A control method, device, equipment and medium for a distributed energy storage system

Abstract

The present invention discloses a control method, device, equipment and medium for a distributed energy storage system, which relates to the field of distributed energy storage technology and includes the following steps: obtaining the global control task of the energy storage system and decoupling the global control task into multiple control fragments; constructing a topology mapping diagram based on the first data of the energy storage system, marking a set of migratable nodes for each control fragment; when a node control failure or a communication link is detected, triggering fragment drift based on the status of the observer and the communication link between the nodes, and continuously monitoring the control effect; when the original control node resumes communication and operation, the fragment control right is recovered based on the control effect. The present invention automatically triggers the drift of control fragments, and the candidate node takes over the control task to ensure that the system control function is not interrupted. It significantly improves the fault self-healing capability and control continuity of the distributed energy storage system, and breaks through the vulnerability problem of the existing centralized control architecture under communication failure.

Description

Technical Field

[0001] The present invention relates to the field of distributed energy storage technology, and more specifically, to a control method, device, equipment and medium for a distributed energy storage system. Background Art

[0002] As the proportion of renewable energy generation continues to increase, the operational characteristics of the power system are undergoing significant changes. The intermittent and uncertain output of clean energy sources such as wind and solar power poses numerous challenges to the stable operation and dispatch of the power grid. To enhance the flexibility and reliability of the power system, distributed energy storage systems are becoming an increasingly important component. They not only enable time-shifting energy management and smooth out fluctuations in renewable energy output, but also provide auxiliary services such as voltage support, frequency regulation, and peak-load shifting, enhancing the intelligence and adaptability of the distribution system.

[0003] Traditional energy storage control methods rely on centralized management models, resulting in long communication and control paths and slow system response. These methods are ill-suited to the large number and geographical dispersion of distributed energy storage devices. With the advancement of information and communication technologies, a growing number of researchers are exploring localized and collaborative control strategies tailored to distributed structures. These approaches emphasize the autonomy and collaboration of energy storage units, aiming to achieve stable and economical global system operation while ensuring optimized local operation.

[0004] The above-mentioned disclosed technical solutions have at least the following technical problems: In a distributed energy storage system, node control rights are usually bound (one node controls one local device). Once the node fails (communication interruption, local failure, etc.), the control chain is broken and the control task cannot be continued. The existing technology can only perform "rescheduling" or "migration", which is still based on the "overall transfer" thinking, is slow and lacks continuity.

[0005] In view of the above problems, the present invention proposes a solution. Summary of the Invention

[0006] To overcome the above-mentioned shortcomings of the prior art, embodiments of the present invention provide a control method, apparatus, device, and medium for a distributed energy storage system. By constructing a heterogeneous graph to mark the set of migratable nodes of the control fragments, and integrating the communication quality index with an extended state observer, early identification of node communication disconnection or control failure is achieved, thereby solving the problem of node failure, control chain breakage, and inability to continue to execute control tasks in the prior art.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A control method for a distributed energy storage system includes the following steps: obtaining a global control task for the energy storage system and decoupling the global control task into multiple control fragments; constructing a topology map based on first data of the energy storage system, marking a set of migratable nodes for each control fragment, wherein the first data includes electrical connection relationship, communication quality, power margin and control coupling degree; triggering fragment drift based on an observer and a communication anomaly recognition algorithm when node control failure or communication link disconnection is detected, and continuously monitoring the control effect; and recovering the fragment control right based on the control effect after the original control node resumes communication and operation.

[0009] In a preferred embodiment, the global control task of the energy storage system is obtained and the global control task is decoupled into multiple control fragments, specifically: second data of the energy storage system is obtained, and a global control task list for the current period is generated based on the second data, wherein the second data includes real-time operating status, grid dispatch instructions, and load forecast results; a task matrix is ​​constructed based on the global control task list, and the task matrix is ​​decomposed into multiple control fragments based on a task decoupling mechanism; fragment metadata is labeled for each control fragment, and the control fragment is mounted on each energy storage node.

[0010] In a preferred embodiment, a topological map is constructed based on the first data of the energy storage system, and a set of migratable nodes is marked for each control fragment, specifically: a capability vector is constructed for each energy storage node based on the first data of the energy storage system; a resource requirement vector is constructed for each control fragment; based on the matching degree between the node capability vector and the fragment requirement vector, a fitness score between the control fragment and each energy storage node is calculated, and a heterogeneous graph is constructed based on the current communication topology; a fitness threshold and a topological distance are set in the heterogeneous graph, nodes that meet the fitness conditions are searched, and a set of migratable nodes for the fragment is generated, wherein the fitness conditions are met as follows: the calculated fitness score is greater than the preset fitness threshold, the communication path is reachable, and the topological distance requirements are met.

[0011] In a preferred embodiment, when node control failure or communication disconnection is detected, fragment drift is triggered based on the observer and communication anomaly identification algorithm, specifically: periodically collect status information of the communication link between nodes and calculate the communication quality index; deploy a dual-modal state perception mechanism that integrates the gradient field and the extended state observer at each energy storage node, dynamically select the observation mode according to the communication quality index, and determine whether the control performance is abnormal; if the observation residual is greater than the preset residual threshold within a preset number of sampling cycles, or the communication quality index is continuously lower than the preset communication threshold and the maintenance time exceeds the preset duration, it is determined that the node has control failure or communication disconnection, and fragment drift is triggered.

[0012] In a preferred embodiment, the observation mode is dynamically selected according to the communication quality index, specifically: when the communication quality index is greater than a preset first threshold, the extended state observer mode is selected to continuously monitor the key operating status of the node, and the observation residual is calculated in combination with the neighborhood state information; when the communication quality index is less than the preset first threshold, the gradient perception mode is switched to calculate the potential energy gradient based on the voltage phase difference for sensing implicit control anomalies; if the observation residual continues to exceed the set residual threshold, or the potential energy gradient continues to increase and tends to be unstable, it is determined that the node control performance has degraded.

[0013] In a preferred embodiment, the triggering of fragment drift is specifically as follows: querying the list of control fragments mounted on the abnormal node, and calling the corresponding set of migratable nodes in the mapping diagram for each fragment, and recalculating the fitness score of each candidate node; sorting the candidate nodes from high to low according to the fitness score, generating a fragment migration scheduling plan, and preparing the control metadata and context state required for migration; initiating a control right declaration or broadcast to the target node based on the fragment migration scheduling plan, and having its controller mount the corresponding fragment and start execution to complete the local takeover of the control function.

[0014] In a preferred embodiment, the continuous monitoring of the control effect is specifically as follows: after the fragment migration is completed, the historical trajectory data of the potential energy gradient field is continuously tracked, the trajectory is real-time feature recognition is performed based on a one-dimensional convolutional neural network, and the second-order curvature of the potential energy field is calculated. If the rate of change of the second-order curvature is less than the rate of change threshold for several consecutive sampling periods, it is determined that the system has entered a steady-state controllable region; according to the potential energy field characteristics when the system enters the steady-state controllable region, the optimal control parameters are matched based on a pre-built potential energy strategy mapping library.

[0015] The technical effects and advantages of the control method, device, equipment and medium of a distributed energy storage system of the present invention are as follows:

[0016] 1. The present invention constructs a heterogeneous graph based on a topology map and capability vector matching mechanism to mark the set of migratable nodes of the control fragments, and integrates the communication quality index and the extended state observer to achieve early identification of node communication disconnection or control failure. In the event of a fault, the system can automatically trigger the drift of the control fragments, and the candidate node will take over the control task to ensure that the system control function is not interrupted. This mechanism has high robustness and self-recovery capabilities, significantly improving the fault self-healing capability and control continuity of the distributed energy storage system, and breaking through the vulnerability of the existing centralized control architecture under communication failures.

[0017] 2. The present invention introduces the potential energy gradient field as a global state perception indicator, uses a one-dimensional convolutional neural network to perform feature recognition on the potential energy trajectory during the system recovery process, and determines whether the system is in a steady-state evolution stage by calculating the second-order curvature change rate. At the same time, a "potential energy-strategy mapping library" based on t-SNE dimensionality reduction and spectral clustering is constructed, which can automatically match the optimal control strategy according to the current system evolution characteristics. After the original control node resumes communication, the system makes an intelligent decision based on the comprehensive control scoring function whether to reclaim the control right of the fragment, avoid control oscillation, and achieve optimal recovery of fragments and closed-loop reconstruction of the system. This method breaks through the passive mode of "restoration after failure" in traditional control systems and improves the decision-making intelligence and stability of control migration. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A schematic flow chart of a control method for a distributed energy storage system according to the present invention;

[0019] Figure 2 This is a structural schematic diagram of a control device for a distributed energy storage system according to the present invention. DETAILED DESCRIPTION

[0020] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0021] Example 1, Figure 1 The present invention provides a control method for a distributed energy storage system, comprising the following steps:

[0022] S1, obtains the global control task of the energy storage system and decouples the global control task into multiple control fragments;

[0023] In this embodiment, the global control tasks of the distributed energy storage system include but are not limited to: frequency support, voltage regulation, active and reactive power distribution, SOC (state of charge) balancing, and black start coordination functional modules.

[0024] The global control task of the energy storage system is obtained and decoupled into multiple control fragments, specifically:

[0025] Acquire second data of the energy storage system, and generate a global control task list for the current period based on the second data, wherein the second data includes real-time operating status, grid dispatch instructions, and load forecast results;

[0026] Build a task matrix based on the global control task list, and decompose the task matrix into multiple control fragments based on the task decoupling mechanism;

[0027] Label each control fragment with fragment metadata and mount the control fragment on each energy storage node.

[0028] In practical applications, global tasks can be fragmented using methods based on graph partitioning or functional matrix decomposition. For example, in a voltage regulation task, multiple control fragments can be created based on node voltage sensitivity, with each fragment controlling a cluster of nodes with tight voltage coupling.

[0029] The task list is represented by a modular modeling approach, and each control task has clear input and output parameters, objective functions, and scopes.

[0030] Each control fragment represents a micro-control subtask with independent operation logic and has the following characteristics:

[0031] Functional independence: Minimize dependencies between control fragments. For example, voltage regulation and SOC balancing can belong to different fragments.

[0032] Migration: Each control fragment has standardized data interfaces and execution conditions, facilitating dynamic migration between multiple nodes.

[0033] Resource adaptability: When generating control fragments, record their demand levels for computing resources, communication bandwidth, and data sampling rate.

[0034] S2, constructing a topology map based on the first data of the energy storage system, marking a set of migratable nodes for each control fragment;

[0035] The first data includes electrical connection relationship, communication quality, power margin and control coupling degree.

[0036] The topology map is constructed based on the first data of the energy storage system, and a set of migratable nodes is marked for each control fragment, specifically:

[0037] constructing a capability vector for each energy storage node based on first data of the energy storage system;

[0038] Construct resource demand vector for each control fragment;

[0039] Based on the matching degree between the node capability vector and the fragment demand vector, the compatibility score between the control fragment and each energy storage node is calculated, and a heterogeneous graph is constructed based on the current communication topology;

[0040] The fitness threshold and topological distance are set in the heterogeneous graph, nodes that meet the fitness conditions are searched, and a fragmented set of migratable nodes is generated.

[0041] The nodes that meet the adaptability condition are specifically:

[0042] The nodes whose calculated fitness score is greater than the preset fitness threshold and whose communication path is reachable and meets the topological distance requirements are selected.

[0043] The heterogeneous graph includes:

[0044] Fragment layer: represents all control fragments to be mapped;

[0045] Node layer: represents all energy storage control nodes;

[0046] Fragment-node edge (heterogeneous edge): represents the adaptation relationship, and the edge weight is the adaptation score;

[0047] Node-node edge (communication edge): represents the communication path and distance between nodes, and the edge weight represents the number of hops, delay, or packet loss rate.

[0048] The capability vector includes voltage / current regulation capability, remaining active / reactive regulation margin, current SOC level and dynamic change rate, control thread resources and computing processing capability, and communication link quality (bandwidth, latency, stability).

[0049] The resource requirement vector includes control accuracy requirements, response time requirements, power regulation strength, and the degree of dependence on communication interaction frequency and bandwidth.

[0050] The specific calculation formula for the fitness score is:

[0051]

[0052]

[0053]

[0054] in, Score the fit. As the basic matching degree, is the weight coefficient of conflict penalty intensity, is the conflict penalty term, is the importance weight of the i-th resource dimension, is the demand of fragment f on the i-th resource, is the performance of node n in the i-th capability dimension, is the total dimension of the resource vector, To control the fragment identification, is the energy storage node identifier, For fragment f and The degree of conflict between The set of shards that have been run on the current node n.

[0055] S3, when node control failure or communication link disconnection is detected, fragment drift is triggered based on the status of the observer and the communication link between nodes, and the control effect is continuously monitored;

[0056] When node control failure or communication disconnection is detected, fragment drift is triggered based on the status of the observer and the communication link between the nodes, and the control effect is continuously monitored. Specifically:

[0057] Periodically collects status information of communication links between nodes and calculates the communication quality index, taking into account packet loss rate, delay fluctuation, and ACK response status to determine whether the communication link is broken or degraded;

[0058] A dual-modal state perception mechanism integrating gradient field and extended state observer is deployed at each energy storage node. The observation mode is dynamically selected based on the communication quality index to determine whether the control performance is abnormal.

[0059] If the observed residual is greater than the preset residual threshold within a preset number of sampling cycles, or the communication quality index is continuously lower than the preset communication threshold and the maintenance time exceeds the preset duration, it is determined that the node has control failure or communication link disconnection, triggering fragment drift.

[0060] The dynamic selection of the observation mode according to the communication quality index is specifically as follows:

[0061] When the communication quality index is greater than a preset first threshold, an extended state observer model is deployed at each energy storage node to continuously monitor the key operating status of the node, including but not limited to voltage, current, active / reactive power, and frequency change rate, and calculate the observation residual in combination with the neighborhood state information;

[0062] When the communication quality index is less than a preset first threshold, the system switches to a gradient sensing mode and calculates the potential energy gradient based on the voltage phase difference to sense implicit control anomalies.

[0063] If the observed residual continues to exceed the set residual threshold, or the potential energy gradient continues to increase and tends to be unstable, it is determined that the node control performance has degraded.

[0064] The triggering of debris drift is specifically:

[0065] Query the list of control fragments mounted on the abnormal node, call the corresponding set of migratable nodes in the mapping graph for each fragment, and recalculate the fitness score of each candidate node;

[0066] Sort candidate nodes from high to low based on their fitness scores, generate a shard migration schedule, and prepare the control metadata and context state required for migration;

[0067] Based on the fragment migration scheduling plan, a control right declaration or broadcast is initiated to the target node, and its controller mounts the corresponding fragment and starts execution to complete the local takeover of the control function.

[0068] The continuous monitoring and control effect is specifically:

[0069] After the debris migration is completed, the historical trajectory data of the potential energy gradient field is continuously tracked. The trajectory is recognized in real time based on a one-dimensional convolutional neural network, and the second-order curvature of the potential energy field is calculated. If the rate of change of the second-order curvature is less than the rate of change threshold for several consecutive sampling periods, it is determined that the system has entered the steady-state controllable region.

[0070] According to the potential energy field characteristics (including trend slope, oscillation frequency, and energy concentration) when the system enters the steady-state controllable zone, the optimal control parameters are matched based on the pre-built potential energy strategy mapping library.

[0071] The potential energy strategy mapping library is specifically constructed as follows:

[0072] Collect potential energy gradient field distribution data under historical fault scenarios;

[0073] The t-SNE algorithm is used to perform nonlinear dimensionality reduction on the high-dimensional feature vectors of the historical potential energy gradient field to obtain a low-dimensional representation in the embedding space;

[0074] Based on low-dimensional representation, similar potential energy trajectories are divided through spectral clustering algorithm, typical feature areas are identified, and the mapping relationship between each area and the control strategy is established.

[0075] The observation residual is specifically:

[0076]

[0077] The communication quality index is specifically:

[0078]

[0079] in, is the observation residual, is the state prediction value, is the measured value of the state, is the communication quality index, and is the preset scale factor, is the packet loss rate, is the average delay.

[0080] S4: When the original control node resumes communication and operation, the fragment control rights are recovered based on the control effect.

[0081] When the original control node resumes communication and operation, the fragment control right is recovered based on the control effect, specifically:

[0082] Based on the observer and communication link monitoring, it is determined whether the original node meets the recyclable state;

[0083] If the recyclable state is met, obtain the control effect of the node where the current fragment is located during the local takeover process, and build a comprehensive control scoring function based on the historical control effect parameters of the original node;

[0084] The system scheduler compares the comprehensive control scores of the current fragment execution node and the original node based on the comprehensive control scoring function;

[0085] If the original node has a higher comprehensive control score, the fragment control is transferred back to the original node and the current node thread is unloaded;

[0086] Update the topology structure and control fragment mapping relationship to complete the system closed-loop reconstruction.

[0087] The comprehensive control scoring function is specifically:

[0088]

[0089] in, is the comprehensive control score, 、 、 are the corresponding weight ratios, is the power balance index in the control effect, is the steady-state deviation in the historical control effect parameter, The system response time.

[0090] Example 2, Figure 2 A schematic diagram of the structure of a control device for a distributed energy storage system is given, which is characterized by including the following modules:

[0091] Control fragment building module: used to obtain the global control task of the energy storage system and decouple the global control task into multiple control fragments;

[0092] A fragment mapping module is configured to construct a topology map based on the first data of the energy storage system, and mark a set of migratable nodes for each control fragment;

[0093] Anomaly detection and fragment drift module: When node control failure or communication disconnection is detected, it triggers fragment drift based on the observer and communication anomaly recognition algorithm, and continuously monitors the control effect;

[0094] Control rights recovery and strategy optimization module: used to recover fragmented control rights based on the control effect after the original control node resumes communication and operation.

[0095] This embodiment includes that the electronic device may include a processor, a memory, a communication bus and a communication interface, and may also include a computer program stored in the memory and executable on the processor, such as a control program for a distributed energy storage system.

[0096] Among them, the processor is the control core (Control Unit) of the electronic device, which uses various interfaces and lines to connect the various components of the entire electronic device, and performs various functions of the electronic device and processes data by running or executing programs or modules stored in the memory, and calling data stored in the memory.

[0097] The memory includes at least one type of readable storage medium. In some embodiments, the memory may be an internal storage unit of the electronic device, such as a mobile hard disk of the electronic device. The memory may be used to store not only application software installed in the electronic device but also various data.

[0098] The communication bus is configured to implement connection and communication between the memory and at least one processor.

[0099] The communication interface is used for communication between the electronic device and other devices, and includes a network interface and a user interface.

[0100] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0101] The control program of the distributed energy storage system stored in the memory of the electronic device is a combination of multiple instructions, and when executed in the processor, the steps of the above-mentioned control method of the distributed energy storage system can be implemented.

[0102] Specifically, the specific implementation system of the processor for the above instructions can refer to the description of the relevant steps in the corresponding embodiment in the accompanying drawings, which will not be repeated here.

[0103] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor of an electronic device, it can implement the steps in the above-mentioned control method of a distributed energy storage system.

[0104] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters in the formulas are set by technicians in this field according to actual conditions.

[0105] The above embodiments may be implemented in whole or in part through software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0106] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0107] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0108] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0109] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A control method for a distributed energy storage system, characterized in that: The following steps are involved: Obtaining a global control task for the energy storage system and decoupling the global control task into multiple control fragments, each representing a micro-control subtask with independent operation logic; Building a topology map based on first data of the energy storage system, marking a set of migratable nodes for each control fragment, wherein the first data includes electrical connection relationship, communication quality, power margin, and control coupling; When node control failure or communication disconnection is detected, fragment drift is triggered based on the status of the observer and the communication link between nodes, and the control effect is continuously monitored; The triggering of debris drift is specifically: Query the list of control fragments mounted on the abnormal node, call the corresponding set of migratable nodes in the mapping graph for each fragment, and recalculate the fitness score of each candidate node; Sort candidate nodes from high to low based on their fitness scores, generate a shard migration schedule, and prepare the control metadata and context state required for migration; Based on the shard migration schedule, a control right declaration or broadcast is initiated to the target node. The controller mounts the corresponding shard and starts execution, completing the partial takeover of the control function. When the original control node resumes communication and operation, the fragment control rights are recovered based on the control effect.

2. The control method of the distributed energy storage system according to claim 1, characterized in that: The global control task of the energy storage system is obtained and decoupled into multiple control fragments, specifically: Acquire second data of the energy storage system, and generate a global control task list for the current period based on the second data, wherein the second data includes real-time operating status, grid dispatch instructions, and load forecast results; Build a task matrix based on the global control task list, and decompose the task matrix into multiple control fragments based on the task decoupling mechanism; Label each control fragment with fragment metadata and mount the control fragment on each energy storage node.

3. The control method of the distributed energy storage system according to claim 2, characterized in that: The topology map is constructed based on the first data of the energy storage system, and a set of migratable nodes is marked for each control fragment, specifically: constructing a capability vector for each energy storage node based on first data of the energy storage system; Construct resource demand vector for each control fragment; Based on the matching degree between the node capability vector and the fragment demand vector, the compatibility score between the control fragment and each energy storage node is calculated, and a heterogeneous graph is constructed based on the current communication topology; The fitness threshold and topological distance are set in the heterogeneous graph, and nodes that meet the fitness conditions are searched to generate a set of migratable nodes of the fragment. The fitness conditions are as follows: the calculated fitness score is greater than the preset fitness threshold, the communication path is reachable, and the topological distance requirements are met.

4. The control method of the distributed energy storage system according to claim 3, characterized in that: When node control failure or communication disconnection is detected, fragment drift is triggered based on the status of the observer and the communication link between the nodes, and the control effect is continuously monitored. Specifically: Periodically collect the status information of the communication link between nodes and calculate the communication quality index; A dual-modal state perception mechanism integrating gradient field and extended state observer is deployed at each energy storage node. The observation mode is dynamically selected based on the communication quality index to determine whether the control performance is abnormal. If the observed residual is greater than the preset residual threshold within a preset number of sampling cycles, or the communication quality index is continuously lower than the preset communication threshold and the maintenance time exceeds the preset duration, it is determined that the node has control failure or communication link disconnection, triggering fragment drift.

5. The control method of the distributed energy storage system according to claim 4, characterized in that: The dynamic selection of the observation mode according to the communication quality index is specifically as follows: When the communication quality index is greater than a preset first threshold, the extended state observer mode is selected to continuously monitor the key operating status of the node and calculate the observation residual in combination with the neighborhood state information; When the communication quality index is less than a preset first threshold, the system switches to a gradient sensing mode and calculates a potential energy gradient based on the voltage phase difference to sense implicit control anomalies. If the observed residual continues to exceed the set residual threshold, or the potential energy gradient continues to increase and tends to be unstable, it is determined that the node control performance has degraded.

6. The control method of the distributed energy storage system according to claim 5, characterized in that: The continuous monitoring and control effect is specifically: After the debris migration is completed, the historical trajectory data of the potential energy gradient field is continuously tracked. The trajectory is recognized in real time based on a one-dimensional convolutional neural network, and the second-order curvature of the potential energy field is calculated. If the rate of change of the second-order curvature is less than the rate of change threshold for several consecutive sampling periods, it is determined that the system has entered the steady-state controllable region. According to the potential energy field characteristics when the system enters the steady-state controllable region, the optimal control parameters are matched based on the pre-built potential energy strategy mapping library.

7. A device using the control method of a distributed energy storage system according to any one of claims 1 to 6, characterized in that: Includes the following modules: Control fragment building module: used to obtain the global control task of the energy storage system and decouple the global control task into multiple control fragments; A fragment mapping module is configured to construct a topology map based on the first data of the energy storage system, and mark a set of migratable nodes for each control fragment; Anomaly detection and fragment drift module: When node control failure or communication disconnection is detected, it triggers fragment drift based on the observer and communication anomaly recognition algorithm, and continuously monitors the control effect; Control rights recovery and strategy optimization module: used to recover fragmented control rights based on the control effect after the original control node resumes communication and operation.

8. An electronic device, characterized in that: The electronic device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the control method of the distributed energy storage system according to any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the control method of the distributed energy storage system according to any one of claims 1 to 6 is implemented.

Images