing interactions between ADN and SESO, and between SESO and microgrid m, respectively; consensus variables , The shared energy storage interaction consistency relationship between ADN-SESO and SESO-microgrid m is represented respectively. By establishing consistency constraints on the local decision variables and consensus variables of each subject, the original multi-subject coupling constraints are explicitly decoupled. Each subject only needs to coordinate consistency around the consensus variables, thereby enabling parallel solving of local subproblems and providing a unified coordination interface for event-triggered communication and privacy-friendly interaction.

[0105]

[0106] in, This indicates the charging and discharging power that energy storage system operators can provide to the ADN after comprehensively considering their own energy storage capacity and leasing plans to the microgrid; This indicates that ADN hopes to lease energy storage charging and discharging power from SESO; This indicates that SESO hopes to lease energy storage charging and discharging power from microgrid m. microgrid m The energy storage charging and discharging power that is willing to be leased after meeting its own needs. This represents the consensus variable between ADN and SESO. This represents the consensus variable between SESO and the microgrid m.

[0107] (1) Establish the augmented Lagrangian function for each subject.

[0108] Taking ADN as an example, the augmented Lagrangian function is as follows:

[0109] in, Represents the augmented Lagrangian function of ADN. The Lagrange multiplier between ADN and SESO The penalty parameter for subproblem 1.

[0110] (2) Introduce an event-triggered communication mechanism (ET): To reduce full communication in each iteration and suppress the differential identifiability of the interaction sequence, define "published value / cached value". Communication is triggered only when the current update deviates significantly from the previous release value; otherwise, the previous release value is used.

[0111] make Representing four types of coupled variables, This indicates that ADN points to the coupling variable category of SESO. This indicates that SESO points to the coupling variable category of ADN. This indicates that the microgrid m points to the category of coupling variables in SESO. If SESO indicates the coupling variable category of microgrid m, then the event triggering criterion and published value update are as follows:

[0112]

[0113] Where k is the number of iterations, As the trigger threshold, This is the trigger indicator variable for the coupling variable x during the k-th iteration. This represents the trigger indicator variable for the coupling variable x in the (k+1)th iteration. This represents the updated value of the coupling variable x in the charging / discharging direction during time period t at the (k+1)th iteration; These are the coupling variables published up to the k-th iteration; Let L be the square of the L2 norm of the current updated value in the (k+1)th iteration and the published cached value in the kth iteration; This represents the coupling variables published up to the (k+1)th iteration.

[0114] (3) Event-triggered ADMM iteration: Each subject first solves locally, then decides whether to publish coupling variables based on event triggering, and then updates consensus variables and multipliers: 0) Initialization: Given k=0, initialize the initial values ​​of the consensus variables. Initial values ​​of Lagrange multipliers and order Set threshold Convergence accuracy And the maximum number of iterations Kmax. This represents the initial value published by the coupling variable. This represents the initial value of the coupled variable.

[0115] 1) Microgrid m Local Update:

[0116] in, This represents the update value of the charge / discharge coupling variable provided by the microgrid m to SESO during the (k+1)th iteration; Describe the augmented Lagrangian function of the microgrid m; is the Lagrange multiplier between the microgrid m and SESO.

[0117] 2) Similarly, perform local decision updates for ADN and SESO: 3) Event-triggered release: Each subject calculates according to formula (33). And update ;like If the value is 0, no new message will be sent in this round.

[0118] 4) Consensus variable update: Taking the consensus variables of ADN and SESO as an example, the update process is as follows:

[0119] in, This represents the consensus variable update value between ADN and SESO at the (k+1)th iteration; This represents the coupling variable with SESO published by the ADN side at the (k+1)th iteration; This represents the coupling variable with ADN published by the SESO side at the (k+1)th iteration; This represents the Lagrange multiplier between SESO and ADN at the k-th iteration; Let represent the Lagrange multiplier between SESO and ADN at the k-th iteration.

[0120] 5) Update the Lagrange multipliers: by For example, the update process of the Lagrange multipliers is as follows:

[0121] in, It represents the Lagrange multiplier between SESO and ADN at the (k+1)th iteration.

[0122] 6) Convergence Criterion and Iterative Update: Define the original residual and the dual residual.

[0123] in: This represents the value of the coupling variable published by ADN to SESO at the (k+1)th iteration; This represents the consensus variable between ADN and SESO at the (k+1)th iteration; This represents the value of the coupling variable published by SESO to ADN during the (k+1)th iteration; This represents the value of the coupling variable published by microgrid m to SESO at the (k+1)th iteration; This represents the consensus variable between SESO and microgrid m at the (k+1)th iteration; This represents the value of the coupling variable that SESO publishes to the microgrid m at the (k+1)th iteration; This represents the consensus variable between SESO and microgrid m in the (k+1)th iteration; This represents the consensus variable between ADN and SESO at the (k+1)th iteration; This represents the consensus variable between ADN and SESO at the k-th iteration; This represents the consensus variable between SESO and microgrid m at the (k+1)th iteration; This represents the consensus variable between SESO and microgrid m at the k-th iteration; r(k+1) is the original residual, s(k+1) is the dual residual, if and If the algorithm converges, then the algorithm is considered to have converged; otherwise, let k = k + 1 and return to step 1.

[0124] Subproblem 2 primarily addresses the distribution of benefits among the stakeholders. To ensure all stakeholders have a sufficient willingness to participate, the cost to each stakeholder after cooperation must be less than their initial cost before cooperation. The model solution process is the same as that of Subproblem 1.

[0125] (iv) Calculation example analysis.

[0126] To verify the effectiveness of the proposed method, a case study was conducted based on the improved IEEE-33-bus system. This section analyzes the corresponding simulation results.

[0127] 4.1 Example setup.

[0128] use Figure 2 The improved IEEE-33 node system shown is used as an example to verify the effectiveness of the proposed active distribution network-multi-microgrid coordinated dispatching method considering shared energy storage leasing. The time interval is 1 hour, and the dispatching cycle is 24 hours. The distribution network is normalized using reference values. S base =10MVA, U base =30kV, S base Indicates the power reference value; MVA represents megavolt-amperes. U base The voltage reference value is represented by kV, which stands for kilovolt. The upper and lower limits of the node voltage are set between 1.1 pu and 0.9 pu, where pu represents the per-unit value. The 33-node system admittance is set to 0, and the marginal network loss cost is set to 0.5. To promote the use of renewable energy, the operating costs of wind and solar power are set to 0. The microgrid adopts a time-of-use pricing strategy for purchasing electricity from the active distribution network.

[0129] exist Figure 2 The specific parameters and load distribution of the access sources are as follows: OLTC is connected between node 1 and node 2. OLTC represents an on-load tap-changing transformer. Nodes 3, 13, and 23 are connected to three microgrids, namely MG1, MG2, and MG3; SESO: Shared energy storage is located at node 17 with a storage capacity of 2000 kWh. The energy storage system has a charge / discharge efficiency of 95%, a maximum charge / discharge power of 500 kW, and a minimum charge / discharge power of 0 kW.

[0130] WT: Distributed wind turbines WT1 and WT2 are connected to nodes 16 and 26 respectively, each with an installed capacity of 1MW.

[0131] PV: Distributed photovoltaic power generation units are connected to node 16, with an installed capacity of 1.5MW.

[0132] 4.2. Comparative Cases.

[0133] 1) Case 1: Without considering the leasing of shared energy storage, each entity uses its own energy storage capacity to absorb the output of new energy and smooth out load fluctuations; 2) Case 2: The method proposed in this embodiment.

[0134] Table 1 shows the operating costs of each entity using the leasing framework proposed in this embodiment. The results show that the introduction of the shared energy storage framework improves energy storage utilization and enhances the operational efficiency of each entity.

[0135] Compared to Case 1, in Case 2, SESO can respond to distribution network time-of-use electricity prices by utilizing its own centralized energy storage and leasing microgrid energy storage, achieving low storage and high sales, thus reducing distribution network operating costs by 1.5% and microgrid operating costs by 13%. This is because SESO's leasing strategy incentivizes microgrids to lease out more energy storage capacity. By leasing this portion of microgrid energy storage, the distribution network's power dependence on the main grid is reduced, thereby lowering operating costs.

[0136] Table 1. Operating costs of each entity under different cases.

[0137]

[0138] 4.3 Algorithm convergence analysis.

[0139] Figure 3 The iterative convergence results of the main objective functions in subproblem 1 are shown. The proposed algorithm converges after 31 iterations. Figure 3 In the diagram, the horizontal axis represents the number of iterations, and the vertical axis represents the objective function value. Different colored and marked lines represent the convergence status of different objective functions. Specifically, the red line with a circle represents ADN, the green line with a square represents MG1, the blue line with a triangle represents MG2, the magenta line with a diamond represents MG3, and the black line with dots represents SESO.

[0140] Figure 4The iterative convergence results of the main objective functions in subproblem 2 are shown. The residuals converge to 10 after 49 iterations. -2 Within the range. Figure 4 In the diagram, the horizontal axis represents the number of iterations, and the vertical axis represents the objective function value. Different colored and marked lines represent the convergence status of different objective functions. Specifically, the red line with a circle represents ADN, the green line with a square represents MG1, the blue line with a triangle represents MG2, the magenta line with a diamond represents MG3, and the black line with dots represents SESO.

[0141] Figure 3 and Figure 4 This demonstrates that the distributed optimization algorithm proposed in this embodiment has good convergence characteristics and computational efficiency, and can meet the time limit requirements for power grid optimization scheduling while protecting the privacy of each subject.

[0142] 4.4 Analysis of microgrid energy storage power leasing.

[0143] Figure 5 and Figure 6 Taking the energy storage leasing and self-use of microgrid 1 as an example, this paper analyzes the effectiveness of the proposed leasing framework in improving the utilization rate of energy storage within the microgrid. Figure 5 and Figure 6 In the diagram, the horizontal axis represents time in hours (h), and the vertical axis represents power in kilowatts (kW). Blue bars represent charging services, and orange bars represent discharging services. When the self-consumption rate of the microgrid's own energy storage is low, the shared energy storage leasing framework proposed in this embodiment allows the microgrid to lease its redundant energy storage power to the SESO, and provide flexibility support to the active distribution network after unified scheduling by the SESO.

[0144] like Figure 6 As shown, the self-consumption rate of microgrid energy storage is relatively low throughout the entire dispatch cycle, with only a portion of its capacity being utilized for local renewable energy consumption during periods of high photovoltaic and wind power output. The shared energy storage leasing strategy proposed in this embodiment allows microgrids to lease out their redundant energy storage capacity, thereby improving the utilization efficiency of energy storage resources.

[0145] 4.5 Analysis of the negotiation results.

[0146] Taking discharge pricing as an example, according to the method proposed in this embodiment, Figure 7 , Figure 8 The rental prices for energy storage were displayed. Figure 7 The figure shows the change in ADN-SESO discharge price over time. Figure 8 The change in the discharge price of SESO-MG1 over time is shown.

[0147] Figure 7In the graph, the horizontal axis represents time in hours (h), and the vertical axis represents price in yuan / kWh; the black dashed line represents the main grid purchase price of electricity, and the red broken line represents the price of electricity for ADN leased SESO discharge power.

[0148] Figure 8 In the graph, the horizontal axis represents time in hours (h), and the vertical axis represents price in yuan / kWh; the black dashed line represents the main grid electricity price, and the green broken line represents the electricity price for SESO leased MG1 discharge.

[0149] The discharge price of ADN leased shared energy storage, derived by the algorithm proposed in this embodiment, is lower than the price of directly purchasing electricity from the main grid, thereby reducing the operating costs of the distribution network. It is worth noting that at some point, the leased discharge price paid by the SESO to the microgrid may exceed the revenue it would receive from leasing this discharge power to the distribution network. This is because, in the profit-sharing process, the profit-sharing mechanism requires the SESO to sacrifice some of its profits to the microgrid to ensure the microgrid has sufficient willingness to participate in cooperation. Although leasing energy storage from the microgrid may be unprofitable for the SESO at any given moment, its overall revenue over the entire dispatch cycle is still higher than before cooperation.

[0150] 4.6 Analysis of scheduling results across multiple time scales.

[0151] During the intraday dispatch phase, rapid adjustment of gas turbine output is not considered; instead, demand response and energy storage output adjustment are used to handle uncertainties on both the source and load sides. If demand response and energy storage adjustment fail to smooth source and load fluctuations, power balance is achieved by adjusting the power purchased from the main grid. Microgrid 1 is equipped only with wind turbine generators, while microgrids 2 and 3 are equipped only with photovoltaic generators.

[0152] Figure 9 The changes in MG1's self-use energy storage are shown. Figure 10 The data shows the intraday changes in MG1 discharge rental. Figure 9 In the graph, the horizontal axis represents time in hours (h), and the vertical axis represents power in kilowatts (kW). The red dashed line represents the charging status of the day before, the blue solid line represents the charging status of the day within the day, the magenta dashed line represents the discharging status of the day before, and the green solid line represents the discharging status of the day within the day.

[0153] Figure 10 In the graph, the horizontal axis represents time in hours (h), and the vertical axis represents power in kilowatts (kW). The red dashed line represents the day-ahead situation, and the blue solid line represents the intraday situation.

[0154] Figure 9Taking the self-consumption of energy storage in microgrid 1 as an example, it exhibits highly consistent adjustment strategies and behavioral patterns on both the day-ahead and intraday timescales. Specifically, microgrid 1 repeatedly experienced actual wind power generation falling below the predicted value between 2:00 and 8:00. Therefore, microgrid 1 must adjust its charging and discharging strategy during this period, switching from charging to discharging, to address the supply-demand imbalance caused by the reduced wind power generation.

[0155] like Figure 10 As shown, taking microgrid 1 as an example, the intraday and day-ahead plans show a generally consistent trend for most periods, indicating that the leased discharge of MG1 can generally be fulfilled as planned, and the lease plan formulated day-ahead has good feasibility. From the perspective of specific time periods, the actual intraday value and the planned value only deviate slightly in the morning, indicating that the system can absorb general fluctuations through its own adjustment. However, more obvious early discharge and amplitude redistribution occurred near the evening peak. That is, the actual intraday discharge first rose to about 300 kW at about 19 hours, and then fell back at about 20 hours. The day-ahead plan, on the other hand, tended to maintain a higher discharge level at 20-21 hours. This indicates that when the actual renewable energy output or load deviates from the forecast, the system will make a small-scale readjustment to the leased discharge plan of MG1 to more timely support ADN balance and renewable energy consumption.

[0156] The method proposed in this embodiment has the following technical advantages: 1) The proposed shared energy storage operation framework, which allows for the leasing of redundant energy storage in multiple microgrids, effectively reduces the overall operating cost of the system by improving the utilization rate of energy storage resources within the microgrid and reducing the power dependence of the ADN on the main grid.

[0157] 2) The risk-privacy-aware Nash game model we constructed considers intraday uncertainty risk and privacy exposure costs in collaborative scheduling, which makes the cooperation strategy more reliable in the face of bilateral uncertainty, and achieves a reasonable distribution of the benefits of multi-party cooperation under the constraints of fairness and individual incentives.

[0158] 3) The proposed ADMM distributed solution algorithm based on consensus variables and event-triggered communication requires only a small amount of boundary information to be exchanged among the subjects, and significantly reduces the number of full communication times in the iteration process through the event-triggered mechanism, thereby further suppressing the privacy risks of interactive sequence difference inference while ensuring convergence and feasibility.

[0159] Example 2 This embodiment discloses an active distribution network-multi-microgrid collaborative optimization scheduling system.

[0160] An active distribution network-multi-microgrid collaborative optimization dispatch system includes: The framework building module is configured to: build a shared energy storage operation framework that includes active distribution networks, shared energy storage operators, and multiple microgrids and multiple stakeholders; The operation strategy design module is configured to: consider the bilateral uncertainties of active distribution networks and multiple microgrids, introduce adjustment costs and default penalties to quantify performance risks, and design multi-timescale operation strategies for shared energy storage operators; The collaborative scheduling model building module is configured to: consider the risk perception cost and the privacy exposure risk cost, establish a multi-agent collaborative scheduling model based on Nash game, and decompose the collaborative scheduling problem into two sub-problems: maximizing social benefits and distributing cooperative benefits. The solution module is configured to introduce the ADMM distributed solution algorithm based on consensus variables and event-triggered communication. The consensus variables explicitly decouple the coupling relationships between the subjects, and based on the event-triggered mechanism, the coupling variables are exchanged only when the triggering conditions are met, and the two sub-problems are solved to achieve collaborative optimization scheduling of multiple subjects.

[0161] Those skilled in the art will understand that the modules or steps of the present invention described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, thereby allowing them to be stored in a storage device for execution by a computer device, or they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. The present invention is not limited to any particular combination of hardware and software.

[0162] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. An active distribution network-multi-microgrid collaborative optimization scheduling method, characterized in that, Includes the following steps: Establish a shared energy storage operation framework that includes active distribution networks, shared energy storage operators, and multiple microgrids and stakeholders; Considering the bilateral uncertainties of active distribution networks and multiple microgrids, adjustment costs and default penalties are introduced to quantify performance risks, and a multi-timescale operation strategy for shared energy storage operators is designed. Considering the risk perception cost and the privacy exposure risk cost, a multi-agent collaborative scheduling model based on Nash game is established, and the collaborative scheduling problem is decomposed into two sub-problems: maximizing social benefits and distributing cooperative benefits. An ADMM distributed solution algorithm based on consensus variables and event-triggered communication is introduced. The coupling relationship between the subjects is explicitly decoupled through consensus variables, and based on the event-triggered mechanism, the coupling variables are exchanged only when the triggering condition is met, and the two sub-problems are solved, realizing the collaborative optimization scheduling of multiple subjects.

2. The active distribution network-multi-microgrid collaborative optimization scheduling method as described in claim 1, characterized in that, Establish a shared energy storage operation framework involving multiple stakeholders, including active distribution networks, shared energy storage operators, and microgrids, specifically including: Active distribution networks do not have internal energy storage, while shared energy storage operators have dispatchable centralized energy storage, and each microgrid contains distributed energy storage. When renewable energy generation within the microgrid is insufficient, electricity is purchased from the active distribution network or discharged through distributed energy storage; when renewable energy generation within the microgrid exceeds load demand, electricity is sold to the active distribution network or charged to distributed energy storage; after meeting the internal demand of the microgrid, the remaining capacity is leased to shared energy storage operators. Active distribution networks prioritize consuming the power generated by their own renewable energy sources. When renewable energy output is insufficient, they lease shared energy storage devices or purchase electricity from microgrids. When there is a surplus of renewable energy generation, they charge energy storage devices or sell electricity to microgrids. Shared energy storage operators have the ability to aggregate distributed energy storage resources within a microgrid, prioritize dispatching their own centralized energy storage to meet the needs of the active distribution network, and only consider leasing microgrid energy storage resources when their own energy storage is insufficient.

3. The active distribution network-multi-microgrid collaborative optimization scheduling method as described in claim 1, characterized in that, Shared energy storage operators' multi-timescale operation strategy includes two phases: day-ahead planning and intraday adjustments, specifically: During the current planning phase, the optimization objective is to minimize the cost for each entity. During the intraday adjustment phase, the optimization objective is to restore the adjusted energy storage status to the day-ahead plan while minimizing adjustment costs and default penalties.

4. The active distribution network-multi-microgrid collaborative optimization scheduling method as described in claim 3, characterized in that, During the intraday adjustment phase, the adjustment costs and penalties for breach of contract for shared energy storage operators are as follows: ； ； ； in, This represents SESO's total intraday cost; This represents the adjustment cost of SESO. This indicates the penalty for SESO's breach of contract. This indicates the intraday time scheduling scale, taken as 15 minutes; This indicates the number of microgrids, where m represents the m-th microgrid. This refers to SESO's adjustment of the unit cost of leasing energy storage for microgrids. , These represent SESO in the time period. The adjustment amount of the leased charging / discharging power of the microgrid m; This is a penalty imposed on SESO units when they fail to meet the energy storage needs of the distribution network. Indicates the distribution network in Energy storage power demand during a given period This indicates the actual energy storage capacity that SESO provides to the distribution network; Indicates charging; Indicates discharge; During the intraday adjustment phase, the adjustment costs and penalties for default of microgrids are as follows: ； ； ； in, This represents the total daily cost of microgrid m; This represents the intraday adjustment cost of microgrid m; This represents the penalty cost for microgrid m failing to lease out energy storage as scheduled; This represents the unit cost of adjusting the self-use energy storage of microgrid m. This represents the unit cost of microgrid m adjusting its demand response; where, This indicates the charging and adjustment status of the microgrid m's self-used energy storage. This indicates the discharge adjustment status of the microgrid m's self-used energy storage; This indicates the adjustment status of the microgrid m's demand response; and This indicates the day-ahead charge / discharge leasing plan for microgrid m, in which, This indicates the day-ahead charging leasing plan for microgrid m. This represents the day-ahead discharge leasing plan for microgrid m; and This indicates the intraday leasing situation of microgrid m, where, This indicates the daily charging rental situation of microgrid m. This indicates the intraday discharge leasing situation of microgrid m; M m This represents the penalty imposed on the microgrid for failing to lease out energy storage as planned. During the intraday adjustment phase, the adjustment costs for the active distribution network are as follows: ； ； in, This represents the intraday cost of the distribution network. For daily adjustment costs of the distribution network; This indicates that ADN has adjusted the unit cost of leasing energy storage from SESO. This represents the unit adjustment cost of demand response; This indicates the adjustment amount for the distribution network leasing energy storage and charging plan. This indicates the adjustment amount of the energy storage discharge plan for the distribution network leasing; This indicates the amount of adjustment to the distribution network demand response plan.

5. The active distribution network-multi-microgrid collaborative optimization scheduling method as described in claim 1, characterized in that, Before establishing the multi-agent cooperative scheduling model based on Nash game, it is also necessary to establish the objective functions of each agent, where: The objective function of an active distribution network is: ； in, This indicates the operating cost of an active distribution network; , , , , These represent the generation cost of the distribution network, the cost of purchasing and selling electricity with the main grid, the demand response cost, the cost of interacting with the microgrid, and the cost of leasing energy storage from SESO, respectively; m represents the microgrid number that interacts with the distribution network. Indicates the number of microgrids; The objective function of the microgrid is: ； in, The total cost of operating microgrid m. This represents the interaction cost between the microgrid m and the distribution network. The gas turbine cost represents the microgrid m. This represents the revenue generated from leasing energy storage within a microgrid m. The objective function of a shared energy storage operator is: ； in, For the operating costs of SESO, The cost for shared energy storage operators to use their own energy storage. This refers to the net income generated during the leasing process.

6. The active distribution network-multi-microgrid collaborative optimization scheduling method as described in claim 5, characterized in that, The established multi-agent cooperative scheduling model based on Nash game is as follows: ； ； ； Where Nn is the number of participants in the game. n Indicates the first n Number of participants and Let these represent the total cost to the nth participant when negotiations break down, and the total cost to the nth participant when negotiations succeed, taking into account risk and privacy perception. n Total cost for each participant; This represents the total cost to participant n after considering risk and privacy perception. This represents the operating cost of the nth participant. This represents the perceived risk cost for the nth participant. Privacy exposure sensitivity coefficient; This indicates the privacy breach status of the nth participant; For the trajectory of the interaction variable, , These are the difference characteristic functions of the interaction variables; Represents the weighting coefficients of subject n with respect to the first-order temporal difference features of the interaction variables; Represents the weighting coefficients of subject n with respect to the second-order temporal difference features of the interaction variables; Indicates time, Indicates the total number of scheduling periods; ； ； in, For the scene The probability of occurrence For a set of uncertain scenarios, For the scene The intraday risk cost of subject n; The risk aversion coefficient for subject n. For confidence level of In the case of [the specific event], the conditional risk value of subject n for intraday costs; For the tail threshold variable, For the scene Excess costs.

7. The active distribution network-multi-microgrid collaborative optimization scheduling method as described in claim 1, characterized in that, The collaborative scheduling problem is decomposed into two sub-problems: maximizing social benefits and distributing cooperative gains. Specifically: Subproblem of maximizing social benefits: ； in, , , Let m represent the total cost of SESO, ADN, and microgrid m, respectively, considering risk and privacy awareness. This indicates the number of microgrids, where m represents the m-th microgrid; Sub-problem of distributing cooperative benefits: ； in, This represents the total cost of ADN before the collaboration. This represents the total cost of ADN after collaboration. This represents SESO's total cost before the collaboration. This represents the total cost of SESO after the collaboration. This represents the total cost of microgrid m before the cooperation. This represents the total cost of microgrid m after the collaboration; , and These represent SESO's revenue from leasing energy storage to ADN, SESO's cost of leasing energy storage from microgrids, and SESO's net profit from leasing energy storage, respectively. , and These represent the bargaining factors for ADN, SESO, and microgrid, respectively.

8. The active distribution network-multi-microgrid collaborative optimization scheduling method as described in claim 1, characterized in that, The coupling relationships between the various entities are explicitly decoupled through consensus variables, specifically including: Introducing coupling variables , , , With consensus variables , ; Through coupling variables , , , The charging / discharging leasing interactions between ADN and SESO, and between SESO and microgrid m are described respectively. Through consensus variables , The shared energy storage interaction consistency relationship between ADN-SESO and SESO-microgrid m is respectively characterized; By establishing consistency constraints between the local decision variables and consensus variables of each subject, the original multi-subject coupling constraints are explicitly decoupled. Each subject only coordinates consistency around the consensus variables, thereby solving the local subproblems in parallel. ； in, This indicates the charging and discharging power that energy storage system operators can provide to the ADN after comprehensively considering their own energy storage capacity and leasing plans to the microgrid; This indicates that ADN hopes to lease energy storage charging and discharging power from SESO; This indicates that SESO hopes to lease energy storage charging and discharging power from microgrid m. microgrid m The energy storage charging and discharging power that is willing to be leased after meeting its own needs; This represents the consensus variable between ADN and SESO. This represents the consensus variable between SESO and the microgrid m.

9. The active distribution network-multi-microgrid collaborative optimization scheduling method as described in claim 1, characterized in that, Based on the event-triggered mechanism, the coupling variables are swapped only when the triggering condition is met, and the two sub-problems are solved. The specific process includes: Establish the augmented Lagrangian functions for each subject; An event-triggered communication mechanism is introduced, which only triggers communication when the current update deviates significantly from the previous release value; otherwise, the previous release value is used. Event-triggered ADMM iteration: Each subject first solves locally, then decides whether to publish coupling variables based on event triggering, and then updates consensus variables and Lagrange multipliers until the algorithm converges. Among them, let Representing four types of coupled variables, the event triggering criteria and published value updates are as follows: ； ； in, This indicates that ADN points to the coupling variable category of SESO. This indicates that SESO points to the coupling variable category of ADN. This indicates that the microgrid m points to the category of coupling variables in SESO. SESO indicates the category of coupling variables in microgrid m; k is the iteration number. As the trigger threshold, This is the trigger indicator variable for the coupling variable x during the k-th iteration. This represents the trigger indicator variable for the coupling variable x during the (k+1)th iteration; This represents the update value of the coupling variable x in the charging / discharging direction during time period t at the (k+1)th iteration; These are the coupling variables published up to the k-th iteration; Let L be the square of the L2 norm of the current updated value in the (k+1)th iteration and the published cached value in the kth iteration; This represents the coupling variables published up to the (k+1)th iteration.

10. An active distribution network-multi-microgrid collaborative optimization dispatching system, characterized in that, include: The framework building module is configured to: build a shared energy storage operation framework that includes active distribution networks, shared energy storage operators, and multiple microgrids and multiple stakeholders; The operation strategy design module is configured to: consider the bilateral uncertainties of active distribution networks and multiple microgrids, introduce adjustment costs and default penalties to quantify performance risks, and design multi-timescale operation strategies for shared energy storage operators; The collaborative scheduling model building module is configured to: consider the risk perception cost and the privacy exposure risk cost, establish a multi-agent collaborative scheduling model based on Nash game, and decompose the collaborative scheduling problem into two sub-problems: maximizing social benefits and distributing cooperative benefits. The solution module is configured to introduce the ADMM distributed solution algorithm based on consensus variables and event-triggered communication. The consensus variables explicitly decouple the coupling relationships between the subjects, and based on the event-triggered mechanism, the coupling variables are exchanged only when the triggering conditions are met, and the two sub-problems are solved to achieve collaborative optimization scheduling of multiple subjects.

Images