Distributed electric power energy storage system and method

Abstract

The invention relates to the technical field of energy, and discloses a distributed electric power energy storage system and method, and the system comprises a decommissioned battery sorting and modularization module which is used for carrying out the performance evaluation, grading and standardized modularization processing of decommissioned power batteries; the energy storage unit module is used for assembling the retired batteries into an energy storage unit and storing and releasing electric energy; the energy management module is used for monitoring the running state of the energy storage unit and optimizing energy distribution; the communication network module is used for carrying out data transmission and cooperative control between the modules; and the power grid and user interface module is used for interacting the energy storage system with a power grid and a user side. According to the method, the energy scheduling unit can optimize the charging and discharging strategy of the energy storage unit in real time by combining reinforcement learning and federated learning algorithms, and meanwhile, the adaptability of a scheduling model in multiple scenes is improved by utilizing cooperative training of distributed nodes.

Description

Technical Field

[0001] The present invention relates to the field of energy technologies, and particularly to a distributed power energy storage system and method. Background Art

[0002] With the rapid development of renewable energy sources (such as photovoltaic and wind power) and the continuous growth in the number of electric vehicles, distributed energy storage systems have gradually become an important research direction in the field of energy management. By connecting dispersed energy storage units to the power grid, distributed energy storage systems can not only achieve peak shaving and valley filling for the power grid, but also improve the grid connection utilization rate of renewable energy sources, and optimize the electricity cost and reliability on the user side. In this context, retired power batteries have become an important part of distributed energy storage due to their low cost and high remaining capacity, and have great potential for cascaded utilization.

[0003] However, existing scheduling algorithms lack the adaptability to different scenarios (such as power grid load fluctuations, electricity price changes, aging of energy storage units, etc.), resulting in low energy distribution efficiency and difficulty in fully exerting the potential of energy storage units. These problems seriously restrict the performance and economic benefits of distributed energy storage systems. Summary of the Invention

[0004] To make up for the above deficiencies, the present invention provides a distributed power energy storage system and method, aiming to improve the problem that existing scheduling algorithms lack the adaptability to different scenarios (such as power grid load fluctuations, electricity price changes, aging of energy storage units, etc.).

[0005] In a first aspect, the present invention provides the following technical solution. A distributed power energy storage system includes: A retired battery sorting and modularization module for performing performance evaluation, grading, and standardized modular processing on retired power batteries; An energy storage unit module for assembling retired batteries into energy storage units and storing and releasing electrical energy; An energy management module for monitoring the operating state of energy storage units and optimizing energy distribution; A communication network module for data transmission and cooperative control between modules; A power grid and user interface module for the interaction between the energy storage system and the power grid and the user side.

[0006] Preferably, the retired battery sorting and modularization module includes: A battery performance evaluation unit for measuring the internal resistance, capacity, and voltage characteristics of retired batteries using electrochemical impedance spectroscopy technology, and evaluating the remaining life of the batteries in combination with a health state prediction model based on a neural network; Battery grading unit, which classifies the performance parameters of retired batteries through multi-sensor data fusion technology, performs weighted scoring on capacity, internal resistance and consistency indicators, and divides battery packs according to performance levels; Modular assembly unit, which assembles retired batteries with similar performance into battery modules with standardized dimensions and interfaces, and integrates thermal management and protection circuits at the same time.

[0007] Preferably, the energy storage unit module includes: Standardized battery module unit, which is composed of retired batteries with consistent performance. A battery management system BMS is embedded in the module. The BMS monitors the voltage, current and temperature of the module in real time through the CAN bus protocol and provides an equalization function to improve the consistency of the module; Bidirectional inverter unit, which adopts a two-stage conversion topology. The first stage is a DC-DC buck-boost circuit for adjusting the output voltage of the battery module, and the second stage is a DC-AC inverter for efficient conversion of direct current and alternating current; Temperature control and heat dissipation unit, which optimizes the heat dissipation structure inside the module through thermoelectric coupling simulation, configures a liquid cooling system to control the operating temperature under high load, and uses thermistors to monitor the battery temperature in real time.

[0008] Preferably, the energy management module includes: Status monitoring unit, which collects the operation data of the energy storage unit based on a distributed sensor network and preprocesses the data using edge computing nodes; Energy scheduling unit, which adopts a reinforcement learning algorithm to dynamically generate charge and discharge optimization strategies according to the status data of the energy storage unit and the grid load demand, and balances the workload of each energy storage unit through a dynamic programming model at the same time; Safety management unit, which evaluates the safety in real time through a fault prediction algorithm combined with the historical operation data of the battery module. If abnormal parameters are detected, including overcharging, overheating or voltage fluctuations, a fast circuit breaker protection mechanism is triggered to isolate the faulty unit.

[0009] Preferably, the communication network module includes: Data acquisition unit, which adopts low-power Bluetooth and ZigBee communication technologies to collect the operation data of the energy storage unit and the grid through embedded sensors; Data transmission unit, which performs multi-point to multi-point data transmission based on the MQTT protocol and combines the AES-256 encryption algorithm to ensure data security; Edge computing unit, which locally preprocesses the collected data through an ARM Cortex-A processor, including anomaly detection and trend prediction, and optimizes the charge and discharge instructions through fuzzy control logic.

[0010] Preferably, the grid and user interface module includes: The grid-connected control unit adopts a control algorithm based on a synchronous phase-locked loop (PLL) to synchronize the energy storage unit with the power grid, and at the same time uses a power factor correction circuit to optimize the grid-connected quality. The island operation control unit conducts island operation after the power grid power failure through a fast switching circuit. The switching circuit is driven by an IGBT module, making the switching time less than 20 milliseconds. The custom power optimization unit analyzes the user's electrical load using a time series prediction model, and optimizes the charge and discharge strategy of the energy storage unit through a mixed integer programming algorithm in combination with real-time electricity price fluctuations.

[0011] Preferably, the energy scheduling unit further combines the federated learning algorithm, uses multiple distributed nodes to jointly train the scheduling model, and improves the efficiency and accuracy of energy allocation in multiple scenarios.

[0012] In a second aspect, the present invention provides the following technical solution. A distributed power energy storage method includes the following steps: S1. Retired battery sorting and modular processing: Measure the remaining capacity, internal resistance, and health status of retired power batteries through electrochemical impedance spectroscopy technology. Classify the performance parameters of retired batteries based on multi-sensor data fusion technology, and grade the batteries according to capacity consistency and health status. Use automated assembly equipment to assemble batteries with similar performance into standardized battery modules, and integrate a balancing protection circuit and a thermal management system. S2. Energy storage unit deployment and operation: Integrate standardized battery modules into each energy storage unit, and equip it with a battery management system (BMS). Real-time monitor the voltage, current, and temperature of the module through the CAN bus. Adopt a bidirectional inverter to complete the efficient conversion of direct current and alternating current. The inverter includes a DC-DC buck-boost circuit and a DC-AC inverter. Use a liquid cooling system to control the temperature of the battery module, and monitor the thermal management effect through a thermistor. S3. Energy management and scheduling: Collect the operation data of the energy storage unit through a distributed sensor network, and use edge computing nodes to locally analyze the data. Use a reinforcement learning algorithm combined with a dynamic programming model to generate an optimized charge and discharge strategy for the energy storage unit, and at the same time balance the operation load of each unit. Based on a fault prediction algorithm, real-time evaluate the safety status of the energy storage unit, and trigger a fast circuit breaker protection mechanism in case of an abnormality to isolate the faulty unit. S4. Communication and data transmission: Collect the operation status data of the energy storage unit and the power grid through low-power Bluetooth and ZigBee communication technologies; Transmit the data of the energy storage unit based on the MQTT protocol, and combine the AES-256 encryption algorithm to enhance data security; Use fuzzy control logic in the edge computing node to optimize the local charge and discharge instructions, and send the data to the cloud through wireless communication; S5. Interaction between the power grid and the user side: Adopt a grid-connected control algorithm based on a synchronous phase-locked loop (PLL) to make the energy storage unit operate synchronously with the power grid, and use a power factor correction circuit to improve the grid-connected quality; In the case of a power grid power outage, trigger the island operation mode, and switch the operation mode of the energy storage unit within 20 milliseconds through a fast switching circuit; Combine the time series prediction model of the user's electricity load and the real-time electricity price fluctuation, and optimize the charge and discharge strategy of the user side through the mixed integer programming algorithm.

[0013] In a third aspect, the present invention provides the following technical solution. A computer device includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the above-mentioned distributed power energy storage method is implemented.

[0014] In a fourth aspect, the present invention provides the following technical solution. A readable storage medium stores a computer program, and when the computer program is executed by a processor, the above-mentioned distributed power energy storage method is implemented.

[0015] The present invention has the following beneficial effects: 1. In the present invention, by combining reinforcement learning and federated learning algorithms, the energy scheduling unit can optimize the charge and discharge strategy of the energy storage unit in real time. At the same time, through the collaborative training of distributed nodes, the adaptability of the scheduling model in multiple scenarios is improved. This innovative scheduling mechanism not only significantly improves the energy distribution efficiency of the system, but also can dynamically respond to load demands, electricity price fluctuations, and changes in the state of the energy storage unit, realizing intelligent and precise energy management.

[0016] 2. In the present invention, the distributed architecture design endows the system with strong fault tolerance. Each energy storage unit operates independently, and when a fault occurs, the overall stability of the system can be ensured through an intelligent monitoring and fast isolation mechanism. In addition, the thermal management system further guarantees the safety of the energy storage unit under high-load operation through liquid cooling or air cooling technologies, thus meeting the safety requirements in complex application scenarios.

[0017] 3. In the present invention, the modular design enables the energy storage system to be flexibly expanded according to requirements, and it can be easily achieved from small-scale household energy storage to large-scale power grid peak shaving. In addition, the system supports the integrated energy storage of multiple energy forms (such as photovoltaic, wind power, etc.), improving its applicability under different energy structures and geographical conditions, and providing important support for the future diversified development of energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 FIG. is a system architecture diagram of a distributed power energy storage system proposed by the present invention; Figure 2 FIG. is a sorting and modular module architecture diagram of retired batteries of a distributed power energy storage system proposed by the present invention; Figure 3 FIG. is a storage unit module architecture diagram of a distributed power energy storage system proposed by the present invention; Figure 4 FIG. is an energy management module architecture diagram of a distributed power energy storage system proposed by the present invention; Figure 5 FIG. is a communication network module architecture diagram of a distributed power energy storage system proposed by the present invention; Figure 6 FIG. is a grid and user interface module architecture diagram of a distributed power energy storage system proposed by the present invention; Figure 7 FIG. is a method flow chart of a distributed power energy storage method proposed by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0020] Embodiment 1 Referring to Figures 1-6 , in the first embodiment of the present invention, the present invention provides a distributed power energy storage system, including: A retired battery sorting and modular module for performing performance evaluation, grading, and standardized modular processing on retired power batteries; A storage unit module for assembling retired batteries into a storage unit and storing and releasing electrical energy; An energy management module for monitoring the operating state of the storage unit and optimizing energy distribution; A communication network module for data transmission and cooperative control between modules; The grid - user interface module is used for the interaction between the energy storage system, the power grid, and the user side.

[0021] Specifically, this distributed power energy storage system combines reinforcement learning and federated learning algorithms to optimize the charging and discharging strategies of energy storage units, achieving dynamic and precise energy distribution, and improving the system efficiency in multiple scenarios. The system adopts a distributed architecture design, enabling each energy storage unit to operate independently and having a strong fault - tolerance ability. Even if a single - point failure occurs, it will not affect the overall stability of the system. Combined with intelligent monitoring, thermal management, and fast isolation mechanisms, the system demonstrates excellent safety in high - load scenarios. In addition, the modular design and multi - energy compatibility provide the system with high scalability and flexibility, enabling it to adapt to various application scenarios from household energy storage to large - scale power grid peak shaving, and providing comprehensive support for diverse energy management needs.

[0022] The retired battery sorting and modularization module includes: The battery performance evaluation unit measures the internal resistance, capacity, and voltage characteristics of retired batteries using electrochemical impedance spectroscopy technology, and evaluates the remaining life of the batteries in combination with a health - state prediction model based on neural networks; The battery grading unit classifies the performance parameters of retired batteries through multi - sensor data fusion technology, performs weighted scoring on capacity, internal resistance, and consistency indicators, and divides battery groups according to performance levels; The modular assembly unit assembles retired batteries with similar performance into battery modules with standardized sizes and interfaces, and simultaneously integrates thermal management and protection circuits.

[0023] Specifically, during the implementation process, the sorting and modularization of retired batteries are completed through the collaboration of high - precision electrochemical analysis equipment and intelligent algorithms. Through electrochemical impedance spectroscopy technology, the internal resistance, capacity, and state of health (SOH) of retired batteries are measured to judge the remaining useful value of the batteries.

[0024] In the grading process, through multi - sensor data fusion technology, parameters such as the capacity, internal resistance, and cycle life of the batteries are comprehensively analyzed, and weighted scoring is calculated to ensure that batteries with consistent performance parameters are grouped into the same category. Automated assembly equipment completes the assembly of the modules according to the classification results, and integrates over - charge protection, balancing circuits, and thermal management modules within the modules to improve the safety and consistency of the modules.

[0025] The energy storage unit module includes: The standardized battery module unit is composed of retired batteries with consistent performance. A battery management system (BMS) is embedded in the module. The BMS monitors the voltage, current, and temperature of the module in real - time through the CAN bus protocol and provides an equalization function to improve the consistency of the module; The bidirectional inverter unit adopts a two-stage conversion topology. The first stage is a DC-DC buck-boost circuit used to regulate the output voltage of the battery module, and the second stage is a DC-AC inverter for efficient conversion between direct current and alternating current. The temperature control and heat dissipation unit optimizes the internal heat dissipation structure of the module through thermoelectric coupling simulation, configures a liquid cooling system to control the operating temperature under high loads, and uses thermistors to monitor the battery temperature in real time.

[0026] Specifically, the core of the energy storage unit is a standardized battery module and an efficient power conversion system. In actual applications, real-time monitoring is achieved through the embedded battery management system (BMS) inside the module. The BMS communicates with other modules and the central management system via the CAN bus protocol to obtain voltage, current, and temperature data of the module.

[0027] To improve the conversion efficiency, the inverter in the energy storage unit adopts a two-stage topology. The first-stage DC-DC buck-boost circuit is used to adjust the module output to a stable DC voltage, and the second-stage DC-AC inverter converts direct current to alternating current and supports grid connection synchronization. The liquid cooling system further enhances the heat dissipation capacity under high loads through an optimized fluid channel design, ensuring the module can operate stably in different environments.

[0028] The energy management module includes: The status monitoring unit collects the operation data of the energy storage unit based on a distributed sensor network and preprocesses the data using edge computing nodes. The energy scheduling unit adopts a reinforcement learning algorithm to dynamically generate charge and discharge optimization strategies according to the status data of the energy storage unit and the grid load demand, and balances the workload of each energy storage unit through a dynamic programming model. The safety management unit evaluates the safety in real time through a fault prediction algorithm combined with the historical operation data of the battery module. If abnormal parameters are detected, including overcharging, overheating, or voltage fluctuations, it triggers a fast circuit breaker protection mechanism and isolates the faulty unit.

[0029] Specifically, the energy management module plays a role of coordinated control in the system, and its core is distributed status monitoring and optimized scheduling. The monitoring part collects key data such as voltage, temperature, and current in real time through distributed sensors installed in each energy storage unit. These data are transmitted to the edge computing nodes to complete the preliminary analysis and preprocessing of the data.

[0030] Optimal scheduling relies on reinforcement learning algorithms, which mainly generate optimal charging and discharging strategies for each energy storage unit through historical load data and electricity price fluctuation prediction. During the scheduling execution process, the system also balances the workload of each energy storage unit to prevent some modules from failing prematurely due to overcharging or over-discharging. In addition, the safety management module can take protective measures in advance to prevent further spread when the module shows overheating or abnormal voltage fluctuations by using a prediction model trained with historical fault data.

[0031] The communication network module includes: The data acquisition unit uses low-power Bluetooth and ZigBee communication technologies to collect the operation data of the energy storage unit and the power grid through embedded sensors; The data transmission unit performs multi-point to multi-point data transmission based on the MQTT protocol and combines the AES-256 encryption algorithm to ensure data security; The edge computing unit locally preprocesses the collected data through an ARM Cortex-A processor, including anomaly detection and trend prediction, and optimizes the charging and discharging instructions through fuzzy control logic.

[0032] Specifically, the communication module adopts a hierarchical architecture design in actual applications to ensure the efficiency and security of data transmission. At the data acquisition end, the low-power Bluetooth and ZigBee modules are responsible for communicating with distributed sensors and uploading the operation data of the battery and the energy storage unit to the edge computing node.

[0033] The transmission part uses the MQTT protocol to achieve multi-point to multi-point data sharing through lightweight topic subscription. In addition, to enhance data security, the system embeds the AES-256 encryption algorithm to encrypt the transmitted data packets to prevent external attacks or data leakage. At the edge computing end, the ARM Cortex-A architecture processor runs fuzzy control logic to optimize the local charging and discharging instructions, and these optimized instructions are then uploaded to the central management system to achieve global regulation.

[0034] The power grid and user interface module includes: The grid connection control unit uses a control algorithm based on a synchronous phase-locked loop (PLL) to synchronize the energy storage unit with the power grid, and at the same time uses a power factor correction circuit to optimize the grid connection quality; The island operation control unit performs island operation after the power grid power failure through a fast switching circuit. The switching circuit is driven by an IGBT module, and the switching time is less than 20 milliseconds; The user power optimization unit analyzes the user's electricity load using a time series prediction model and optimizes the charging and discharging strategy of the energy storage unit through a mixed integer programming algorithm combined with real-time electricity price fluctuations.

[0035] Specifically, when the system interacts with the power grid, a synchronous phase-locked loop (PLL) technology is adopted to achieve grid connection synchronization, ensuring that the alternating current output by the energy storage unit is consistent with the grid voltage and frequency. In addition, the built-in power factor correction circuit in the system can dynamically adjust the phase angle of the output current to improve the grid connection quality. When the power grid is powered off, the fast switching circuit can complete the mode switching within less than 20 milliseconds, switching from grid-connected operation to island operation to continue powering critical loads.

[0036] The interface at the user end analyzes the user's electricity consumption habits through a prediction model, and combines real-time electricity price fluctuations to automatically optimize the charge and discharge strategies of the energy storage unit, helping users reduce electricity costs. At the same time, through an intelligent monitoring interface, users can view the operating status of the energy storage unit and the electricity price changes in real time, so as to more flexibly control electricity consumption.

[0037] The energy scheduling unit further combines the federated learning algorithm, uses multiple distributed nodes to jointly train the scheduling model, and improves the efficiency and accuracy of energy allocation in multiple scenarios.

[0038] Specifically, in the energy scheduling unit, in order to meet the energy allocation requirements in multiple scenarios (such as grid fluctuations, user load changes, energy storage unit performance differences, etc.), the system introduces a federated learning algorithm, uses multiple distributed nodes to jointly train the scheduling model, thereby improving the efficiency and accuracy of the system.

[0039] Design of the Federated Learning Architecture Deployment of Distributed Nodes Each energy storage unit is regarded as an independent distributed node, and the node includes a local status monitoring unit and a data processing unit.

[0040] The data stored and run locally by the node includes the real-time status of the energy storage unit (such as SOC, SOH, temperature, voltage) and historical operation records (such as charge and discharge cycles, grid load demands).

[0041] Each node independently performs model training, avoiding global data centralized storage, and ensuring privacy protection and data security.

[0042] Central Coordinator The system sets up a central coordinator (such as a cloud server or an edge computing cluster), which is responsible for receiving the local model parameters uploaded by each node and aggregating them to generate a global model.

[0043] The central coordinator does not directly access the original data of the nodes, but only integrates and optimizes the model parameters, which conforms to the privacy protection mechanism of federated learning.

[0044] Design and Training Process of the Federated Learning Model Local Model Training Each distributed node uses the data collected by itself to train the scheduling model locally.

[0045] The model adopts the deep reinforcement learning (DRL) framework, takes the state of the energy storage unit as input, optimizes the charge and discharge decisions, and the goal is to maximize the economic benefits and extend the life of the energy storage unit.

[0046] Local model training uses an adaptive learning rate algorithm, combines dynamic programming (DP) and long short-term memory network (LSTM) to predict the grid load demand, and improves the timeliness of the scheduling strategy.

[0047] Model parameter upload and aggregation After each node completes local training, it encrypts and uploads the updated model parameters (such as weights and bias values) to the central coordinator.

[0048] The uploaded data is protected by an encryption mechanism based on differential privacy (Differential Privacy) to ensure that the private data of each node will not be leaked.

[0049] Global model update The central coordinator uses the federated averaging (FedAvg) algorithm to perform weighted averaging on the model parameters uploaded by all nodes to generate a globally optimized model.

[0050] The updated global model is sent to each node for guiding the next round of local training and real-time scheduling.

[0051] Multi-round iterative training The federated learning process continuously optimizes the model performance through multiple rounds of iteration to ensure efficient energy scheduling in complex and changing scenarios.

[0052] Specific applications of federated learning in energy scheduling Multi-scenario scheduling adaptation The federated learning algorithm captures diverse operating scenarios (such as electricity price fluctuations, the impact of weather changes on photovoltaic systems, peak-valley differences in grid demand, etc.) by integrating data from different nodes, and optimizes the generalization ability of the scheduling model in multiple scenarios.

[0053] For example, when some nodes are in the scenario of excessive photovoltaic power generation, the global model can balance the energy distribution between these nodes and other high-load nodes, reducing energy waste.

[0054] Dynamic energy allocation Real-time monitor the states (such as SOC, temperature) and demands of each energy storage unit, combine the federated learning model to predict load changes, and generate dynamic charge and discharge strategies for each node.

[0055] During peak load, high-SOC nodes are preferentially called to release electrical energy; during low load, low-SOC nodes are allocated for charging to extend the life of the energy storage unit.

[0056] Improvement of Fault Tolerance and Robustness The distributed architecture of federated learning has high fault tolerance. Even if some nodes cannot participate in training or data is lost, the global model can still be optimized continuously through the data of other nodes.

[0057] In extreme cases (such as a node going offline due to a fault), the global scheduling model can infer a reasonable energy allocation strategy based on historical training results to ensure the stability of the system.

[0058] Performance Improvement and Evaluation Efficiency Improvement Federated learning significantly reduces the processing burden of the central system on global data through collaborative training, improving the model training speed.

[0059] Each node uses local computing power for training, avoiding the communication bottleneck caused by large-scale data upload.

[0060] Accuracy Improvement The global model integrates the scenario data of different nodes and can more accurately predict the load demand and the operating state of the energy storage unit in a complex environment.

[0061] In the test scenario, the federated learning model improves the scheduling accuracy by more than 20% compared with the traditional centralized optimization algorithm.

[0062] Privacy Protection and Security Through differential privacy and encrypted transmission mechanisms, the privacy of the operating data of each node is protected, avoiding security risks caused by data leakage.

[0063] The system adopts distributed storage and encrypted synchronization, reducing the risk of single-point failure and improving the overall security.

[0064] Scalability of Federated Learning in the Energy Scheduling Unit Cross-Regional Collaboration The federated learning architecture supports collaborative training of cross-regional nodes. For example, nodes in photovoltaic energy storage scenarios and wind power energy storage scenarios are collaborated to form a more comprehensive global model.

[0065] Multi-Energy Fusion Scheduling In the future, it can be extended to the fusion scheduling of multiple energy forms (such as thermal energy, electrical energy, hydrogen energy). Through federated learning, the energy allocation strategy is unifiedly optimized to improve the efficiency of the integrated energy system.

[0066] Embodiment 2: Refer to Figure 7, in the second embodiment of the present invention, the present invention provides a distributed power energy storage method, including the following steps: S1. Retired battery sorting and modular processing: Measure the remaining capacity, internal resistance, and health status of retired power batteries through electrochemical impedance spectroscopy technology; Based on multi-sensor data fusion technology, classify the performance parameters of retired batteries, and grade the batteries according to capacity consistency and health status; Use automated assembly equipment to assemble batteries with similar performance into standardized battery modules, and integrate a balancing protection circuit and a thermal management system; S2. Energy storage unit deployment and operation: Integrate standardized battery modules in each energy storage unit, and equip it with a battery management system BMS to monitor the voltage, current, and temperature of the module in real time through the CAN bus; Adopt a bidirectional inverter to complete the efficient conversion of direct current and alternating current, where the inverter includes a DC-DC buck-boost circuit and a DC-AC inverter; Use a liquid cooling system to control the temperature of the battery module, and monitor the thermal management effect through a thermistor; S3. Energy management and scheduling: Collect the operation data of the energy storage unit through a distributed sensor network, and perform local analysis on the data using an edge computing node; Use a reinforcement learning algorithm combined with a dynamic programming model to generate an optimized charge and discharge strategy for the energy storage unit, and balance the operation load of each unit at the same time; Based on a fault prediction algorithm, evaluate the safety status of the energy storage unit in real time, and trigger a fast circuit breaker protection mechanism and isolate the faulty unit in case of an anomaly; S4. Communication and data transmission: Collect the operation status data of the energy storage unit and the power grid through low-power Bluetooth and ZigBee communication technologies; Perform data transmission of the energy storage unit based on the MQTT protocol, and increase data security by combining the AES-256 encryption algorithm; Use fuzzy control logic in the edge computing node to optimize local charge and discharge instructions, and send the data to the cloud through wireless communication; S5. Interaction between the power grid and the user side: Adopt a grid-connected control algorithm based on a synchronous phase-locked loop (PLL) to make the energy storage unit operate synchronously with the power grid, and use a power factor correction circuit to improve the grid-connected quality; In the case of a power grid power outage, trigger the island operation mode, and switch the operation mode of the energy storage unit within 20 milliseconds through a fast switching circuit; Combined with the timing prediction model of the user's electrical load and the real-time electricity price fluctuations, the charging and discharging strategy at the user side is optimized through a mixed-integer programming algorithm.

[0067] Embodiment 3 In the third embodiment of the present invention, based on the same inventive concept, a computer-readable storage medium is proposed by the present invention. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the distributed power energy storage method in the above embodiment are implemented.

[0068] Embodiment 4 In the fourth embodiment of the present invention, based on the same inventive concept, a computer device is proposed by the present invention. The terminal includes: a processor and a memory; the processor and the memory communicate with each other; the memory is used to store instructions; the processor is used to execute the instructions in the memory and execute the distributed power energy storage method in the above embodiment.

[0069] It should be understood that each part of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiment, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following technologies well known in the art can be used: discrete logic circuits with logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits with appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0070] Finally, it should be noted that the above are only the preferred embodiments of the present invention and are not used to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A distributed power energy storage system, characterized in that: include: Retired battery sorting and modularization module, used for performance evaluation, grading and standardized modular processing of retired power batteries; Energy storage unit module, used to assemble retired batteries into energy storage units and store and release electrical energy; Energy management module, used to monitor the operating status of the energy storage unit and optimize energy distribution; Communication network module, used for data transmission and coordinated control between modules; The grid and user interface module is used for interaction between the energy storage system and the grid and user end.

2. The distributed power energy storage system according to claim 1, characterized in that: The retired battery sorting and modularization module includes: The battery performance evaluation unit uses electrochemical impedance spectroscopy to measure the internal resistance, capacity and voltage characteristics of retired batteries, and combines it with a neural network-based health status prediction model to evaluate the remaining life of the battery; The battery grading unit classifies the performance parameters of retired batteries through multi-sensor data fusion technology, performs weighted scoring on capacity, internal resistance and consistency indicators, and divides battery packs according to performance levels; The modular assembly unit assembles retired batteries with similar performance into battery modules with standardized dimensions and interfaces, while integrating thermal management and protection circuits.

3. The distributed power energy storage system according to claim 1, characterized in that: The energy storage unit module comprises: Standardized battery module units, composed of retired batteries with consistent performance, with a battery management system (BMS) embedded in the module. The BMS monitors the voltage, current and temperature of the module in real time through the CAN bus protocol, and provides a balancing function to improve module consistency; The bidirectional inverter unit adopts a two-stage conversion topology. The first stage is a DC-DC buck-boost circuit, which is used to adjust the output voltage of the battery module. The second stage is a DC-AC inverter, which performs efficient conversion between DC and AC. The temperature control and heat dissipation unit optimizes the heat dissipation structure inside the module through thermoelectric coupling simulation, and configures a liquid cooling system to control the operating temperature under high load, while using thermistors to monitor the battery temperature in real time.

4. The distributed power energy storage system according to claim 1, characterized in that: The energy management module comprises: The state monitoring unit collects the operating data of the energy storage unit based on the distributed sensor network and pre-processes the data using the edge computing node; The energy dispatch unit uses a reinforcement learning algorithm to dynamically generate charging and discharging optimization strategies based on the status data of the energy storage unit and the grid load demand, while balancing the workload of each energy storage unit through a dynamic programming model; The safety management unit uses a fault prediction algorithm combined with the historical operating data of the battery module to evaluate safety in real time. If abnormal parameters are detected, including overcharging, overheating or voltage fluctuations, the rapid circuit breaker protection mechanism is triggered and the faulty unit is isolated.

5. The distributed power energy storage system according to claim 1, characterized in that: The communication network module comprises: The data acquisition unit uses low-power Bluetooth and ZigBee communication technologies to collect operating data of the energy storage unit and the power grid through embedded sensors; The data transmission unit performs multi-point to multi-point data transmission based on the MQTT protocol and combines the AES-256 encryption algorithm to ensure data security; The edge computing unit performs local pre-processing of collected data through the ARM Cortex-A processor, including anomaly detection and trend prediction, and optimizes charging and discharging instructions through fuzzy control logic.

6. The distributed power energy storage system according to claim 1, characterized in that: The power grid and user interface module comprises: The grid-connected control unit uses a control algorithm based on a synchronous phase-locked loop (PLL) to synchronize the energy storage unit with the grid, and uses a power factor correction circuit to optimize the grid-connected quality. An island operation control unit performs island operation after a power outage in the grid by means of a fast switching circuit, wherein the switching circuit is driven by an IGBT module so that the switching time is less than 20 milliseconds; The user power optimization unit uses a time series prediction model to analyze the user's power load, and optimizes the charging and discharging strategy of the energy storage unit through a mixed integer programming algorithm in combination with real-time electricity price fluctuations.

7. The distributed power energy storage system according to claim 1, characterized in that: The energy scheduling unit further combines the federated learning algorithm and utilizes multiple distributed nodes to collaboratively train the scheduling model to improve the efficiency and accuracy of energy allocation in multiple scenarios.

8. A distributed power energy storage method, characterized in that: The distributed power energy storage system according to any one of claims 1 to 7 comprises the following steps: S1. Retired battery sorting and modular processing: Measure the remaining capacity, internal resistance and health status of retired power batteries through electrochemical impedance spectroscopy technology; Based on multi-sensor data fusion technology, the performance parameters of retired batteries are classified and the batteries are graded according to capacity consistency and health status; Use automated assembly equipment to assemble batteries with similar performance into standardized battery modules, and integrate balancing protection circuits and thermal management systems; S2. Energy storage unit deployment and operation: Standardized battery modules are integrated into each energy storage unit and equipped with a battery management system (BMS) to monitor the voltage, current and temperature of the module in real time through the CAN bus; A bidirectional inverter is used to achieve efficient conversion between DC and AC, where the inverter includes a DC-DC buck-boost circuit and a DC-AC inverter. Use liquid cooling system to control battery module temperature and monitor thermal management effect through thermistor; S3. Energy management and scheduling: Collect the operating data of the energy storage unit through a distributed sensor network and use edge computing nodes to analyze the data locally; Use reinforcement learning algorithms combined with dynamic programming models to generate charging and discharging optimization strategies for energy storage units while balancing the operating load of each unit; Based on the fault prediction algorithm, the safety status of the energy storage unit is evaluated in real time, and the fast circuit breaker protection mechanism is triggered and the faulty unit is isolated under abnormal conditions; S4. Communication and data transmission: Collect the operating status data of the energy storage unit and the power grid through low-power Bluetooth and ZigBee communication technologies; The data transmission of energy storage units is based on the MQTT protocol, and the AES-256 encryption algorithm is combined to increase data security; Use fuzzy control logic in edge computing nodes to optimize local charge and discharge instructions and send data to the cloud via wireless communication; S5. Interaction between power grid and user end: A grid-connected control algorithm based on a synchronous phase-locked loop (PLL) is used to synchronize the energy storage unit with the grid, and a power factor correction circuit is used to improve the quality of grid connection. In the event of a power outage, the island operation mode is triggered, and the operation mode of the energy storage unit is switched within 20 milliseconds through a fast switching circuit; Combining the time series prediction model of user power load and real-time electricity price fluctuations, the user-side charging and discharging strategy is optimized through a mixed integer programming algorithm.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the distributed power storage method as claimed in claim 8 is implemented.

10. A readable storage medium, characterized in that: The readable storage medium stores a computer program, and when the computer program is executed by a processor, the distributed power energy storage method according to claim 8 is implemented.

Images