Reconfigurable energy storage control method and system based on group game

Abstract

The application discloses a reconfigurable energy storage control method and system based on group game, relates to the technical field of energy storage control, and comprises the following steps: deploying a digital twin platform mapped with a reconfigurable energy storage system, and synchronously collecting operation parameters of each access energy storage module; inputting collected photovoltaic and load sequences into a trained prediction model for processing, and establishing an operation scene set by using predicted photovoltaic and load output curves; calculating value coefficients of each access energy storage module under different operation scenes according to the operation scene set and the operation parameters; defining a benefit function according to the value coefficients, calculating the Sharpe value of each access energy storage module by using the benefit function, initiating a game election, and determining a master control module; and driving each access energy storage module to solve an optimization problem generated by the master control module by using a distributed negotiation algorithm embedded with the value coefficients, outputting a target power control strategy, and improving the control efficiency of the energy storage system.

Description

Technical Field

[0001] This invention relates to the field of energy storage control technology, and in particular to a reconfigurable energy storage control method and system based on group game theory. Background Technology

[0002] Against the backdrop of the current construction of a new power system characterized by high-proportion renewable energy integration and the development of distributed energy, user-side energy storage, as an important carrier for enhancing energy autonomy and participating in grid interaction, is undergoing a profound transformation in its technological evolution from fixed and centralized to flexible and intelligent. Modular design has become the mainstream technological approach to achieve flexible system expansion and configuration.

[0003] Existing modular energy storage systems are mostly limited to "static expansion" and "in-situ replacement" modes. Their system topology is fixed, making dynamic reconfiguration difficult. They cannot adapt to flexible scenarios such as load migration and temporary power consumption. Furthermore, their control relies on rules driven by static parameters (such as SOC), lacking global optimization capabilities and resulting in low control efficiency.

[0004] Therefore, how to effectively control reconfigurable energy storage and improve the reliability of new power system operation has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] This invention provides a reconfigurable energy storage control method and system based on swarm game theory, to solve how to achieve global collaborative optimization and reliable operation of energy storage systems by constructing a swarm intelligence hybrid game mechanism.

[0006] To address the aforementioned technical problems, embodiments of the present invention provide a reconfigurable energy storage control method based on group game theory, comprising:

[0007] A digital twin platform that maps to a reconfigurable energy storage system is deployed to synchronously collect the operating parameters of each connected energy storage module;

[0008] The collected photovoltaic and load sequences are input into the trained prediction model for processing, and the photovoltaic and load output curves output by the prediction model are used to establish a set of operating scenarios.

[0009] Based on the set of operating scenarios and the operating parameters, calculate the value coefficient of each access energy storage module reflecting cycle life loss under different operating scenarios;

[0010] A revenue function is defined based on the value coefficient. The Shapley value of each access energy storage module is calculated using the revenue function, and a game election is initiated to determine the master control module.

[0011] The system drives each of the access energy storage modules to solve the optimization problem generated by the main control module using a distributed negotiation algorithm embedded with the value coefficient, and outputs a target power control strategy.

[0012] Furthermore, the step of establishing an operating scenario set using the photovoltaic and load output curves output by the prediction model includes:

[0013] Obtain the prediction error generated by the prediction model and determine the uncertainty parameters;

[0014] The uncertain parameters are used as the amplitude reference to inject disturbances into the photovoltaic and load output curves to generate disturbance curves;

[0015] The disturbance curve is encapsulated into the set of operating scenarios.

[0016] Furthermore, the step of calculating the value coefficient reflecting cycle life loss of each of the access energy storage modules under different operating scenarios based on the set of operating scenarios and the operating parameters includes:

[0017] Using pre-constructed historical energy storage operation data and the set of operation scenarios as input, a cycle life prediction model is trained.

[0018] In each operating scenario, the operating parameters corresponding to each access energy storage module are input into the cycle life prediction model for simulation and deduction, and the life loss of each access energy storage module in the current operating scenario is output.

[0019] The value coefficient corresponding to each access energy storage module is determined by accumulating the lifespan loss under all operating scenarios.

[0020] Furthermore, defining the revenue function based on the value coefficient includes:

[0021] An alliance is formed by any one or more of the access energy storage modules, and the sum of the maximum chargeable and dischargeable power and the average response speed of each access energy storage module in the current state of charge are obtained. The sum of the maximum chargeable and dischargeable power and the average response speed are jointly used as the energy storage efficiency item.

[0022] Each alliance was simulated under different operating scenarios, and the performance loss function of each alliance was established based on the simulation results.

[0023] The value coefficient is weighted and coordinated with the energy storage efficiency term and the efficiency loss function to determine the revenue function.

[0024] Furthermore, the step of calculating the Shapley value of each of the access energy storage modules using the revenue function and initiating a game-theoretic election to determine the master control module includes:

[0025] Each of the access energy storage modules bids according to the corresponding Shapley value in a Vickrey auction, and the access energy storage module with the highest bid is initially selected as the main control module;

[0026] Each non-master control module in the access energy storage module competes with the master control module, and the access energy storage module that wins the competition is replaced by the master control module.

[0027] Furthermore, each non-master control module initiates competition with the master control module, and the successfully competing access energy storage module is replaced by the master control module, including:

[0028] Calculate the first benefit function value of the alliance under the first power allocation strategy issued by the current main control module;

[0029] During the competition, each non-master module submits its corresponding second power allocation strategy and calculates the alliance's second revenue function value under the second power allocation strategy.

[0030] The access energy storage module whose second revenue function value is higher than the first revenue function value is replaced by the main control module.

[0031] Furthermore, the process of driving each of the access energy storage modules to solve the optimization problem generated by the main control module using a distributed negotiation algorithm embedded with the value coefficient includes:

[0032] The main control module extracts the scheduling task within the target control cycle from the digital twin platform, and transforms the scheduling task into an optimization problem with the objectives of minimizing lifetime decay and minimizing grid dispatch command tracking error, and with the output of each of the access energy storage modules as the decision variable.

[0033] The value coefficient is embedded as a penalty factor into the Lagrange function of the preset negotiation algorithm to obtain a distributed negotiation algorithm, which is then used to iteratively solve the optimization problem.

[0034] Another embodiment of the present invention provides a reconfigurable energy storage control system based on group game theory, comprising:

[0035] The deployment module is used to deploy a digital twin platform that maps to the reconfigurable energy storage system and synchronously collect the operating parameters of each connected energy storage module.

[0036] The scenario establishment module is used to input the collected photovoltaic and load sequences into the trained prediction model for processing, and to establish a set of operating scenarios by using the photovoltaic and load output curves output by the prediction model.

[0037] The value coefficient determination module is used to calculate the value coefficient of each access energy storage module reflecting cycle life loss under different operating scenarios based on the operating scenario set and the operating parameters.

[0038] The game-theoretic election module is used to define a payoff function based on the value coefficient, calculate the Shapley value of each of the access energy storage modules using the payoff function, and initiate a game-theoretic election to determine the master control module.

[0039] The strategy generation module is used to drive each of the access energy storage modules to solve the optimization problem generated by the main control module using a distributed negotiation algorithm embedded with the value coefficient, and output the target power control strategy.

[0040] Another embodiment of the present invention provides a computer device including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the reconfigurable energy storage control method based on group game theory as described above.

[0041] In another embodiment of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the reconfigurable energy storage control method based on group game theory as described above.

[0042] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following:

[0043] This invention improves the accuracy of real-time energy storage status perception by deploying a digital twin platform that maps in real time to a reconfigurable energy storage system and synchronously collecting operating parameters of each module. By inputting photovoltaic and load sequences into a prediction model and establishing an operating scenario set, it enhances the coverage of various energy storage operating scenarios and avoids scheduling deviations caused by environmental uncertainties. By calculating value coefficients reflecting cycle life loss based on the scenario set and operating parameters, defining a revenue function with the value coefficients, and initiating a game election based on the Shapley value, it improves the reliability and execution efficiency of collaborative decision-making. By using a distributed negotiation algorithm with embedded value coefficients to solve the optimization problem and output the power control strategy for each energy storage module, it avoids conflicts between energy storage modules and ensures the temperature operation of the power system. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of the reconfigurable energy storage control method based on group game theory in one embodiment of the present invention.

[0045] Figure 2 This is a schematic diagram of a reconfigurable energy storage control system based on group game theory in one embodiment of the present invention.

[0046] Figure 3 This is a structural block diagram of a preferred embodiment of a computer device provided by the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0048] In the description of this application, the terms "first," "second," "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," "third," etc., may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0049] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; they can refer to mechanical connections or electrical connections; they can refer to direct connections or indirect connections through an intermediate medium; and they can refer to the internal communication between two components. The terms "vertical," "horizontal," "left," "right," "upper," "lower," and similar expressions used herein are for illustrative purposes only and do not indicate or imply that the device or component referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0050] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art. The terminology used in this specification is for the purpose of describing specific embodiments only and is not intended to limit the invention. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0051] One embodiment of the present invention provides a reconfigurable energy storage control method based on group game theory. For details, please refer to [link to relevant documentation]. Figure 1 , Figure 1 The diagram shown is a schematic flowchart of a reconfigurable energy storage control method based on group game theory in one embodiment of the present invention, including the following steps:

[0052] S1, a digital twin platform for deploying and mapping reconfigurable energy storage systems, synchronously collecting operating parameters of each connected energy storage module.

[0053] To achieve plug-and-play control and scheduling of each energy storage module in the reconfigurable energy storage system, this embodiment deploys a self-evolving digital twin platform on the edge computing node. This platform establishes a real-time communication link with the reconfigurable energy storage system via the OPC UA protocol. When any energy storage module with a globally unique identifier (UUID, e.g., "ModX", representing a specific access energy storage module Module X) accesses the digital twin platform, the platform immediately triggers an object-based instantiation process. By synchronizing the operating parameters of the access energy storage module, including voltage, current, and full lifecycle state parameters (including SOH, SOC, dynamic internal resistance, and temperature), a virtual image mapped to each access energy storage module is constructed. In this embodiment, the specific format of the virtual image can be understood as a structured data packet (Struct or JSON object) containing vector representations of the operating states [SOH (health), SOC (state of charge), Rin (internal resistance), T (temperature)].

[0054] Understandably, in terms of physical form, these access energy storage modules have intelligent composite interfaces, which include both power interfaces for transmitting electricity and communication interfaces for transmitting data. They integrate cell arrays (such as lithium iron phosphate batteries), battery management systems (BMS), and matching power switching devices.

[0055] In one implementation of this embodiment, a reconfigurable energy storage system can be constructed by configuring twenty sets of standardized energy storage modules. During a three-month operation cycle, a scenario is simulated in which ten modules are migrated to a new plant and reassembled due to production line adjustments, thereby achieving efficient control of reconfigurable energy storage.

[0056] S2. Input the collected photovoltaic and load sequences into the trained prediction model for processing, and use the photovoltaic and load output curves output by the prediction model to establish a set of operating scenarios.

[0057] To achieve precise competition among the connected energy storage modules, this embodiment introduces a "value coefficient" as an indicator to measure the value of each module. Based on this, this step aims to construct a standardized set of typical operating scenarios covering various extreme and normal operating conditions encountered during energy storage operation.

[0058] First, photovoltaic (PV) and load sequence data will be acquired. Using a digital twin platform, output sequence data for both PV and load will be collected in real time via corresponding data interfaces. This output data will be input into a trained prediction model to obtain the output curves and error data. In this embodiment, a Long Short-Term Memory (LSTM) network is preferred to generate short-term load and PV output curves for future short periods (e.g., 30 minutes). To construct differentiated operating conditions, the uncertain parameters generated by the LSTM are used as amplitude benchmarks to inject perturbations into the PV and load output curves, generating perturbation curves. These perturbation curves are then encapsulated into a set of operating scenarios.

[0059] The uncertain parameters are determined by the prediction errors generated by the prediction model. Specifically, in this embodiment, historical load, weekday indicators, temperature, and other data are used as load training samples to train an LSTM to generate a load prediction model. Similarly, historical power, irradiance, cloud cover, and power data are used as photovoltaic training samples to train an LSTM to generate a photovoltaic prediction model. Then, the two trained LSTM models are run in parallel within a digital twin platform to predict photovoltaic power and load respectively. During this process, the prediction model generates a standard deviation or variance that characterizes the prediction uncertainty at each time point. This standard deviation is used as the magnitude of the perturbation, correspondingly perturbing the generated photovoltaic output curve and load output curve. At time t, the baseline photovoltaic output value = 0.5, and the standard deviation σ = 0.05. The system may generate a set of perturbed photovoltaic output values ​​such as [0.425, 0.455, 0.5, 0.545, 0.575].

[0060] In other embodiments of the present invention, real-time electricity price data is obtained by connecting a digital twin platform to an electricity trading center, and the data is input into an ARIMA time series prediction model to generate a real-time electricity price curve for a target period. Similarly, a standard deviation of the ARIMA output is used as an uncertainty parameter to inject a disturbance into the real-time electricity price curve.

[0061] S3. Based on the set of operating scenarios and operating parameters, calculate the value coefficient of each access energy storage module reflecting cycle life loss under different operating scenarios.

[0062] It should be understood that under certain extreme operating conditions, the high internal resistance of energy storage can lead to overheating and a sharp decline in lifespan. In energy storage systems, the cycle life of batteries is a finite and non-renewable core asset. Every charge and discharge cycle depletes this asset.

[0063] Based on this, this embodiment quantifies the "value" of each connected energy storage module by analyzing the level of energy storage cycle life loss. Specifically, a cycle life prediction model is first trained using pre-constructed historical energy storage operation data and a set of operation scenarios as input. The cycle life prediction model can be a deep feedforward neural network, such as a multilayer perceptron or residual network; in this embodiment, it is denoted as the V-Net network. During training, feature vectors extracted from the historical energy storage operation data are input into the V-Net network for training, including the energy storage's SOH, SOC, internal resistance, and temperature.

[0064] In each operating scenario, the vector representation of the real-time operating parameters of each access energy storage module is input into the cycle life prediction model for simulation and deduction. The output is the life loss of each access energy storage module in the current operating scenario. The life loss corresponding to all operating scenarios is accumulated to determine the value coefficient corresponding to each access energy storage module.

[0065] This value coefficient V pre This reflects the equivalent cycle life loss rate generated by the connected energy storage module performing one standard charge-discharge cycle. The following formula shows its mapping relationship with key parameters:

[0066]

[0067] in, σ is the state vector of energy storage module i at time t (including SOH, SOC, and internal resistance), σ is the uncertainty parameter, and H is the historical energy storage operation data.

[0068] S4. Define the revenue function based on the value coefficient, calculate the Shapley value of each access energy storage module using the revenue function, and initiate a game election to determine the master control module.

[0069] After determining the value coefficients, a group game among the energy storage modules will be conducted. This embodiment preferably uses a game architecture driven by digital twins at preset intervals (e.g., minutes) to determine the Shapley Value (SV). In this architecture, the payoff function is improved using the value coefficients. Specifically, the payoff function v(S) is designed as a three-term synergy, expressed by the following formula:

[0070]

[0071] In the formula, v raw ( S ) for the alliance S Comprehensive energy storage efficiency; V pre ( i ) is an energy storage module i Pre-access value coefficient;β 1 is V pre Weighting factors; R (S, σ ) is the alliance S under uncertainty parameters σ The risk loss function is used to quantify the performance loss that alliance S may face in uncertain scenarios; λ This is the risk aversion coefficient, dynamically determined based on historical assessments of environmental uncertainty by the digital twin platform. Specifically, the platform calculates in real time the normalized prediction error variance between the photovoltaic and load forecast sequences and actual operating data within a set time window (e.g., 24 hours). The system presets a positive correlation mapping function between the error variance and λ; when historical prediction errors fluctuate significantly, it automatically increases the value of λ to improve risk aversion. R (S, σ The weight of λ is adjusted to favor conservative strategies in the game to avoid potential default risks caused by uncertainty; conversely, when the forecast data is stable, the value of λ is reduced to encourage the pursuit of higher energy storage efficiency.

[0072] Specifically, in this embodiment, an alliance is formed by any one or more access energy storage modules. In the digital twin platform, the sum of the maximum chargeable and dischargeable power and the average response speed of each access energy storage module in the current state of charge are obtained. The energy storage efficiency term is defined as the sum of the maximum chargeable and dischargeable power and the average response speed.

[0073] for R (S, σ In this embodiment, within the digital twin platform, each alliance is placed under different operating scenarios for simulation. Under each operating scenario, the efficiency loss generated by the alliance within the set scheduling cycle is calculated, such as the additional lifespan reduction of modules within the alliance due to fluctuations in the operating scenario, or the tracking deviation of power grid dispatch instructions. Based on these simulation results, the efficiency loss function of each alliance is established.

[0074] After determining the above two items, assign weights to the value coefficients. β The revenue function is determined by the coordinated energy storage efficiency term and efficiency loss function. Based on this, high... V pre High-quality energy storage modules can join any alliance S This will significantly increase the total value of the alliance. v ( S This involves increasing the Shapley value during the game. In game theory, v(S) represents the total value that a "storage module alliance (subset S)" can generate as a whole, while SV(i) represents the fair allocation value that "individual module i" should receive in this game, specifically expressed as follows:

[0075] In the game-theoretic election process, this embodiment is designed with two stages. Specifically: In the first stage, each access energy storage module bids according to its corresponding Shapley value in a Vickrey auction, and the access energy storage module with the highest bid is initially selected as the master control module.

[0076] Next, the second phase of execution is triggered: each non-master control module in the access energy storage module initiates competition with the master control module, and the access energy storage module that wins the competition is replaced by the master control module. That is, other access energy storage modules that are not master control modules are allowed to "compete" with this master control module in the second phase. This competition mechanism is based on the benefit value generated by the power allocation strategies of both modules. Specifically, the first benefit function value of the alliance is calculated under the first power allocation strategy issued by the current master control module. Then, during the competition process, each non-master control module submits its corresponding second power allocation strategy and calculates the second benefit function value of the alliance under the second power allocation strategy. Finally, the access energy storage module whose second benefit function value is higher than the first benefit function value is replaced by the master control module.

[0077] S5. Drive each access energy storage module to solve the optimization problem generated by the main control module using a distributed negotiation algorithm with embedded value coefficients, and output the target power control strategy.

[0078] In this embodiment, with the establishment of the main control module, the power control of the energy storage group under the multi-objective optimization problem will begin.

[0079] At this point, the main control module extracts the scheduling tasks within the target control cycle from the digital twin platform, transforming the scheduling tasks into an optimization problem with the objectives of minimizing lifetime decay and minimizing grid dispatch command tracking error, and using the output of each connected energy storage module as the decision variable. For example, the scheduling tasks for the next control cycle (5 minutes) can be extracted and deconstructed into a multi-objective optimization problem that includes minimizing the lifetime decay (f1) and minimizing the grid dispatch command tracking error (f2).

[0080] This embodiment preferably uses the Alternating Direction Multiplier Method (ADMM) as the distributed negotiation algorithm. To resolve conflicts of interest among multiple modules, each energy storage module is mapped in the digital twin space as a rational intelligent agent with independent computing capabilities. Furthermore, the value coefficient is embedded as a penalty factor into the Lagrangian function of the preset negotiation algorithm, resulting in an improved distributed negotiation algorithm. γ i Specifically, it is expressed as follows:

[0081] γ i = 1 / (1+ V pre,i )

[0082] The above optimization problem is solved iteratively using this distributed negotiation algorithm. During the solution process, V pre Lower-ranking modules are given larger penalty weights. γ i This means that the decision variables in this optimization problem x i If the variable deviates from the global consensus variable z, it will face a huge loss function penalty, forcing it to abandon its "lazy" intention and strictly converge to the global optimal solution to prioritize satisfying the power grid service objective f2; conversely, V pre Higher-quality modules have smaller penalty constraints, and the algorithm allows them to retain more degrees of freedom that deviate from the global mean, thus prioritizing their own lifetime protection target f1.

[0083] Furthermore, the algorithm stops iterations based on both the original residual and the dual residual, and a convergence threshold is set as follows: =10 -4 When the iteration error meets the condition, for example: when the current power output x of all energy storage modules... i The difference (norm) between the global consensus variable z and the global consensus variable z (less than) And the change (norm) of the global variable z itself between the two iterations. It is also smaller than When the time is right, the algorithm terminates and outputs a set of Pareto optimal power allocation vectors. For example, if the total scheduling demand is 100kW, it is shared by module A (new battery, high Vpre) and module B (old battery, low Vpre). After solving using ADMM, a set of target power control strategies is finally output. P ModA =40kW, P ModB =60kW. This strategy is used to guide each connected energy storage module to execute actual charging and discharging power commands, thereby achieving power control of reconfigurable energy storage.

[0084] In some embodiments of the present invention, an objective that includes maximizing economic benefits (f3) may also be introduced. In this case, the real-time electricity price curve previously obtained by the ARIMA model needs to be taken into account in the process of constructing the set of operating scenarios, analyzing the value coefficient, and defining the revenue function.

[0085] In one implementation of this embodiment, when maintenance personnel initiate a disconnection request for any target access module "ModY", the networked physical system (CPS) collaborative disconnection procedure is immediately initiated. At this time, the main control module first instructs the digital twin platform to mark "ModY" as pending disconnection in virtual space. When "ModY" is physically removed, a "communication interruption signal" is generated, and the digital twin platform will immediately determine the physical topology change of the reconfigurable energy storage system. Since the Shapley value (SV) is a relative indicator reflecting the "marginal contribution of individuals in the alliance", when the number of modules changes from N to N-1, that is, the number of topology nodes decreases from N to N-1, the topology structure and total value function v(S) of the entire alliance change. In order to satisfy the output of the global Pareto optimal solution by the distributed negotiation algorithm, the game election will be forcibly retried.

[0086] It is worth noting that after each energy storage module executes the instructions of the target power control strategy, the actual operating data of the reconstructable energy storage system (covering the real SOC evolution trajectory, electrochemical decay rate, and power execution deviation, etc.) is collected in real time and fed back to the digital twin platform. This data is then compared with the prediction results of V-Net before the strategy was executed to generate quantified prediction errors and value biases. Based on this, a deep reinforcement learning algorithm is introduced, with the optimization guideline being the minimization of the comprehensive objective function L, which includes the sum of squared biases and the cost of strategy correction. Backpropagation is dynamically executed to correct the internal parameters of V-Net. The comprehensive objective function L is specifically expressed as follows:

[0087]

[0088] In the formula, ΔV represents the value realization deviation, and Δσ represents the risk prediction error. C cost Let α, β2, and γ be the cost of adjusting the game strategy, and let α, β2, and γ be the weighting coefficients.

[0089] In summary, this embodiment achieves real-time mapping of reconfigurable energy storage through a digital twin platform, utilizes an embedded prediction model to analyze the output of photovoltaics and loads to construct a set of operating scenarios, calculates the value coefficient of each energy storage module under different operating scenarios, and this coefficient fully reflects the cycle life loss of the energy storage module; it also deeply couples the value coefficient into the alliance payoff function of the Shapley game, and initiates a master control module election mechanism based on the reconfigured Shapley value; subsequently, in the power allocation stage, it solves the optimization problem by embedding the value coefficient as a penalty factor into the alternating direction multiplier method, and finally outputs an energy storage power control strategy that takes into account both global optimization and individual protection.

[0090] One embodiment of the present invention provides a reconfigurable energy storage control system based on group game theory. For details, please refer to [link to relevant documentation]. Figure 2 , Figure 2The diagram shown illustrates a reconfigurable energy storage control system based on group game theory in one embodiment of the present invention, comprising:

[0091] Deployment module M1 is used to deploy a digital twin platform mapped to the reconfigurable energy storage system and synchronously collect the operating parameters of each connected energy storage module;

[0092] The scenario establishment module M2 is used to input the collected photovoltaic and load sequences into the trained prediction model for processing, and to establish a set of operating scenarios by outputting the photovoltaic and load output curves from the prediction model.

[0093] The value coefficient determination module M3 is used to calculate the value coefficient of each access energy storage module reflecting cycle life loss under different operating scenarios based on the operating scenario set and the operating parameters.

[0094] The game election module M4 is used to define a payoff function based on the value coefficient, calculate the Shapley value of each of the access energy storage modules using the payoff function, and initiate a game election to determine the main control module.

[0095] The strategy generation module M5 is used to drive each of the access energy storage modules to solve the optimization problem generated by the main control module using a distributed negotiation algorithm embedded with the value coefficient, and output the target power control strategy.

[0096] like Figure 3 As shown, this embodiment of the invention also provides a computer device. Figure 3 This is a structural block diagram of a preferred embodiment of a computer device provided by the present invention. The computer device includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the method described above.

[0097] Preferably, the computer program can be divided into one or more modules / units (such as computer program 1, computer program 2, ...), and the one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program in the computer device.

[0098] The processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor, or the processor can be any conventional processor. The processor is the control center of the terminal device, connecting various parts of the terminal device through various interfaces and lines.

[0099] The memory mainly includes a program storage area and a data storage area. The program storage area can store the operating system, applications required for at least one function, etc., while the data storage area can store related data, etc. Furthermore, the memory can be a high-speed random access memory, or a non-volatile memory, such as a plug-in hard drive, a SmartMedia Card (SMC), a Secure Digital (SD) card, and a Flash Card, or other volatile solid-state storage devices.

[0100] It should be noted that the aforementioned terminal devices may include, but are not limited to, processors and memory, as will be understood by those skilled in the art. Figure 3 The structural block diagram is merely an example of a terminal device and does not constitute a limitation on the terminal device. It may include more or fewer components than shown, or combine certain components, or use different components. Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium may be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0101] Accordingly, embodiments of the present invention provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the steps in the method of the above embodiments, for example... Figure 1 Steps S1 to S5 as described above.

[0102] The technical features and effects of the reconfigurable energy storage control system based on group game theory proposed in this embodiment of the invention are the same as those of the reconfigurable energy storage control method based on group game theory proposed in this embodiment of the invention, and will not be repeated here.

[0103] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A reconfigurable energy storage control method based on group game theory, characterized in that, include: A digital twin platform that maps to a reconfigurable energy storage system is deployed to synchronously collect the operating parameters of each connected energy storage module; The collected photovoltaic and load sequences are input into the trained prediction model for processing, and the photovoltaic and load output curves output by the prediction model are used to establish a set of operating scenarios. Based on the set of operating scenarios and the operating parameters, the value coefficient reflecting the cycle life loss of each access energy storage module under different operating scenarios is calculated; specifically: using pre-constructed historical energy storage operating data and the set of operating scenarios as input, a cycle life prediction model is trained; under each operating scenario, the operating parameters corresponding to each access energy storage module are input into the cycle life prediction model for simulation and deduction, and the life loss of each access energy storage module under the current operating scenario is output. Accumulate the lifespan loss corresponding to all operating scenarios to determine the value coefficient corresponding to each of the access energy storage modules; A revenue function is defined based on the value coefficient. The Shapley value of each access energy storage module is calculated using the revenue function, and a game-theoretic election is initiated to determine the master control module. Specifically, each access energy storage module bids with the corresponding Shapley value in a Vickrey auction, and the access energy storage module with the highest bid is initially elected as the master control module. Each non-master control module among the access energy storage modules competes with the master control module, and the access energy storage module that wins the competition is replaced by the master control module. The system drives each of the access energy storage modules to solve the optimization problem generated by the main control module using a distributed negotiation algorithm embedded with the value coefficient, and outputs a target power control strategy.

2. The reconfigurable energy storage control method based on group game theory as described in claim 1, characterized in that, The establishment of an operating scenario set based on the photovoltaic and load output curves output by the prediction model includes: Obtain the prediction error generated by the prediction model and determine the uncertainty parameters; The uncertain parameters are used as the amplitude reference to inject disturbances into the photovoltaic and load output curves to generate disturbance curves; The disturbance curve is encapsulated into the set of operating scenarios.

3. The reconfigurable energy storage control method based on group game theory as described in claim 1, characterized in that, The definition of the revenue function based on the value coefficient includes: An alliance is formed by any one or more of the access energy storage modules, and the sum of the maximum chargeable and dischargeable power and the average response speed of each access energy storage module in the current state of charge are obtained. The sum of the maximum chargeable and dischargeable power and the average response speed are jointly used as the energy storage efficiency item. Each alliance was simulated under different operating scenarios, and the performance loss function of each alliance was established based on the simulation results. The value coefficient is weighted and coordinated with the energy storage efficiency term and the efficiency loss function to determine the revenue function.

4. The reconfigurable energy storage control method based on group game theory as described in claim 1, characterized in that, Each non-master control module initiates competition with the master control module, and the successfully competing access energy storage module is replaced by the master control module, including: Calculate the first benefit function value of the alliance under the first power allocation strategy issued by the current main control module; During the competition, each non-master module submits its corresponding second power allocation strategy and calculates the alliance's second revenue function value under the second power allocation strategy. The access energy storage module whose second revenue function value is higher than the first revenue function value is replaced by the main control module.

5. The reconfigurable energy storage control method based on group game theory as described in claim 1, characterized in that, The driving mechanism for each of the access energy storage modules utilizes a distributed negotiation algorithm embedded with the value coefficient to solve the optimization problem generated by the main control module, including: The main control module extracts the scheduling task within the target control cycle from the digital twin platform, and transforms the scheduling task into an optimization problem with the objectives of minimizing lifetime decay and minimizing grid dispatch command tracking error, and with the output of each connected energy storage module as the decision variable. The value coefficient is embedded as a penalty factor into the Lagrange function of the preset negotiation algorithm to obtain a distributed negotiation algorithm, which is then used to iteratively solve the optimization problem.

6. A reconfigurable energy storage control system based on group game theory, characterized in that, include: The deployment module is used to deploy a digital twin platform that maps to the reconfigurable energy storage system and synchronously collect the operating parameters of each connected energy storage module. The scenario establishment module is used to input the collected photovoltaic and load sequences into the trained prediction model for processing, and to establish a set of operating scenarios by using the photovoltaic and load output curves output by the prediction model. The value coefficient determination module is used to calculate the value coefficient reflecting cycle life loss of each access energy storage module under different operating scenarios based on the operating scenario set and the operating parameters. Specifically, a cycle life prediction model is trained using pre-constructed historical energy storage operation data and the operating scenario set as input. Under each operating scenario, the operating parameters corresponding to each access energy storage module are input into the cycle life prediction model for simulation and deduction, and the life loss of each access energy storage module under the current operating scenario is output. Accumulate the lifespan loss corresponding to all operating scenarios to determine the value coefficient corresponding to each of the access energy storage modules; The game-theoretic election module is used to define a payoff function based on the value coefficient, calculate the Shapley value of each of the access energy storage modules using the payoff function, and initiate a game-theoretic election to determine the master control module. Specifically, each of the access energy storage modules bids with the corresponding Shapley value in a Vickrey auction, and the access energy storage module with the highest bid is initially elected as the master control module. Each non-master control module among the access energy storage modules competes with the master control module, and the access energy storage module that wins the competition is replaced by the master control module. The strategy generation module is used to drive each of the access energy storage modules to solve the optimization problem generated by the main control module using a distributed negotiation algorithm embedded with the value coefficient, and output the target power control strategy.

7. A computer device, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the reconfigurable energy storage control method based on group game theory as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the device containing the computer-readable storage medium executes the computer program, it implements the reconfigurable energy storage control method based on group game theory as described in any one of claims 1 to 5.

Images