A distributed energy scheduling method for optical storage charging station based on double-layer game

Abstract

The application discloses a kind of distributed light storage charging station energy scheduling methods based on double-layer game, including as follows: constructing distributed light storage charging station system, defining optimization goal;Establish upper Stackelberg game model;Establish lower layer Nash bargaining game model, upper Stackelberg game model is converted into optimization problem;Based on the distributed solution strategy of neurodynamics algorithm, each light storage charging station is only exchanged auxiliary variable with neighbor node as agent, realizes global collaborative optimization under the premise of not sharing decision variable;Agent adjusts the iteration step of neurodynamics algorithm by multi-armed tiger machine self-adaptive step algorithm;Based on double-layer optimization algorithm, upper Stackelberg game and lower layer Nash bargaining game are solved, leader strategy and follower strategy are updated, until Stackelberg equilibrium is reached, and the optimal strategy is output.The application optimizes the total income of light storage charging station, reduces the impact of load on power system.

Description

Technical Field

[0001] This invention relates to the field of energy scheduling for photovoltaic-storage-charging stations, and more particularly to a distributed energy scheduling method for photovoltaic-storage-charging stations based on a two-layer game theory approach. Background Technology

[0002] With the increasing severity of the global energy crisis and environmental pollution, electric vehicles (EVs), as an important component of green transportation, have experienced explosive growth in their fleet size. However, the unregulated charging behavior of large-scale EVs is random and volatile. Direct connection to the power grid would lead to a greater load peak-valley difference, severely impacting the stable operation of the distribution network. Photovoltaic-storage charging stations integrate photovoltaic systems, energy storage systems, and charging facilities, enabling the local consumption of clean energy through integrated intelligent scheduling, thus alleviating grid pressure.

[0003] To further improve operational efficiency, multiple geographically dispersed distributed photovoltaic-storage-charging stations often form cooperative alliances to address fluctuations in photovoltaic output and uncertainties in charging demand through energy sharing. However, in actual operation, this model faces the following main challenges: First, there is a conflict of interest between distribution network operators and photovoltaic-storage charging station alliances. Distribution networks aim to maintain grid stability by guiding load through price signals, achieving "peak shaving and valley filling" (peak shaving: during peak electricity consumption periods, price signals or dispatch strategies guide photovoltaic-storage charging stations to reduce charging or even discharging, thereby reducing the peak load on the grid and avoiding problems such as line overload and voltage instability; valley filling: during off-peak electricity consumption periods, encouraging photovoltaic-storage charging stations to charge and store more energy, "storing" the off-peak electricity, improving the utilization rate of grid equipment, and avoiding waste of generation and transmission resources), while photovoltaic-storage charging stations pursue maximizing their own operating profits. Existing research often focuses on one-sided optimization and lacks a two-layer game mechanism that can simultaneously balance grid stability and charging station profits.

[0004] Secondly, within the photovoltaic-storage charging station alliance, the various sites often belong to different stakeholders. Achieving a fair distribution of benefits through cooperation is crucial for maintaining the alliance's stability. While traditional centralized optimization methods can achieve global optimization, they require collecting detailed photovoltaic, energy storage, and load data from each site, posing serious risks of privacy breaches and single points of failure.

[0005] Finally, for distributed optimization problems, existing solution algorithms typically rely on a fixed iteration step size. A step size that is too small leads to slow convergence, failing to meet real-time scheduling requirements; a step size that is too large may cause system oscillations or even non-convergence. Currently, there is a lack of an efficient distributed solution strategy that can adaptively adjust the step size to accelerate convergence while protecting privacy.

[0006] Therefore, there is still a lack of a distributed photovoltaic-storage-charging station energy scheduling method that can coordinate the interests of the distribution network and charging stations, ensure fairness and privacy within the alliance, and has rapid convergence capability. Summary of the Invention

[0007] In view of this, the purpose of this invention is to propose a distributed photovoltaic-storage-charging station energy scheduling method based on two-layer game theory, which optimizes the total revenue of the photovoltaic-storage-charging station and reduces the impact of load on the power system.

[0008] To achieve the above-mentioned technical objectives, the technical solution adopted by this invention is as follows: This invention provides a distributed photovoltaic-storage-charging station energy scheduling method based on a two-layer game theory approach, comprising the following steps: Step 1: Construct a distributed photovoltaic-storage charging station system, including the power distribution network and a photovoltaic-storage charging station alliance composed of multiple photovoltaic-storage charging stations, and define the optimization objectives of the photovoltaic-storage charging station system; Step 2: Based on the optimization objective, establish an upper-level Stackelberg game model consisting of the distribution network operator as the leader and the photovoltaic-storage-charging station alliance as the follower. Step 3: Within the photovoltaic-storage-charging station alliance, establish a lower-level Nash bargaining game model and transform the upper-level Stackelberg game model into an optimization problem. Achieve fair cooperation among the photovoltaic-storage-charging stations by maximizing cooperative benefits and distributing benefits. Step 4: Based on the distributed solution strategy of the neurodynamics algorithm, each photovoltaic-storage-charging station acts as an agent and only exchanges auxiliary variables with its neighboring nodes, so as to achieve global collaborative optimization without sharing decision variables. Step 5: In each iteration, each agent adjusts the iteration step size of the neurodynamics algorithm through the multi-armed slot machine adaptive step size algorithm to accelerate the convergence of the neurodynamics algorithm; Step 6: A distributed two-level optimization algorithm based on the multi-armed slot machine adaptive step size algorithm-neurodynamics algorithm is used to solve the upper-level Stackelberg game and the lower-level Nash bargaining game through nested inner and outer layers. The leader strategy and follower strategy are alternately updated in each time period until the Stackelberg equilibrium is reached and the optimal strategy is output.

[0009] Furthermore, step 1 specifically includes: Step 11: The distributed photovoltaic-storage charging station system consists of a power distribution network and a photovoltaic-storage charging station alliance composed of multiple photovoltaic-storage charging stations; Step 12: The power distribution network serves as the external power source for the photovoltaic-storage charging station, supplying power to the station unidirectionally. Its operation must meet multiple constraints, including total power purchase constraints, electricity price boundary constraints, and global balance constraints. Step 13: The photovoltaic-energy storage charging station consists of a photovoltaic system, an energy storage system, a charging pile system, and an energy management system; specifically as follows: Step 131: The operation mechanism of the photovoltaic system is as follows: the energy generated by the photovoltaic system is first delivered to the charging pile system for consumption, the surplus is stored in the energy storage system or sold to other photovoltaic-storage charging stations, and the excess energy that cannot be consumed is discarded. Step 132: The energy storage system provides energy regulation for the photovoltaic-energy storage charging station; Step 133: The charging pile system serves as the interface between the photovoltaic energy storage charging station and the electric vehicle. Step 134: The energy management system is responsible for communicating with the external power distribution network, other photovoltaic-storage charging stations and electric vehicles to formulate energy dispatch strategies with the goal of maximizing the operating revenue of photovoltaic-storage charging stations, and send them to the energy storage system and charging pile system for execution. Step 14: In the scheduling of the distributed photovoltaic-storage-charging station system, the optimization objective of the distribution network operator is to minimize power fluctuations, and the optimization objective of the photovoltaic-storage-charging station alliance is to maximize the total operating revenue of the alliance. Based on the optimization objectives of the distribution network operator and the photovoltaic-storage-charging station alliance, the scheduling problem of the distributed photovoltaic-storage-charging station system is transformed into an optimization objective.

[0010] Furthermore, step 11 specifically includes: definition One photovoltaic energy storage charging station is ,in, This indicates the first photovoltaic-storage charging station. This indicates the second photovoltaic-storage charging station. Indicates the first i One photovoltaic energy storage charging station, Indicates the first One photovoltaic energy storage charging station; the scheduling time period set T is divided into... A period of time, namely ; The total power purchase constraint, electricity price boundary constraint, and global balance constraint in step 12 are as follows: (a) Total power purchase constraint: Total power purchased by the photovoltaic-storage-charging station alliance during the period The sum of the purchased power of each photovoltaic-storage charging station is expressed as: (1) in, Photovoltaic and energy storage charging stations exist Electricity purchase capacity during the time period This represents the maximum power that the distribution network can supply. (b) Electricity price boundary constraints: Based on the time-of-use pricing, the distribution network operator releases a dynamic electricity price according to the current load to guide the photovoltaic-storage charging station to adjust its power purchase capacity; Dynamic electricity price of time-of-use distribution network Represented as: (2) in, for Time-of-use pricing for distribution network operators This is the electricity price adjustment coefficient. The average power purchase capacity of photovoltaic-storage charging stations. For the minimum electricity price, The maximum electricity price; (c) Global Balance Constraint: The energy exchange power of each photovoltaic-storage-charging station in the distribution network must satisfy the global balance constraint, expressed as: (3) in, In order to be in Time-of-use photovoltaic energy storage charging station The energy exchange power between other photovoltaic and energy storage charging stations; a positive value indicates that energy is being transferred to other photovoltaic and energy storage charging stations, while a negative value indicates that energy is being received from other photovoltaic and energy storage charging stations. Step 131 specifically includes: Photovoltaic and energy storage charging stations exist Actual photovoltaic power consumed during the period Determined by supply and demand balance, expressed as: (4) in, In order to be in Time-of-use photovoltaic energy storage charging station The actual output power of the photovoltaic system In order to be in Time-of-use photovoltaic energy storage charging station The total charging power demand of electric vehicles in China In order to be in Time-of-use photovoltaic energy storage charging station The charging and discharging power of the energy storage system is negative when it indicates charging and positive when it indicates discharging. In order to be in Time-of-use photovoltaic energy storage charging station The power purchased from the distribution network; Step 132 specifically includes: exist Time-of-use photovoltaic energy storage charging station The charging and discharging power of medium energy storage system To optimize the variables, it is represented as: (5) in, In order to be in Time-of-use photovoltaic energy storage charging station The energy storage charge-discharge coefficient, Photovoltaic and energy storage charging stations Maximum charging power, Photovoltaic and energy storage charging stations Maximum discharge power; exist During the period, photovoltaic and energy storage charging stations State of charge of medium energy storage system The formula is: (6) in, and Photovoltaic and energy storage charging stations Upper and lower limits of the state of charge of medium-capacity energy storage systems Photovoltaic and energy storage charging stations Medium energy storage battery capacity, The length of a time period; The energy conversion coefficient of the energy storage system during the charging and discharging process is expressed as: (7) in, These refer to the charging efficiency and discharging efficiency of the energy storage system, respectively. The energy storage system during charging and discharging Time-of-use photovoltaic energy storage charging station Energy storage loss costs Represented as: (8) in, This is the energy storage loss cost coefficient; Step 133 specifically includes: exist Time-of-use photovoltaic energy storage charging station The total charging power supplied to electric vehicles by the charging pile system , represented as: (9) in, In order to be in Time-of-use photovoltaic energy storage charging station The total charging power required by electric vehicles in China is expressed as: (10) in, In order to be in Time-of-use photovoltaic energy storage charging station The increased demand for charging power for electric vehicles in China; In the first Time-of-use photovoltaic energy storage charging station Total charging power demand for electric vehicles in China; In the first Time-of-use photovoltaic energy storage charging station The actual charging power supplied to electric vehicles by the charging pile system in China; Step 134 specifically includes: exist Time-of-use photovoltaic energy storage charging station Power purchased from the distribution network Represented as: (11) exist Time-of-use photovoltaic energy storage charging station Self-operated revenue Represented as: (12) in, In order to be in Time-of-use photovoltaic energy storage charging station The price of charging services, In order to be in Dynamic electricity pricing for time-of-use distribution networks; Step 14 specifically includes: Step 141, in Negative effects of time-of-use distribution network operators for: (13) Step 142, in The effectiveness of the photovoltaic-storage-charging station alliance for: (14) in, for Time-of-use photovoltaic energy storage charging station In the photovoltaic-storage charging station alliance, the transfer payments used for revenue distribution, when When, it means Obtaining energy from other photovoltaic and energy storage charging stations in the photovoltaic and energy storage charging station alliance requires paying a fee. When, it means There is no charge for transmitting energy to other photovoltaic and energy storage charging stations in the photovoltaic and energy storage charging station alliance. Step 143, the optimization objective of the distribution network operator Represented as: (15) in, For a given Time period At that time, the total power purchased will maximize the effectiveness of the photovoltaic-storage charging station alliance. Step 144: Optimization objective of the photovoltaic-storage-charging station alliance Represented as: (16) in, For a given Time period The dynamic electricity price that minimizes the negative utility for distribution network operators.

[0011] Furthermore, the construction process of the upper-level Stackelberg game model in step 2 includes: Step 21: Model the optimization objective as a Stackelberg game process, for each time period... The upper-level Stackelberg game model is represented as follows: The specific definitions of each element are as follows: (a) Player: As a leader, namely the distribution network operator; As a follower, namely the Photovoltaic-Storage-Charging Station Alliance; (b) Strategy: For leaders' strategies, namely Dynamic electricity price of time-of-use distribution network In the In the round of iteration, Updated by the following formula: (17) For follower strategy, i.e. Total power purchased by the photovoltaic-storage charging station alliance In the photovoltaic-storage charging station alliance, each photovoltaic-storage charging station makes collaborative decisions. (c) Utility: The negative utility of the leader is to minimize load fluctuations; For the utility of followers, that is, to maximize the total revenue of the alliance operation; Step 22, in the upper-level Stackelberg game model In the middle, strategy group A Stackelberg equilibrium point is formed if and only if the following condition is satisfied: (a) The leader's subgame Nash equilibrium, i.e. (18); (b) The subgame Nash equilibrium of the follower, i.e. (19).

[0012] Furthermore, the construction process of the lower-level Nash bargaining game model in step 3 includes: Step 31: Model the cooperation in the photovoltaic-storage-charging station alliance using asymmetric Nash bargaining. For each time period... The lower-level Nash bargaining game model is expressed as follows: The specific definitions of each element are as follows: (a) Player: For players to gather, namely, a collection of photovoltaic and energy storage charging stations; (b) Point of divergence: For the player's branching points; in Time period The point of contention in collaboration, i.e., the maximum benefit when operating independently. , represented as: (20) (c) Bargaining power: This represents the collective bargaining power of the players. Time period Bargain coefficient in cooperation Represented as: (twenty one) (d) Strategy: For the player's strategy set, among which, for Time period The strategy group; (e) Utility: For the player's utility set, Time period The utility Defined as: (twenty two) Step 32, for the photovoltaic-storage charging station alliance The asymmetric Nash bargaining problem is formalized using the Nash product, resulting in the optimization problem of the lower-level Nash bargaining game model. , represented as: (twenty three) st(1)-(12),(20)-(22) Step 33: Based on Nash's axioms, the optimization problem... Decomposed into subproblems of maximizing cooperative benefits and the problem of income distribution , respectively represented as: (twenty four) (25) (20), (21) in, For cooperation Time period The optimal self-operating revenue is determined by The solution is obtained; A collection of photovoltaic energy storage and charging stations where energy exchange exists; Step 34: Maximizing the subproblem of cooperative benefits Income distribution sub-problem The formal transformation is performed as follows: (a) The subproblem of maximizing the benefits of cooperation Converted to standard form, it is expressed as: (26) in, for The operating revenue of the entire photovoltaic-storage charging station alliance during the period. for Time period Decision variables, for Time period The linear equality constraint matrix, for The vector of constant terms constrained by the linear equality of the time period. for Time period Inequality constraint functions; (b) Sub-problem of revenue allocation Converted to logarithmic form, it can be expressed as: (27) (20), (21) Step 35: Construct the Lagrange function, expressed as: (28) in, For Lagrange multipliers; For transfer payments and Lagrange multipliers The Lagrangian function of the variable; Due to the conditions of Karush-Kun-Tucker, ,get: (29) because and ,get ,in, for Time period The optimal operating revenue for the entire photovoltaic-storage-charging station alliance; Step 36: Substitute the result of formula (29) back into formula (25) to obtain... The analytical solution is: (30) in, for for Time period Optimal transfer payments within a cooperative alliance.

[0013] Furthermore, the distributed solution strategy based on the neurodynamics algorithm in step 4 specifically includes: Step 41: In distributed neurodynamics, each photovoltaic-storage-charging station is modeled as an agent composed of a recurrent neural network, whose internal variables include decision variables and auxiliary variables; each agent only exchanges auxiliary variables with neighboring nodes in the topology network and does not share decision variables. The internal variables of the agent are defined as follows: (b) Decision variables : For photovoltaic and energy storage charging stations exist The decision variable for a given time period does not need to be shared with other agents; (b) Auxiliary variables , , and Used for exchanging data between neighboring nodes to help the photovoltaic-storage-charging station alliance reach a consensus; Photovoltaic and energy storage charging stations exist The Lagrange multipliers for the time interval correspond to the dual variables of the equality constraints; Photovoltaic and energy storage charging stations exist The Lagrange multipliers for the time interval correspond to the dual variables of the inequality constraints; Photovoltaic and energy storage charging stations exist Time periods are used as auxiliary variables to convey information related to consistency constraints, promoting consensus among multipliers; Photovoltaic and energy storage charging stations exist Additional auxiliary variables for the time period are used to facilitate consensus on inequality multipliers between adjacent nodes; Step 42: Iteratively update the decision variables and auxiliary variables using discretized neurodynamic equations until the convergence condition is met; (a) Decision variables of the agent The update equation is expressed as: (31) in, For the first Decision variables in round iteration , For the first Decision variables in round iteration , In order to be in Photovoltaic and energy storage charging stations during the time period The The step size of the round iteration, For projection operators over the feasible region, The gradient of the site's own operating revenue function. This is the equality constraint matrix. This is an inequality constraint function; For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration ; (b) the auxiliary variables The update equation is expressed as: (32) in, For the first Auxiliary variables of round iteration , for Neighboring nodes in a topological network for The set of neighboring nodes in a topological network; Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration, For the first Auxiliary variables of round iteration , Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration; (c) the auxiliary variables The update equation is expressed as: (33) in, For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration , As a dual variable, it is used to accelerate the auxiliary variable. The convergence process; (d) The auxiliary variables The update equation is expressed as: (34) in, For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration , Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration, Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration, For operators projecting onto non-negative quadrants; (e) the auxiliary variables The update equation is expressed as: (35) in, For the first Auxiliary variables of round iteration ; (f) The convergence condition of the neural dynamics algorithm is measured by the change in decision variables in the form of a sliding window, and is expressed as: (36) in, To determine the convergence measure, the number of rounds is used, i.e., the length of the sliding window. Index for iteration count, In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration This is the preset convergence accuracy threshold.

[0014] Furthermore, the adaptive step-size algorithm based on multi-armed slot machines in step 5 specifically includes: Step 51, Initialization Phase: Initialize each Instant rewards Changes in decision variables With each average reward and the number of times selected ;in, For the collection of photovoltaic and energy storage charging stations, For the set of step-size strategies, For iteration rounds; Step 52, Reward Calculation: Based on the... Calculating the change in decision variables in each iteration Then calculate the instant reward. The formula is as follows:

[0015]

[0016] in, for During the period The The immediate reward for each iteration step is used to evaluate the contribution of the current iteration step to the convergence of the neurodynamics algorithm. for During the period No. The change in decision variables during rounds of iteration. for During the period No. The change in decision variables during each iteration; In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration; Step 53, Average Reward Update: Update instant rewards Used to update the Average reward of the selected action in the round The formula is as follows:

[0017] in, for During the period In the Selecting actions during round iteration The average reward for During the period In the Actions during round iteration The average reward; for During the period No. Actions during round iteration The number of times it was selected; Step 54, Action Selection: Based on the number of times all actions have been selected. and average reward The confidence upper limit strategy is used to select the action for this iteration. The formula is as follows:

[0018] in, for Photovoltaic and energy storage charging stations during the period No. The step size calculation strategy for each iteration. To explore the weighting coefficients; Step 55, Step Length Calculation: Based on the selected action The corresponding step size calculation formula determines the step size for this iteration. ; like For fixed step size ,but ;in, This is a preset fixed step size value; like Cosine annealing step ,but ;in, For the maximum iteration step size, To be the minimum iteration step size, This represents the maximum number of iterations. like linear decay step size ,but ; Step 56, Update Selection Count: Increment the selection count of the selected action by 1, i.e. ;in, for During the period No. Actions during round iteration Number of times selected for During the period No. Actions during round iteration The number of times it was selected; Step 57, Iteration Loop: Repeat steps 52 to 56 until the maximum number of iterations is reached. Output iteration step size .

[0019] Furthermore, the specific process of the distributed two-layer optimization algorithm in step 6 is as follows: Step 61, Initialization Phase: Set the scheduling time set T and initialize the leader policy. Follower strategy Smart agent strategy for photovoltaic-storage charging stations Decision variables and auxiliary variables of the photovoltaic-storage-charging station intelligent agent, as well as the iteration step size. ; Step 62, Outer Loop: For each time period Optimize sequentially; Step 63, Leader Initialization: At the beginning of each period, the leader provides an initial strategy. ; Step 64, Inner Loop: All photovoltaic-storage-charging station agents execute the neural dynamics algorithm and the multi-armed slot machine adaptive step size algorithm in parallel to solve the response strategy under the current electricity price; Step 65: Strategy Update: Update the agent strategy for each photovoltaic-storage-charging station based on the inner layer solution results. and follower strategy ; Step 66: Leader Update: The leader adjusts its strategy according to the followers' response strategies and the update rules of the Stackelberg game. ; Step 67, Convergence Judgment: Iteration Rounds Repeat steps 64 to 66 until the leader strategy meets the convergence condition. Furthermore, the follower strategy satisfies the convergence condition. This achieves the Stackelberg equilibrium; among which, In order to be in Time period Leader strategy during round iteration In order to be in Time period Leader strategy during round iteration In order to be in Time period Follower strategy during round iteration In order to be in Time period Follower strategy during round iteration; Step 68, Time Period Cycle: Enter the next time period and repeat steps 63 to 67 until all time periods have been optimized, and output the optimal strategy group. ,in, The optimal strategy for leaders. The optimal strategy for followers The optimal strategy for the intelligent agent of the photovoltaic-storage charging station.

[0020] Furthermore, step 64 specifically includes: Step 641, Order ; Step 642: Call the multi-armed slot machine adaptive step size algorithm to update the iteration step size. ; Step 643: Update the decision variables according to formula (31) ; Step 644: Update the auxiliary variables according to formulas (32) and (35). , ,、 and ; Step 645 With neighboring nodes in the topology network Communication, exchanging auxiliary variables; Step 646, Iteration Rounds ; Step 647: Repeat steps 642 to 646 until the inner convergence condition is met. Or reach the maximum number of iterations ,in, Index for iteration count, In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration In order to be in Time-of-use photovoltaic energy storage charging station In the -1 round of iterations of decision variables.

[0021] Furthermore, the output of the distributed two-layer optimization algorithm includes: (a) Optimal dynamic electricity price for each time period; (b) The optimal power purchase, charging and discharging power of the energy storage system, and energy exchange power between the photovoltaic and energy storage charging stations at each time period; (c) Optimal transfer payments and optimal operating revenue for each photovoltaic-storage-charging station at each time period.

[0022] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art: The purpose of this invention is to address the energy scheduling problem of distributed photovoltaic-storage-charging stations under privacy protection. Combining the profitability requirements of photovoltaic-storage-charging stations with the stability requirements of the power system, it proposes an optimal energy scheduling strategy and a pricing strategy for distribution network operators. To this end, a distributed energy scheduling strategy based on a multi-armed slot machine-neurodynamics algorithm is proposed. This strategy aims to balance the stability of the distribution network and the revenue of the photovoltaic-storage-charging station alliance by solving a two-level game model, achieving a fair distribution of benefits within the alliance. The proposed distributed algorithm achieves rapid solution while protecting privacy.

[0023] To achieve the aforementioned objectives, the technical solution of this invention is as follows: A distributed photovoltaic-storage-charging station system model is established; an upper-level Stackelberg game model is constructed between the distribution network operator and the photovoltaic-storage-charging station alliance; and a lower-level Nash bargaining game model is constructed within the photovoltaic-storage-charging station alliance. Based on this, a distributed energy dispatching strategy based on a multi-armed slot machine-neurodynamics algorithm is proposed. This invention solves the problem of cooperation among distributed photovoltaic-storage-charging stations under privacy protection, optimizes the total revenue of photovoltaic-storage-charging stations, and reduces the impact of load on the power system. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is an execution flowchart of a distributed photovoltaic-storage-charging station energy scheduling method based on two-layer game theory provided in an embodiment of the present invention.

[0026] Figure 2 This is a framework diagram of a distributed photovoltaic energy storage and charging station system with privacy protection as provided in an embodiment of the present invention.

[0027] Figure 3 This is a diagram of the internal network structure of the intelligent agent RNN provided in an embodiment of the present invention. Detailed Implementation

[0028] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be particularly noted that the following embodiments are for illustrative purposes only and do not limit the scope of the invention. Similarly, the following embodiments are only some, not all, embodiments of the present invention, and all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] Please see Figures 1-3The present invention provides a distributed photovoltaic-storage-charging station energy scheduling method based on a two-layer game theory, comprising the following steps: Step 1: Construct a distributed photovoltaic-storage charging station system, including the power distribution network and a photovoltaic-storage charging station alliance composed of multiple photovoltaic-storage charging stations, and define the optimization objectives of the photovoltaic-storage charging station system; In this embodiment, step 1 specifically includes: Step 11: The distributed photovoltaic-storage charging station system consists of a power distribution network and a photovoltaic-storage charging station alliance composed of multiple photovoltaic-storage charging stations, such as... Figure 2 As shown, the Photovoltaic-Storage Charging Station Alliance is a virtual platform for distributed photovoltaic-storage charging stations to cooperate, enabling them to exchange energy and negotiate prices. To protect the privacy of operational information, photovoltaic-storage charging stations only need to exchange anonymized information with neighboring nodes to achieve global optimization. Based on this, the alliance aggregates the electricity purchase demands of distributed photovoltaic-storage charging stations and interacts uniformly with distribution network operators. Within the alliance framework, photovoltaic-storage charging stations cooperate only through anonymized information and utilize the physical infrastructure of the distribution network for energy exchange; specifically including: definition One photovoltaic energy storage charging station is ,in, This indicates the first photovoltaic-storage charging station. This indicates the second photovoltaic-storage charging station. Indicates the first i One photovoltaic energy storage charging station, Indicates the first One photovoltaic energy storage charging station; the scheduling time period set T is divided into... A period of time, namely It is assumed that the power of each part of the system is approximately constant during each time period, and energy loss during transmission is ignored.

[0030] Step 12: The power distribution network serves as the external power source for the photovoltaic-storage charging station, supplying power unidirectionally to the station. Its operation must meet several constraints, including total power purchase constraints, electricity price boundary constraints, and global balance constraints; specifically as follows: (a) Total power purchase constraint: Total power purchased by the photovoltaic-storage-charging station alliance during the period The sum of the purchased power of each photovoltaic-storage charging station is expressed as: (1) in, Photovoltaic and energy storage charging stations exist Electricity purchase capacity during the time period This represents the maximum power that the distribution network can supply. (b) Electricity price boundary constraints: In order to mitigate the impact of load peak-valley differences on the power system, distribution network operators issue dynamic electricity prices based on the current load on the basis of time-of-use pricing, guiding photovoltaic and energy storage charging stations to adjust their power purchase capacity; Dynamic electricity price of time-of-use distribution network Represented as: (2) in, for Time-of-use pricing for distribution network operators This is the electricity price adjustment coefficient. The average power purchase capacity of photovoltaic-storage charging stations. For the minimum electricity price, The maximum electricity price; (c) Global Balance Constraints: Furthermore, the distribution network provides the physical infrastructure for energy exchange between photovoltaic-storage-charging stations. The energy exchange power of each photovoltaic-storage-charging station in the distribution network must satisfy global balance constraints, expressed as: (3) in, In order to be in Time-of-use photovoltaic energy storage charging station The energy exchange power between other photovoltaic and energy storage charging stations; a positive value indicates that energy is being transferred to other photovoltaic and energy storage charging stations, while a negative value indicates that energy is being received from other photovoltaic and energy storage charging stations. Step 13: The photovoltaic-energy storage charging station is the main body of the system scheduling, and consists of a photovoltaic system, an energy storage system, a charging pile system, and an energy management system; specifically as follows: Step 131: The photovoltaic system provides free clean energy to the photovoltaic-storage charging station. To maximize the utilization rate of clean energy and reduce the occurrence of photovoltaic curtailment, the photovoltaic system operates as follows: energy generated by the photovoltaic system is prioritized for consumption by the charging pile system; surplus energy is stored in the energy storage system or sold to other photovoltaic-storage charging stations; and excess energy that cannot be consumed is discarded. Specifically, this includes: Photovoltaic and energy storage charging stations exist Actual photovoltaic power consumed during the period Determined by supply and demand balance, expressed as: (4) in, In order to be in Time-of-use photovoltaic energy storage charging station The actual output power of the photovoltaic system In order to be in Time-of-use photovoltaic energy storage charging station The total charging power demand of electric vehicles in China In order to be in Time-of-use photovoltaic energy storage charging station The charging and discharging power of the energy storage system is negative when it indicates charging and positive when it indicates discharging. In order to be in Time-of-use photovoltaic energy storage charging station The power purchased from the distribution network; Step 132: The energy storage system provides energy regulation for the photovoltaic-energy storage charging station, reducing the operating costs of the station while helping the distribution network achieve peak shaving and valley filling to reduce the impact of EV charging load on the power system. Specifically, this includes: exist Time-of-use photovoltaic energy storage charging station The charging and discharging power of medium energy storage system To optimize the variables, it is represented as: (5) in, In order to be in Time-of-use photovoltaic energy storage charging station The energy storage charge-discharge coefficient, Photovoltaic and energy storage charging stations Maximum charging power, Photovoltaic and energy storage charging stations Maximum discharge power; exist During the period, photovoltaic and energy storage charging stations State of Charge (SOC) of a medium-sized energy storage system The formula is: (6) in, and Photovoltaic and energy storage charging stations Upper and lower limits of the state of charge of medium-capacity energy storage systems Photovoltaic and energy storage charging stations Medium energy storage battery capacity, The length of a time period; The energy conversion coefficient of the energy storage system during the charging and discharging process is expressed as: (7) in, These refer to the charging efficiency and discharging efficiency of the energy storage system, respectively. During the charging and discharging process of an energy storage system, the performance of the energy storage battery will degrade with the increase of usage cycles. The energy storage system during charging and discharging... Time-of-use photovoltaic energy storage charging station Energy storage loss costs Represented as: (8) in, This is the energy storage loss cost coefficient; Step 133: The charging pile system serves as the interface between the photovoltaic energy storage charging station and the electric vehicle; specifically, it includes: Limited by the energy supply capacity of power distribution networks, photovoltaic systems, and energy storage systems, Time-of-use photovoltaic energy storage charging station The total charging power supplied to electric vehicles by the charging pile system , represented as: (9) in, In order to be in Time-of-use photovoltaic energy storage charging station The total charging power required by electric vehicles in China is expressed as: (10) in, In order to be in Time-of-use photovoltaic energy storage charging station The increased demand for charging power for electric vehicles in China; In the first 1-period photovoltaic-storage charging station Total charging power demand for electric vehicles in China; In the first 1-period photovoltaic-storage charging station The actual charging power supplied to electric vehicles by the charging pile system in China; Step 134: The energy management system is responsible for communicating with the external power distribution network, other photovoltaic-storage charging stations, and electric vehicles to formulate energy dispatch strategies with the goal of maximizing the operating revenue of photovoltaic-storage charging stations, and then sending these strategies to the energy storage system and charging pile system for execution; specifically including: exist Time-of-use photovoltaic energy storage charging station Power purchased from the distribution network Represented as: (11) exist Time-of-use photovoltaic energy storage charging station Self-operated revenue Represented as: (12) in, In order to be in Time-of-use photovoltaic energy storage charging station The price of charging services, In order to be in Dynamic electricity pricing for time-of-use distribution networks; Step 14: In the scheduling of the distributed photovoltaic-storage charging station system, the optimization objective of the distribution network operator is to minimize power fluctuations in order to maintain the stable operation of the distribution network; the optimization objective of the photovoltaic-storage charging station alliance is to maximize the total operating revenue of the alliance; based on the optimization objectives of the distribution network operator and the photovoltaic-storage charging station alliance, the scheduling problem of the distributed photovoltaic-storage charging station system is transformed into optimization objectives; specifically including: Step 141, in Negative effects of time-of-use distribution network operators for: (13) Step 142, in The effectiveness of the photovoltaic-storage-charging station alliance for: (14) in, for Time-of-use photovoltaic energy storage charging station Transfer payments used for revenue distribution within the photovoltaic-storage-charging station alliance are negotiated within the alliance. When, it means Obtaining energy from other photovoltaic and energy storage charging stations in the photovoltaic and energy storage charging station alliance requires paying a fee. When, it means There is no charge for transmitting energy to other photovoltaic and energy storage charging stations in the photovoltaic and energy storage charging station alliance. Step 143, the optimization objective of the distribution network operator Represented as: (15) in, For a given Time period At that time, the total power purchased will maximize the effectiveness of the photovoltaic-storage charging station alliance. Step 144: Optimization objective of the photovoltaic-storage-charging station alliance Represented as: (16) in, For a given Time period The dynamic electricity price that minimizes the negative utility for distribution network operators.

[0031] Step 2: Based on the optimization objective, establish an upper-level Stackelberg game model consisting of the distribution network operator as the leader and the photovoltaic-storage-charging station alliance as the follower. In this embodiment, the construction process of the upper-level Stackelberg game model in step 2 includes: Step 21: In the distributed photovoltaic-storage charging station system of this invention, the distribution network operator first publishes a dynamic electricity price during each scheduling period to guide the photovoltaic-storage charging station alliance to optimize its power purchase capacity, thereby achieving peak shaving and valley filling and ensuring the stable operation of the power system. The photovoltaic-storage charging station alliance adjusts its power purchase capacity through internal coordination to maximize overall benefits. Therefore, the optimization objective is modeled as a Stackelberg game process for each time period. The upper-level Stackelberg game model is represented as follows: The specific definitions of each element are as follows: (a) Player: As a leader, namely the distribution network operator; As a follower, namely the Photovoltaic-Storage-Charging Station Alliance; (b) Strategy: For leaders' strategies, namely Dynamic electricity price of time-of-use distribution network In the In the round of iteration, Updated by the following formula: (17) For follower strategy, i.e. Total power purchased by the photovoltaic-storage charging station alliance In the photovoltaic-storage-charging station alliance, each photovoltaic-storage-charging station makes collaborative decisions, and the total power purchased by the entire alliance is given by the equilibrium solution of the lower-level game. (c) Utility: The negative utility of the leader is to minimize load fluctuations, as shown in formula (13). For the follower's utility, i.e., to maximize the total revenue of the alliance operation, as in formula (14). Step 22: In game theory, the stable state of a system is a Nash equilibrium. For the Stackelberg game of this invention, its stable state can be represented by a Stackelberg Equilibrium (SE), which is a refined Nash equilibrium. The SE point is defined as follows: The upper-level Stackelberg game model In the middle, strategy group A Stackelberg equilibrium point is formed if and only if the following condition is satisfied: (a) The leader's subgame Nash equilibrium, i.e. (18); (b) The subgame Nash equilibrium of the follower, i.e. (19).

[0032] When the system is at the SE point, neither the leader nor the followers change their strategies to gain less negative utility or greater utility. Therefore, none of the players have an incentive to deviate from the SE point.

[0033] Step 3: Within the photovoltaic-storage-charging station alliance, establish a lower-level Nash bargaining game model and transform the upper-level Stackelberg game model into an optimization problem. Achieve fair cooperation among the photovoltaic-storage-charging stations by maximizing cooperative benefits and distributing benefits. In this embodiment, the construction process of the lower-level Nash bargaining game model in step 3 includes: Step 31: The Photovoltaic-Storage-Charging Station Alliance is a cooperative platform for energy exchange among photovoltaic (PV) and energy storage (ESS) charging stations. Within the system, the PV and ESS charging stations represent different stakeholders and are perfectly rational, reaching consensus through negotiation within the alliance. To ensure the stability and fairness of the cooperation, asymmetric Nash negotiation is used to model the cooperation within the PV and ESS charging station alliance. For each time period... The lower-level Nash bargaining game model is expressed as follows: The specific definitions of each element are as follows: (a) Player: For players to gather, namely, a collection of photovoltaic and energy storage charging stations; (b) Point of divergence: This refers to the set of divergence points for players; to ensure individual benefits, photovoltaic-storage charging stations will only participate in energy exchange if they can obtain higher returns than operating independently; this independent operating cost is called the divergence point. Time period The point of contention in collaboration, i.e., the maximum benefit when operating independently. , represented as: (20) (c) Bargaining power: This represents the bargaining power of the players; since different photovoltaic and energy storage charging stations contribute differently to energy exchange in the cooperation, they also have different bargaining power in the distribution of benefits. Time period Bargain coefficient in cooperation Represented as: (twenty one) (d) Strategy: For the player's strategy set, among which, for Time period The strategy group determines the grid power purchase capacity within the photovoltaic-storage charging station, the charging and discharging power of the energy storage system, and the energy exchange power between photovoltaic-storage charging stations; (e) Utility: As a collection of utility benefits for players, the photovoltaic charging station aims to maximize its own benefits within the cooperative alliance, therefore... Time period The utility Defined as: (twenty two) Step 32, for the photovoltaic-storage charging station alliance The asymmetric Nash bargaining problem is formalized using the Nash product, resulting in the optimization problem of the lower-level Nash bargaining game model. , represented as: (twenty three) st(1)-(12),(20)-(22) Step 33: Since the lower-level Nash bargaining game model is non-convex, it is difficult to solve directly. Based on Nash axioms, the optimization problem... Decomposed into subproblems of maximizing cooperative benefits and the problem of income distribution , respectively represented as: (twenty four) (25) (20), (21) in, For cooperation Time period The optimal self-operating revenue is determined by The solution is obtained; A collection of photovoltaic energy storage and charging stations where energy exchange exists; Step 34: To facilitate the solution and calculation, we will solve the subproblem of maximizing cooperative benefits. Income distribution sub-problem The formal transformation is performed as follows: (a) The subproblem of maximizing the benefits of cooperation Converted to standard form, it is expressed as: (26) in, for The operating revenue of the entire photovoltaic-storage charging station alliance during the period. for Time period Decision variables, for Time period The linear equality constraint matrix, for The vector of constant terms constrained by the linear equality of the time period. for Time period Inequality constraint functions; (b) Sub-problem of revenue allocation Converted to logarithmic form, it can be expressed as: (27) (20), (21) Step 35: To calculate the analytical solution for the optimal transfer payment, construct the Lagrange function, expressed as: (28) in, For Lagrange multipliers; For transfer payments and Lagrange multipliers The Lagrangian function of the variable; Conditions provided by Karush-Kuhn-Tucker (KKT) ,get: (29) because and ,get ,in, for Time period The optimal operating revenue for the entire photovoltaic-storage-charging station alliance; Step 36: Substitute the result of formula (29) back into formula (25) to obtain... The analytical solution is: (30) in, for for Time period Optimal transfer payments within a cooperative alliance.

[0034] Therefore, it is only necessary to use a distributed algorithm to... By solving this problem, we can obtain the solution. The solution, in short This reduces the computational complexity of solving the asymmetric Nash bargaining problem.

[0035] For the two-layer game framework proposed in this invention, the key to solving the problem is how to achieve collaborative consensus among multiple sites in a distributed environment without a central controller. To address this issue, this invention proposes a Multi-Armed Slot Machine Adaptive Step Size Algorithm-Neurodynamics Algorithm (MAB-NDA), which enables global equilibrium under privacy protection to be efficiently achieved by exchanging only limited information among the nodes of each photovoltaic-storage-charging station. The energy scheduling strategy based on MAB-NDA is as follows.

[0036] Step 4: Based on the distributed solution strategy of the neurodynamics algorithm, each photovoltaic-storage-charging station acts as an agent and only exchanges auxiliary variables with its neighboring nodes, so as to achieve global collaborative optimization without sharing decision variables. To protect privacy among different photovoltaic, energy storage, and charging stations, the Distributed Neurodynamic Algorithm (NDA) allows these stations to achieve global optimization without sharing decision information, simply by exchanging partial operators with neighboring nodes on the topology network. In NDA, each photovoltaic, energy storage, and charging station is considered an agent composed of a Recurrent Neural Network (RNN), and the internal network structure of each RNN is as follows: Figure 3 As shown. It serves as a decision variable for photovoltaic-storage charging stations and does not need to be shared with other photovoltaic-storage charging stations. , , and As an auxiliary variable, it is used to exchange among neighboring nodes to help the photovoltaic-storage-charging station alliance reach a consensus.

[0037] In this embodiment, the distributed solution strategy based on the neurodynamics algorithm in step 4 specifically includes: Step 41: In distributed neurodynamics, each photovoltaic-storage-charging station is modeled as an agent composed of a recurrent neural network, whose internal variables include decision variables and auxiliary variables; each agent only exchanges auxiliary variables with neighboring nodes in the topology network and does not share decision variables. The internal variables of the agent are defined as follows: (c) Decision variables : For photovoltaic and energy storage charging stations exist The decision variable for a given time period does not need to be shared with other agents; (b) Auxiliary variables , , and Used for exchanging data between neighboring nodes to help the photovoltaic-storage-charging station alliance reach a consensus; Photovoltaic and energy storage charging stations exist The Lagrange multipliers for the time interval correspond to the dual variables of the equality constraints; Photovoltaic and energy storage charging stations exist The Lagrange multipliers for the time interval correspond to the dual variables of the inequality constraints; Photovoltaic and energy storage charging stations exist Time periods are used as auxiliary variables to convey information related to consistency constraints, promoting consensus among multipliers; Photovoltaic and energy storage charging stations exist Additional auxiliary variables for the time period are used to facilitate consensus on inequality multipliers between adjacent nodes; Step 42: Iteratively update the decision variables and auxiliary variables using discretized neurodynamic equations until the convergence condition is met; (a) Decision variables of the agent The update equation is expressed as: (31) in, For the first Decision variables in round iteration , For the first Decision variables in round iteration , In order to be in Photovoltaic and energy storage charging stations during the time period The The step size of each iteration is updated using a multi-armed slot machine adaptive step size algorithm. For projection operators over the feasible region, The gradient of the site's own operating revenue function. This is the equality constraint matrix. This is an inequality constraint function; For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration ; (b) the auxiliary variables The update equation is expressed as: (32) in, For the first Auxiliary variables of round iteration , for Neighboring nodes in a topological network for The set of neighboring nodes in a topological network; Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration, For the first Auxiliary variables of round iteration , Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration; (c) the auxiliary variables The update equation is expressed as: (33) in, For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration , As a dual variable, it is used to accelerate the auxiliary variable. The convergence process; (d) The auxiliary variables The update equation is expressed as: (34) in, For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration , Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration, Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration, For operators projecting onto non-negative quadrants; (e) the auxiliary variables The update equation is expressed as: (35) in, For the first Auxiliary variables of round iteration ; (f) The convergence condition of the neural dynamics algorithm is measured by the change in decision variables in the form of a sliding window, and is expressed as: (36) in, To determine the convergence measure, the number of rounds is used, i.e., the length of the sliding window. Index for iteration count, In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration The preset convergence accuracy threshold is a sufficiently small positive real number.

[0038] Step 5: In each iteration, each agent adjusts the iteration step size of the neurodynamics algorithm through the multi-armed slot machine adaptive step size algorithm to accelerate the convergence of the neurodynamics algorithm; In the iterative process of distributed neurodynamics algorithms, the choice of step size is crucial to the algorithm's solution efficiency. If the step size is too small, the convergence speed will decrease significantly; if the step size is too large, it may cause oscillations or even prevent convergence. Therefore, this invention introduces a Multi-Armed Bandit (MAB) mechanism to dynamically and adaptively update the iteration step size, thereby accelerating the process at each time interval. Distributed equilibrium solution within the system.

[0039] In this embodiment, step 5, the adaptive step size algorithm based on multi-armed slot machines, specifically includes: Step 51, Initialization Phase: Initialize each Instant rewards Changes in decision variables With each average reward and the number of times selected ;in, For the collection of photovoltaic and energy storage charging stations, For the set of step-size strategies, For iteration rounds; Step 52, Reward Calculation: Based on the... Calculating the change in decision variables in each iteration Then calculate the instant reward. The formula is as follows: (37) (38) in, for During the period The The immediate reward for each iteration step is used to evaluate the contribution of the current iteration step to the convergence of the neurodynamics algorithm. for During the period No. The change in decision variables during rounds of iteration. for During the period No. The change in decision variables during each iteration; In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration; Step 53, Average Reward Update: Update instant rewards Used to update the Average reward of the selected action in the round The formula is as follows: (39) in, for During the period In the Actions during round iteration The average reward for During the period In the Actions during round iteration The average reward; for During the period No. Actions during round iteration The number of times it was selected; Step 54, Action Selection: To balance the utilization of known high-reward arms with the exploration of insufficiently tried arms, the action selection is based on the number of times all actions have been selected. and average reward The confidence upper limit strategy is used to select the action for this iteration. The formula is as follows: (40) in, for Photovoltaic and energy storage charging stations during the period (i.e., intelligent agent) ) No. The step size calculation strategy for each iteration. To explore the weighting coefficients; Step 55, Step Length Calculation: Based on the selected action The corresponding step size calculation formula determines the step size for this iteration. ; For distributed neurodynamics algorithms in time periods The first In each iteration, if we consider the step-size calculation strategy as the "arms" of a multi-armed slot machine, then the candidate action set (i.e., the step-size strategy set) is used... It indicates. Among them, , , They are defined as follows.

[0040] arm (Fixed step size): (41) arm (Cosine annealing): (42) arm (Linear decay): (43) like For fixed step size ,but ;in, This is a preset fixed step size value; like Cosine annealing step ,but ;in, For the maximum iteration step size, To be the minimum iteration step size, This represents the maximum number of iterations. like linear decay step size ,but ; Step 56, Update Selection Count: Increment the selection count of the selected action by 1, i.e. ;in, for During the period No. Actions during round iteration Number of times selected for During the period No. Actions during round iteration The number of times it was selected; Step 57, Iteration Loop: Repeat steps 52 to 56 until the maximum number of iterations is reached. Output iteration step size .

[0041] The specific steps of the MAB adaptive step-size algorithm are shown in Algorithm 1. For each distributed agent, in each iteration, the average reward of the step-size policy selected in the previous iteration is first updated (lines 3-4). Secondly, ensuring that all step-size policies have been selected, the step-size policy with the highest score under the UCB policy is selected as the action for this iteration (lines 5-12). Finally, the step size for this iteration is obtained through the corresponding step-size calculation method (lines 13-19).

[0042]

[0043] Step 6: A distributed two-level optimization algorithm based on the multi-armed slot machine adaptive step size algorithm-neurodynamics algorithm is used to solve the upper-level Stackelberg game and the lower-level Nash bargaining game through nested inner and outer layers. The leader strategy and follower strategy are alternately updated in each time period until the Stackelberg equilibrium is reached and the optimal strategy is output.

[0044] In this embodiment, the specific process of the distributed two-layer optimization algorithm in step 6 is as follows: Step 61, Initialization Phase: Set the scheduling time set T and initialize the leader policy. Follower strategy Smart agent strategy for photovoltaic-storage charging stations Decision variables and auxiliary variables of the photovoltaic-storage-charging station intelligent agent, as well as the iteration step size. ; Step 62, Outer Loop: For each time period Optimize sequentially; Step 63, Leader Initialization: At the beginning of each period, the leader provides an initial strategy. ; Step 64, Inner Loop: All photovoltaic-storage-charging station agents execute the neural dynamics algorithm and the multi-armed slot machine adaptive step-size algorithm in parallel to solve the response strategy under the current electricity price; specifically including: Step 641, Order ; Step 642: Call the multi-armed slot machine adaptive step size algorithm to update the iteration step size. ; Step 643: Update the decision variables according to formula (31) ; Step 644: Update the auxiliary variables according to formulas (32) and (35). , ,、 and ; Step 645 With neighboring nodes in the topology network Communication, exchanging auxiliary variables; Step 646, Iteration Rounds ; Step 647: Repeat steps 642 to 646 until the inner convergence condition is met. Or reach the maximum number of iterations ,in, Index for iteration count, In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration In order to be in Time-of-use photovoltaic energy storage charging station In the -1 round of iterations of decision variables.

[0045] Step 65: Strategy Update: Update the agent strategy for each photovoltaic-storage-charging station based on the inner layer solution results. and follower strategy ; Step 66: Leader Update: The leader adjusts its strategy according to the followers' response strategies and the update rules of the Stackelberg game. ; Step 67, Convergence Judgment: Iteration Rounds Repeat steps 64 to 66 until the leader strategy meets the convergence condition. Furthermore, the follower strategy satisfies the convergence condition. This achieves the Stackelberg equilibrium; among which, In order to be in Time period Leader strategy during round iteration In order to be in Time period Leader strategy during round iteration In order to be in Time period Follower strategy during round iteration In order to be in Time period Follower strategy during round iteration; Step 68, Time Period Cycle: Enter the next time period and repeat steps 63 to 67 until all time periods have been optimized, and output the optimal strategy group. ,in, The optimal strategy for leaders. The optimal strategy for followers The optimal strategy for the intelligent agent of the photovoltaic-storage charging station.

[0046] In this embodiment, the output of the distributed two-layer optimization algorithm includes: (a) Optimal dynamic electricity price for each time period; (b) The optimal power purchase, charging and discharging power of the energy storage system, and energy exchange power between the photovoltaic and energy storage charging stations at each time period; (c) Optimal transfer payments and optimal operating revenue for each photovoltaic-storage-charging station at each time period.

[0047] To solve the bi-level optimization problem of this invention, a distributed bi-level optimization algorithm based on MAB-NDA is designed, with specific steps as shown in Algorithm 2. For each scheduling period, the leader (distribution network operator) first formulates an initial electricity price strategy (line 3). Next, the lower-level Nash bargaining game is entered for solution. Each distributed photovoltaic-storage-charging station agent executes in parallel, dynamically updating the iteration step size by calling Algorithm 1 (line 8) and combining information interaction between neighbors to achieve distributed collaborative optimization (lines 9-13) to obtain the optimal response strategy under the current electricity price (line 15). Finally, after updating the Nash bargaining strategy and the follower (photovoltaic-storage-charging station alliance) strategies (lines 17-18), the algorithm returns to the upper-level Stackelberg game. The leader updates its strategy based on the follower's response strategy (line 19) until a Stackelberg equilibrium is reached (line 21).

[0048]

[0049] In Algorithm 2, Algorithm 1 is called as a subroutine to calculate the iteration step size in each distributed agent. Since the size of the MAB action set in Algorithm 1 is constant, and both reward and step size calculations are constant-order algebraic operations, the time complexity of Algorithm 1 is O(log n). In the Nash bargaining game solution phase, assuming the worst-case scenario of a fully connected node topology, the time complexity of distributed neurodynamics computation is O(n log n). The time complexity of solving the Nash bargaining game is O(n log n). In the Stackelberg game solution phase, the follower strategy update involves the aggregation of all Nash bargaining players' strategies, with a time complexity of O(n log n). This is less than the time complexity of internal Nash bargaining. Therefore, within a scheduling period... The overall time complexity of the MAB-NDA-based two-layer optimization algorithm is... It is suitable for real-time scheduling scenarios.

[0050] The above description is only a part of the embodiments of the present invention and does not limit the scope of protection of the present invention. Any equivalent device or equivalent process transformation made based on the content of the present invention specification and drawings, or direct or indirect application in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A distributed photovoltaic-storage-charging station energy scheduling method based on a two-layer game theory approach, characterized in that, Includes the following steps: Step 1: Construct a distributed photovoltaic-storage charging station system, including the power distribution network and a photovoltaic-storage charging station alliance composed of multiple photovoltaic-storage charging stations, and define the optimization objectives of the photovoltaic-storage charging station system; Step 2: Based on the optimization objective, establish an upper-level Stackelberg game model consisting of the distribution network operator as the leader and the photovoltaic-storage-charging station alliance as the follower. Step 3: Within the photovoltaic-storage-charging station alliance, establish a lower-level Nash bargaining game model and transform the upper-level Stackelberg game model into an optimization problem. Achieve fair cooperation among the photovoltaic-storage-charging stations by maximizing cooperative benefits and distributing benefits. Step 4: Based on the distributed solution strategy of the neurodynamics algorithm, each photovoltaic-storage-charging station acts as an agent and only exchanges auxiliary variables with its neighboring nodes, so as to achieve global collaborative optimization without sharing decision variables. Step 5: In each iteration, each agent adjusts the iteration step size of the neurodynamics algorithm through the multi-armed slot machine adaptive step size algorithm to accelerate the convergence of the neurodynamics algorithm; Step 6: A distributed two-level optimization algorithm based on the multi-armed slot machine adaptive step size algorithm-neurodynamics algorithm is used to solve the upper-level Stackelberg game and the lower-level Nash bargaining game through nested inner and outer layers. The leader strategy and follower strategy are alternately updated in each time period until the Stackelberg equilibrium is reached and the optimal strategy is output.

2. The energy scheduling method for distributed photovoltaic-storage-charging stations based on a two-layer game theory as described in claim 1, characterized in that, Step 1 specifically includes: Step 11: The distributed photovoltaic-storage charging station system consists of a power distribution network and a photovoltaic-storage charging station alliance composed of multiple photovoltaic-storage charging stations; Step 12: The power distribution network serves as the external power source for the photovoltaic-storage charging station, supplying power to the station unidirectionally. Its operation must meet multiple constraints, including total power purchase constraints, electricity price boundary constraints, and global balance constraints. Step 13: The photovoltaic-energy storage charging station consists of a photovoltaic system, an energy storage system, a charging pile system, and an energy management system; specifically as follows: Step 131: The operation mechanism of the photovoltaic system is as follows: the energy generated by the photovoltaic system is first delivered to the charging pile system for consumption, the surplus is stored in the energy storage system or sold to other photovoltaic-storage charging stations, and the excess energy that cannot be consumed is discarded. Step 132: The energy storage system provides energy regulation for the photovoltaic-energy storage charging station; Step 133: The charging pile system serves as the interface between the photovoltaic energy storage charging station and the electric vehicle. Step 134: The energy management system is responsible for communicating with the external power distribution network, other photovoltaic-storage charging stations and electric vehicles to formulate energy dispatch strategies with the goal of maximizing the operating revenue of photovoltaic-storage charging stations, and send them to the energy storage system and charging pile system for execution. Step 14: In the scheduling of the distributed photovoltaic-storage-charging station system, the optimization objective of the distribution network operator is to minimize power fluctuations, and the optimization objective of the photovoltaic-storage-charging station alliance is to maximize the total operating revenue of the alliance. Based on the optimization objectives of the distribution network operator and the photovoltaic-storage-charging station alliance, the scheduling problem of the distributed photovoltaic-storage-charging station system is transformed into an optimization objective.

3. The energy scheduling method for distributed photovoltaic-storage-charging stations based on a two-layer game theory as described in claim 2, characterized in that, Step 11 specifically includes: definition Each photovoltaic energy storage charging station is for ,in, This indicates the first photovoltaic-storage charging station. This indicates the second photovoltaic-storage charging station. Indicates the first i One photovoltaic energy storage charging station Indicates the first One photovoltaic energy storage charging station; the scheduling time period set T is divided into... A period of time, namely ; The total power purchase constraint, electricity price boundary constraint, and global balance constraint in step 12 are as follows: (a) Total power purchase constraint: Total power purchased by the photovoltaic-storage-charging station alliance during the period The sum of the purchased power of each photovoltaic-storage charging station is expressed as: （1） in, Photovoltaic and energy storage charging stations exist Electricity purchase capacity during the time period This represents the maximum power that the distribution network can supply. (b) Electricity price boundary constraints: Based on the time-of-use pricing, the distribution network operator releases a dynamic electricity price according to the current load to guide the photovoltaic-storage charging station to adjust its power purchase capacity; Dynamic electricity price of time-of-use distribution network Represented as: （2） in, for Time-of-use pricing for distribution network operators This is the electricity price adjustment coefficient. The average power purchase capacity of photovoltaic-storage charging stations. For the minimum electricity price, The maximum electricity price; (c) Global Balance Constraint: The energy exchange power of each photovoltaic-storage-charging station in the distribution network must satisfy the global balance constraint, expressed as: （3） in, In order to be in Time-of-use photovoltaic energy storage charging station The energy exchange power between other photovoltaic and energy storage charging stations; a positive value indicates that energy is being transferred to other photovoltaic and energy storage charging stations, while a negative value indicates that energy is being received from other photovoltaic and energy storage charging stations. Step 131 specifically includes: Photovoltaic and energy storage charging stations exist Actual photovoltaic power consumed during the period Determined by supply and demand balance, expressed as: （4） in, In order to be in Time-of-use photovoltaic energy storage charging station The actual output power of the photovoltaic system In order to be in Time-of-use photovoltaic energy storage charging station The total charging power demand of electric vehicles in China In order to be in Time-of-use photovoltaic energy storage charging station The charging and discharging power of the energy storage system is negative when it indicates charging and positive when it indicates discharging. In order to be in Time-of-use photovoltaic energy storage charging station The power purchased from the distribution network; Step 132 specifically includes: exist Time-of-use photovoltaic energy storage charging station The charging and discharging power of medium-energy storage system To optimize the variables, it is represented as: （5） in, In order to be in Time-of-use photovoltaic energy storage charging station The energy storage charge-discharge coefficient, Photovoltaic and energy storage charging stations Maximum charging power, Photovoltaic and energy storage charging stations Maximum discharge power; exist During the period, photovoltaic and energy storage charging stations State of charge of medium energy storage system The formula is: （6） in, and Photovoltaic and energy storage charging stations Upper and lower limits of the state of charge of medium-capacity energy storage systems Photovoltaic and energy storage charging stations Medium energy storage battery capacity, The length of a time period; The energy conversion coefficient of the energy storage system during the charging and discharging process is expressed as: （7） in, These refer to the charging efficiency and discharging efficiency of the energy storage system, respectively. The energy storage system during charging and discharging Time-of-use photovoltaic energy storage charging station Energy storage loss costs Represented as: （8） in, This is the energy storage loss cost coefficient; Step 133 specifically includes: exist Time-of-use photovoltaic energy storage charging station The total charging power supplied to electric vehicles by the charging pile system , represented as: （9） in, In order to be in Time-of-use photovoltaic energy storage charging station The total charging power required by electric vehicles in China is expressed as: （10） in, In order to be in Time-of-use photovoltaic energy storage charging station The increased demand for charging power for electric vehicles in China; In the first Time-of-use photovoltaic energy storage charging station Total charging power demand for electric vehicles in China; In the first Time-of-use photovoltaic energy storage charging station The actual charging power supplied to electric vehicles by the charging pile system in China; Step 134 specifically includes: exist Time-of-use photovoltaic energy storage charging station Power purchased from the distribution network Represented as: （11） exist Time-of-use photovoltaic energy storage charging station Self-operated revenue Represented as: （12） in, In order to be in Time-of-use photovoltaic energy storage charging station The price of charging services, In order to be in Dynamic electricity pricing for time-of-use distribution networks; Step 14 specifically includes: Step 141, in Negative effects of time-of-use distribution network operators for: （13） Step 142, in The effectiveness of the photovoltaic-storage charging station alliance for: （14） in, for Time-of-use photovoltaic energy storage charging station In the revenue distribution of the photovoltaic-storage-charging station alliance, when... When, it means Obtaining energy from other photovoltaic and energy storage charging stations in the photovoltaic and energy storage charging station alliance requires paying a fee. When, it means There is no charge for transmitting energy to other photovoltaic and energy storage charging stations in the photovoltaic and energy storage charging station alliance. Step 143, the optimization objective of the distribution network operator Represented as: (15) in, For a given Time period At that time, the total power purchased will maximize the effectiveness of the photovoltaic-storage charging station alliance. Step 144: Optimization objective of the photovoltaic-storage-charging station alliance Represented as: (16) in, For a given Time period The dynamic electricity price that minimizes the negative utility for distribution network operators.

4. The energy scheduling method for distributed photovoltaic-storage-charging stations based on a two-layer game theory as described in claim 1, characterized in that, The construction process of the upper-level Stackelberg game model in step 2 includes: Step 21: Model the optimization objective as a Stackelberg game process, for each time period... The upper-level Stackelberg game model is represented as follows: The specific definitions of each element are as follows: (a) Player: As a leader, namely the distribution network operator; As a follower, namely the Photovoltaic-Storage-Charging Station Alliance; (b) Strategy: For leaders' strategies, namely Dynamic electricity price of time-of-use distribution network In the In the round of iteration, Updated by the following formula: （17） For follower strategy, i.e. Total power purchased by the photovoltaic-storage charging station alliance In the photovoltaic-storage charging station alliance, each photovoltaic-storage charging station makes collaborative decisions. (c) Utility: The negative utility of the leader is to minimize load fluctuations; For the utility of followers, that is, to maximize the total revenue of the alliance operation; Step 22, in the upper-level Stackelberg game model In the middle, strategy group A Stackelberg equilibrium point is formed if and only if the following condition is satisfied: (a) The leader's subgame Nash equilibrium, i.e. (18); (b) The subgame Nash equilibrium of the follower, i.e. (19).

5. The energy scheduling method for distributed photovoltaic-storage-charging stations based on a two-layer game theory as described in claim 1, characterized in that, The construction process of the lower-level Nash bargaining game model in step 3 includes: Step 31: Model the cooperation in the photovoltaic-storage-charging station alliance using asymmetric Nash bargaining. For each time period... The lower-level Nash bargaining game model is expressed as follows: The specific definitions of each element are as follows: (a) Player: For players to gather, namely, a collection of photovoltaic and energy storage charging stations; (b) Point of divergence: For the player's branching points; in Time period The point of contention in collaboration, i.e., the maximum benefit when operating independently. , represented as: （20） (c) Bargaining power: This represents the collective bargaining power of the players. Time period Bargain coefficient in cooperation Represented as: （21） (d) Strategy: For the player's strategy set, among which, for Time period The strategy group; (e) Utility: For the player's utility set, Time period The utility Defined as: （22） Step 32, for the photovoltaic-storage charging station alliance The asymmetric Nash bargaining problem is formalized using the Nash product, resulting in the optimization problem of the lower-level Nash bargaining game model. , represented as: （23） st(1)-(12),(20)-(22) Step 33: According to Nash's axioms, the optimization problem... Decomposed into subproblems of maximizing cooperative benefits and the problem of income distribution , respectively represented as: （24） （25） ，（20），（21） in, For cooperation Time period The optimal self-operating revenue is determined by The solution is obtained; A collection of photovoltaic energy storage and charging stations where energy exchange exists; Step 34: Maximizing the subproblem of cooperation benefits Income distribution sub-problem The formal transformation is performed as follows: (a) The subproblem of maximizing the benefits of cooperation Converted to standard form, it is expressed as: （26） in, for The operating revenue of the entire photovoltaic-storage charging station alliance during the period. for Time period Decision variables, for Time period The linear equality constraint matrix, for The vector of constant terms constrained by the linear equality of the time period. for Time period Inequality constraint functions; (b) Sub-problem of revenue allocation Converted to logarithmic form, it can be expressed as: （27） ，（20），（21） Step 35: Construct the Lagrange function, expressed as: （28） in, For Lagrange multipliers; For transfer payments and Lagrange multipliers The Lagrangian function of the variable; Due to the conditions of Karush-Kun-Tucker, ,get: （29） because and ,get ,in, for Time period The optimal operating revenue for the entire photovoltaic-storage-charging station alliance; Step 36: Substitute the result of formula (29) back into formula (25) to obtain... The analytical solution is: （30） in, for for Time period Optimal transfer payments within a cooperative alliance.

6. The energy scheduling method for distributed photovoltaic-storage-charging stations based on a two-layer game theory as described in claim 1, characterized in that, The distributed solution strategy based on the neurodynamics algorithm in step 4 specifically includes: Step 41: In distributed neurodynamics, each photovoltaic-storage-charging station is modeled as an agent composed of a recurrent neural network, whose internal variables include decision variables and auxiliary variables; each agent only exchanges auxiliary variables with neighboring nodes in the topology network and does not share decision variables. The internal variables of the agent are defined as follows: (a) Decision variables : For photovoltaic and energy storage charging stations exist The decision variable for a given time period does not need to be shared with other agents; (b) Auxiliary variables , , and Used for exchanging data between neighboring nodes to help the photovoltaic-storage-charging station alliance reach a consensus; Photovoltaic and energy storage charging stations exist The Lagrange multipliers for the time interval correspond to the dual variables of the equality constraints; Photovoltaic and energy storage charging stations exist The Lagrange multipliers for the time interval correspond to the dual variables of the inequality constraints; Photovoltaic and energy storage charging stations exist Time periods are used as auxiliary variables to convey information related to consistency constraints, promoting consensus among multipliers; Photovoltaic and energy storage charging stations exist Additional auxiliary variables for the time period are used to facilitate consensus on inequality multipliers between adjacent nodes; Step 42: Iteratively update the decision variables and auxiliary variables using discretized neurodynamic equations until the convergence condition is met; (a) Decision variables of the agent The update equation is expressed as: （31） in, For the first Decision variables in round iteration , For the first Decision variables in round iteration , In order to be in Photovoltaic and energy storage charging stations during the time period The The step size of the round iteration, For projection operators over the feasible region, The gradient of the site's own operating revenue function. This is the equality constraint matrix. This is an inequality constraint function; For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration ; (b) the auxiliary variables The update equation is expressed as: （32） in, For the first Auxiliary variables of round iteration , for Neighboring nodes in a topological network for The set of neighboring nodes in a topological network; Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration, For the first Auxiliary variables of round iteration , Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration; (c) the auxiliary variables The update equation is expressed as: （33） in, For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration , As a dual variable, it is used to accelerate the auxiliary variable. The convergence process; (d) The auxiliary variables The update equation is expressed as: （34） in, For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration , For the first Auxiliary variables of round iteration , Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration, Photovoltaic and energy storage charging stations exist The first period Auxiliary variables for round iteration, For operators projecting to the non-negative quadrants; (e) the auxiliary variable The update equation is expressed as: （35） in, For the first Auxiliary variables of round iteration ; (f) The convergence condition of the neural dynamics algorithm is measured by the change in decision variables in the form of a sliding window, and is expressed as: （36） in, To determine the convergence measure, the number of rounds is used, i.e., the length of the sliding window. Index for iteration count, In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration This is the preset convergence accuracy threshold.

7. The energy scheduling method for distributed photovoltaic-storage-charging stations based on a two-layer game theory as described in claim 1, characterized in that, The adaptive step size algorithm based on multi-armed slot machines in step 5 specifically includes: Step 51, Initialization Phase: Initialize each Instant rewards Changes in decision variables With each average reward and the number of times selected ;in, For the collection of photovoltaic and energy storage charging stations, For the set of step-size strategies, For iteration rounds; Step 52, Reward Calculation: Based on the... Calculating the change in decision variables in each iteration Then calculate the instant reward. The formula is as follows: in, for During the period The The immediate reward for each iteration step is used to evaluate the contribution of the current iteration step to the convergence of the neurodynamics algorithm. for During the period No. The change in decision variables during rounds of iteration. for During the period No. The change in decision variables during each iteration; In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration; Step 53, Average Reward Update: Update instant rewards Used to update the Average reward of the selected action in the round The formula is as follows: in, for During the period In the Selecting actions during round iteration The average reward, for During the period In the Actions during round iteration The average reward; for During the period No. Actions during round iteration The number of times it was selected; Step 54, Action Selection: Based on the number of times all actions have been selected. and average reward The confidence upper limit strategy is used to select the action for this iteration. The formula is as follows: in, for Photovoltaic and energy storage charging stations during the period No. The step size calculation strategy for each iteration. To explore the weighting coefficients; Step 55, Step Length Calculation: Based on the selected action The corresponding step size calculation formula determines the step size for this iteration. ; like For fixed step size ,but ;in, This is a preset fixed step size value; like Cosine annealing step ,but ;in, For the maximum iteration step size, To be the minimum iteration step size, This represents the maximum number of iterations. like linear decay step size ,but ; Step 56, Update Selection Count: Increment the selection count of the selected action by 1, i.e. ;in, for During the period No. Actions during round iteration Number of times selected for During the period No. Actions during round iteration The number of times it was selected; Step 57, Iteration Loop: Repeat steps 52 to 56 until the maximum number of iterations is reached. Output iteration step size .

8. The energy scheduling method for distributed photovoltaic-storage-charging stations based on a two-layer game theory as described in claim 1, characterized in that, The specific process of the distributed two-layer optimization algorithm in step 6 is as follows: Step 61, Initialization Phase: Set the scheduling time set T and initialize the leader policy. Follower strategy Smart agent strategy for photovoltaic-storage charging stations Decision variables and auxiliary variables of the photovoltaic-storage-charging station intelligent agent, as well as the iteration step size. ; Step 62, Outer Loop: For each time period Optimize sequentially; Step 63, Leader Initialization: At the beginning of each period, the leader provides an initial strategy. ; Step 64, Inner Loop: All photovoltaic-storage-charging station agents execute the neural dynamics algorithm and the multi-armed slot machine adaptive step size algorithm in parallel to solve the response strategy under the current electricity price; Step 65: Strategy Update: Update the agent strategy for each photovoltaic-storage-charging station based on the inner layer solution results. and follower strategy ; Step 66: Leader Update: The leader adjusts its strategy according to the followers' response strategies and the update rules of the Stackelberg game. ; Step 67, Convergence Judgment: Iteration Rounds Repeat steps 64 to 66 until the leader strategy meets the convergence condition. Furthermore, the follower strategy satisfies the convergence condition. This achieves the Stackelberg equilibrium; among which, In order to be in Time period Leader strategy during round iteration In order to be in Time period Leader strategy during round iteration In order to be in Time period Follower strategy during round iteration In order to be in Time period Follower strategy during round iteration; Step 68, Time Period Cycle: Enter the next time period and repeat steps 63 to 67 until all time periods have been optimized, and output the optimal strategy group. ,in, The optimal strategy for leaders. The optimal strategy for followers The optimal strategy for the intelligent agent of the photovoltaic-storage charging station.

9. The energy scheduling method for distributed photovoltaic-storage-charging stations based on a two-layer game theory as described in claim 8, characterized in that, Step 64 specifically includes: Step 641, Order ; Step 642: Call the multi-armed slot machine adaptive step size algorithm to update the iteration step size. ; Step 643: Update the decision variables according to formula (31) ; Step 644: Update the auxiliary variables according to formulas (32) and (35). , ,、 and ; Step 645 With neighboring nodes in the topology network Communication, exchanging auxiliary variables; Step 646, Iteration Rounds ; Step 647: Repeat steps 642 to 646 until the inner convergence condition is met. Or reach the maximum number of iterations ,in, Index for iteration count, In order to be in Time-of-use photovoltaic energy storage charging station In the Decision variables during round iteration In order to be in Time-of-use photovoltaic energy storage charging station In the -1 round of iterations. Decision variables.

10. The energy scheduling method for distributed photovoltaic-storage-charging stations based on a two-layer game theory as described in claim 1, characterized in that, The output of the distributed two-level optimization algorithm includes: (a) Optimal dynamic electricity price for each time period; (b) The optimal power purchase, charging and discharging power of the energy storage system, and energy exchange power between the photovoltaic and energy storage charging stations at each time period; (c) Optimal transfer payments and optimal operating revenue for each photovoltaic-storage-charging station at each time period.

Images