Battery charging monitoring and adjusting system based on multiple areas

Abstract

The invention discloses a battery charging monitoring and adjusting system based on multiple areas, and relates to the technical field of battery management. According to the system, electrochemical impedance spectroscopy, infrared thermal imaging and electric heating parameters are synchronously collected, a physical connection and heat conduction topological relation of a battery pack is constructed in combination with a graph neural network, and monomer aging distribution and a cross-regional heat diffusion path are accurately recognized. Based on multi-region data aggregation, battery health protection, temperature equalization and a power grid demand target are fused, and charging current constraint and equalization priority parameters are dynamically generated. And through a fuzzy control algorithm, an active energy transfer mode and a passive risk isolation mode are intelligently switched, the energy distribution efficiency is optimized, fault spreading is blocked, the risk level of the system is evaluated in real time, a hierarchical regulation and control instruction is triggered, and the weight is dynamically adjusted and optimized through a feedback mechanism. The problems of extensive monitoring, strategy stiffness and safety response lag of a multi-area battery system are solved, the charging safety and economy in a complex scene are remarkably improved, and the service life of the system is remarkably prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of battery management, and in particular to a multi-region based battery charging monitoring and regulating system. Background Art

[0002] Against the backdrop of the rapid development of new energy vehicles, energy storage systems, and smart grids, battery technology has become key to supporting energy transformation and sustainable development. However, with the expansion of battery system scale and the complexity of application scenarios, the imbalance within the battery pack, battery health degradation, and thermal management issues have gradually become prominent. Especially in multi-region battery charging scenarios, due to the different aging rates of individual batteries, uneven temperature distribution, and grid load fluctuations, it is difficult for the charging process to balance safety, balance, and efficiency, resulting in system performance degradation and even safety hazards. Therefore, how to achieve intelligent monitoring and dynamic adjustment of battery charging in a multi-region environment has become a core technical issue in improving the overall performance of the battery pack, extending its service life, and ensuring safe operation.

[0003] Existing battery charging management systems mainly rely on centralized or distributed control strategies to set charging parameters through voltage, current and temperature thresholds. However, these methods have the following defects: on the one hand, traditional battery monitoring is mainly based on single modal data, lacks a comprehensive understanding of the complex physical connections and thermal conduction characteristics of the battery pack, and is difficult to effectively evaluate the degree of cell aging and inter-regional thermal coupling relationships; on the other hand, existing optimization strategies usually only adjust the charge and discharge of local units, and fail to fully integrate the three optimization goals of battery health attenuation, temperature balance and grid demand, resulting in poor overall coordination of the system. In addition, the balancing control mode is single and fails to dynamically adapt to different battery states. In terms of safety management, there is a lack of real-time cross-domain risk judgment and regulation mechanisms, making it difficult to effectively respond to sudden failures of abnormal cells. Therefore, it is difficult for existing methods to achieve accurate charging monitoring and efficient dynamic adjustment in a multi-region environment. Summary of the invention

[0004] In view of the deficiencies of the prior art, the present invention provides a multi-region based battery charging monitoring and regulation system, which solves the problems of the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a multi-regional battery charging monitoring and regulation system, comprising the following modules: a multi-modal perception module, a federal collaborative optimization module, a dynamic hybrid balancing strategy module, and a cross-domain safety decision-making module; the multi-modal perception module is used to synchronously collect electrochemical impedance spectroscopy, infrared thermal imaging, and electrothermal parameters in the battery partition, and construct the physical connection and thermal conduction topological relationship of the battery pack through a graph neural network, and output a single cell aging degree map and cross-regional thermal coupling correlation characteristics; the federal collaborative optimization module is used to perform multi-region data aggregation according to the single cell aging degree map and cross-regional thermal coupling correlation characteristics, and integrate battery health attenuation suppression, temperature gradient balancing, and power grid implementation. The dynamic hybrid balancing strategy module is used to dynamically switch between active energy transfer and passive risk suppression modes through a fuzzy control algorithm according to the inter-regional balancing mode priority parameters, realize cross-regional energy redistribution through an inductor-capacitor network in the active mode, and perform rapid isolation of abnormal cells based on charge state prediction in the passive mode; the cross-domain safety decision module is used to comprehensively determine the battery cluster-level risk level according to the safety constraint parameters of the dynamic charging current, trigger local load reduction and global strategy reset instructions, and feedback the risk penalty factor to the federal collaborative optimization module to dynamically adjust the weight distribution of health attenuation and temperature balance.

[0006] Furthermore, the specific process of constructing the physical connection and thermal conduction topological relationship of the battery pack through the graph neural network is as follows: each single cell in the battery pack is defined as a node of the graph neural network, and the node attributes include real-time voltage and internal resistance parameters; the physical connection relationship and thermal conduction path of adjacent single cells are defined as edges, and the edge weights are dynamically calculated through the voltage fluctuation correlation and the temperature gradient changes of infrared thermal imaging data to generate a topological network of the battery pack coupling characteristics.

[0007] Furthermore, the specific process of outputting the monomer aging degree map and cross-regional thermal coupling correlation characteristics is as follows: the aging coefficient of the monomer battery is inverted according to the electrochemical impedance spectroscopy data, and the voltage-internal resistance correlation characteristics extracted from the graph neural network topological network are combined to generate an aging degree distribution heat map with regions as units; the cross-regional thermal coupling correlation matrix is ​​generated by calculating the thermal diffusion intensity of infrared thermal imaging data and the edge weights of the topological network to identify high-risk thermal diffusion paths.

[0008] Furthermore, according to the monomer aging degree map and cross-regional thermal coupling correlation characteristics, the specific process of multi-regional data aggregation is as follows: cross-regional hierarchical aggregation of the aging degree distribution heat map and the cross-regional thermal coupling correlation matrix: Regional layer: align the aging degree distribution characteristics of each region, and weightedly generate a global aging trend map; Cross-domain layer: extract high-risk paths in the thermal coupling correlation matrix, and construct a cross-regional thermal interference chain relationship.

[0009] Furthermore, the three optimization objectives of integrating battery health decay suppression, temperature gradient balancing and real-time grid demand are as follows: mapping the global aging trend map into a health decay suppression item, mapping the thermal interference chain relationship into a temperature gradient balancing item, and mapping the real-time grid demand signal into a charging and discharging cost item. The three optimization objectives are merged into a unified reward function for federal collaborative optimization through a dynamic weight allocation function.

[0010] Furthermore, the specific process of generating the safety constraint parameters of the dynamic charging current and the priority parameters of the inter-regional balancing mode is as follows: based on the unified reward function, through interactive game training of distributed edge nodes and cloud models, the maximum allowable charging current value of each region is output as the safety constraint parameter, and combined with the thermal interference chain relationship and the global aging trend map, the priority score of the inter-regional balancing operation is calculated as the priority parameter.

[0011] Furthermore, according to the inter-regional equilibrium mode priority parameters, the specific process of dynamically switching between active energy transfer and passive risk suppression modes through the fuzzy control algorithm is as follows: the inter-regional equilibrium mode priority parameters, thermal coupling association matrix and global aging trend map are used as fuzzy control inputs, the membership functions of active mode and passive mode are defined, and the mode switching threshold is calculated through fuzzy rule base matching to trigger energy transfer path selection or abnormal monomer isolation instructions.

[0012] Furthermore, based on the safety constraint parameters of the dynamic charging current, the battery cluster-level risk level is comprehensively determined, and the specific process of triggering local load reduction and global strategy reset instructions is as follows: Based on the current over-limit ratio, the number of temperature gradient over-limit areas, and the number of abnormal aging trend areas in the safety constraint parameters, a multi-level risk assessment model is constructed: Level 1 risk: triggering local load reduction to limit the charging current in the abnormal area; Level 2 risk: resetting the global charging strategy and switching to the backup balancing path.

[0013] Furthermore, the risk penalty factor is fed back to the federated collaborative optimization module, and the specific process of dynamically adjusting the weight distribution of health decay and temperature balance is as follows: according to the battery cluster-level risk level assessment results, a risk penalty factor that is nonlinearly positively correlated with the risk level is generated; the risk penalty factor is input into the weight distribution function of the federated collaborative optimization module to dynamically increase the weight coefficient of the health decay inhibition item or the temperature gradient balance item; through the gradient feedback mechanism of the federated collaborative optimization module, the weight ratio of health decay and temperature balance in the cloud global model is updated, and the updated model parameters are sent to the edge nodes to achieve adaptive optimization of multi-region collaborative strategies.

[0014] The present invention has the following beneficial effects: (1) A multi-regional battery charging monitoring and regulation system that achieves accurate perception and optimized regulation of the battery pack status through the synergy of a multimodal sensing module and a federated collaborative optimization module. The multimodal sensing module can simultaneously collect electrochemical impedance spectroscopy, infrared thermal imaging, and electrothermal parameters, and use graph neural networks to construct the physical connection and thermal conduction topological relationship of the battery pack, thereby improving the recognition accuracy of the battery aging status and thermal coupling relationship. The federated collaborative optimization module aggregates multi-region data based on the sensing data, and integrates the three optimization goals of battery health attenuation suppression, temperature gradient balancing, and real-time grid demand to generate safety constraint parameters for dynamic charging current and priority parameters for inter-regional balancing modes, making the charging strategy safer and more globally optimized.

[0015] (2) A multi-regional battery charging monitoring and regulation system improves the flexibility and safety of energy scheduling during charging through a dynamic hybrid balancing strategy module and a cross-domain safety decision module. The dynamic hybrid balancing strategy module can intelligently switch between active energy transfer and passive risk suppression modes based on the fuzzy control algorithm according to the priority parameters of the inter-regional balancing mode, making the system more adaptable in cross-regional energy redistribution and abnormal single-cell isolation. The cross-domain safety decision module comprehensively evaluates the battery cluster-level risk level based on the safety constraint parameters of the dynamic charging current, triggers local load reduction and global strategy reset, and feeds back the risk penalty factor to the federal collaborative optimization module to dynamically adjust the weight distribution of health decay and temperature balance, thereby improving the safety and long-term stability of the charging system.

[0016] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a flow chart of a multi-zone based battery charging monitoring and regulation system of the present invention. DETAILED DESCRIPTION

[0018] The embodiment of the present application solves the problems of uneven health decay, excessive temperature gradient and difficult safety guarantee of battery packs during multi-region charging through a multi-region battery charging monitoring and regulation system. The system uses multimodal perception, federated collaborative optimization, dynamic hybrid balancing strategy and cross-domain security decision-making and other technical means to achieve accurate perception of battery status, dynamic energy balancing and active control of safety risks, thereby improving the overall charging efficiency of the battery pack, extending its service life and enhancing the safety of the charging process.

[0019] The overall idea of ​​the solution in the embodiments of this application is as follows: Electrochemical impedance spectroscopy, infrared thermal imaging and electrothermal parameters are collected synchronously within the battery partition, and the physical connection and thermal conduction topological relationship of the battery pack is constructed through a graph neural network to output a single cell aging degree map and cross-regional thermal coupling correlation characteristics.

[0020] According to the monomer aging degree map and cross-regional thermal coupling correlation characteristics, multi-regional data aggregation is carried out, and the three optimization goals of battery health attenuation suppression, temperature gradient balance and real-time grid demand are integrated to generate safety constraint parameters of dynamic charging current and priority parameters of inter-regional balancing mode.

[0021] According to the priority parameters of the inter-regional balancing mode, the fuzzy control algorithm is used to dynamically switch between the active energy transfer and passive risk suppression modes. In the active mode, the inductor and capacitor network is used to realize cross-regional energy redistribution. In the passive mode, the abnormal monomers are quickly isolated based on the charge state prediction.

[0022] According to the safety constraint parameters of the dynamic charging current, the battery cluster-level risk level is comprehensively determined, triggering local load reduction and global strategy reset instructions, and the risk penalty factor is fed back to the federal collaborative optimization module to dynamically adjust the weight distribution of health attenuation and temperature balance.

[0023] See also Figure 1 The embodiment of the present invention provides a technical solution: a battery charging monitoring and regulation system based on multiple regions, comprising the following modules: a multi-modal perception module, a federated collaborative optimization module, a dynamic hybrid balancing strategy module, and a cross-domain safety decision module; the multi-modal perception module is used to synchronously collect electrochemical impedance spectroscopy, infrared thermal imaging, and electrothermal parameters in the battery partition, and construct the physical connection and thermal conduction topological relationship of the battery pack through a graph neural network, and output a single cell aging degree map and cross-regional thermal coupling correlation characteristics; the federated collaborative optimization module is used to perform multi-region data aggregation based on the single cell aging degree map and cross-regional thermal coupling correlation characteristics, and integrate the three aspects of battery health attenuation suppression, temperature gradient balancing, and real-time grid demand. The optimization objectives are set to generate safety constraint parameters of dynamic charging current and priority parameters of inter-regional balancing mode; the dynamic hybrid balancing strategy module is used to dynamically switch between active energy transfer and passive risk suppression modes through a fuzzy control algorithm according to the priority parameters of the inter-regional balancing mode, to realize cross-regional energy redistribution through an inductor-capacitor network in the active mode, and to perform rapid isolation of abnormal cells based on charge state prediction in the passive mode; the cross-domain safety decision module is used to comprehensively determine the risk level of the battery cluster level according to the safety constraint parameters of the dynamic charging current, to trigger local load reduction and global strategy reset instructions, and to feed back the risk penalty factor to the federal collaborative optimization module, so as to dynamically adjust the weight distribution of health attenuation and temperature balance.

[0024] In this implementation scheme, the multimodal sensing module is responsible for synchronously collecting multiple key parameters within the battery partition to achieve accurate battery health status analysis. The physical connection and thermal conduction topological relationship of the battery pack is constructed through the graph neural network to obtain the monomer aging degree map and cross-region thermal coupling correlation characteristics. Electrochemical impedance spectroscopy: used to analyze the electrochemical characteristics inside the battery and determine its health status. Infrared thermal imaging: detect the temperature distribution on the battery surface, analyze the heating situation, and determine whether there is an overheating risk. Electrothermal parameters: including battery voltage, current, temperature and other parameters, used to comprehensively evaluate the working status of the battery. Graph neural network: a deep learning method suitable for graph structured data, used to analyze the current flow and thermal conduction relationship between battery cells. Fusion analysis through multiple sensing methods makes aging prediction more accurate. Analyze the thermal coupling relationship between batteries to prevent local overheating from causing system failure. Through GNN modeling, dynamically update the battery status information and optimize the subsequent charging strategy. The federated collaborative optimization module uses multi-region data aggregation to optimize between the three goals of battery health attenuation suppression, temperature gradient balance, and real-time grid demand, and generate safety constraint parameters for charging current and priority parameters for inter-regional balancing mode. Battery health decay suppression: refers to optimizing the charging method, slowing down the battery capacity decay, and extending the service life. Temperature gradient balancing: refers to adjusting the charging strategy between different battery areas to prevent local temperatures from being too high or too low and improve the uniformity of the battery pack. Real-time grid demand: refers to dynamically adjusting the charging power in combination with the load conditions of the grid to reduce the impact on the grid. Safety constraint parameters of charging current: refers to the safe charging range calculated based on the battery health status to prevent the battery from being overcharged or overloaded. Inter-regional balancing mode priority parameters: refers to adjusting the priority of power flow between different regions to ensure balanced charging. Prevent excessive battery decay: control the charging rate through intelligent optimization to slow down battery aging. Balanced charging temperature: avoid high temperature concentration in some battery areas and improve system stability. Adapt to grid demand: dynamically adjust the charging strategy, reduce peak load pressure, and improve energy utilization. The dynamic hybrid balancing strategy module dynamically switches between active energy transfer mode and passive risk suppression mode according to the inter-regional balancing mode priority parameters to achieve the optimal charging strategy. Active mode: use the inductor and capacitor network to redistribute energy across regions to improve charging balance. Passive mode: Based on the state of charge prediction, the abnormal battery cells are quickly isolated to prevent them from affecting the overall system. Fuzzy control algorithm: A control method based on fuzzy logic that can achieve smooth switching in complex environments and improve control accuracy. Inductor capacitor network: A circuit structure used for energy storage and energy transfer that can achieve power regulation between different batteries. State of charge prediction: Through historical data and real-time measurements, the remaining power of the battery is predicted and used to optimize the charging strategy. Improve battery balance: In active mode, power is distributed across regions to reduce uneven charging. Quick response to abnormal situations: Isolate faulty batteries through passive mode to prevent them from affecting the entire system.Intelligent adaptation to different situations: Utilize fuzzy control algorithm to flexibly switch charging modes and improve system intelligence. The cross-domain safety decision module evaluates the risk level of battery clusters in real time based on the safety constraint parameters of dynamic charging current, and triggers local load reduction and global policy reset when necessary. Feedback the risk penalty factor to the federated collaborative optimization module to adjust the optimization weights of health management and temperature balance strategies. Battery cluster-level risk level: Determine the safety level of the current system by evaluating the status of the entire battery pack. Local load reduction: When an abnormal situation is detected, reduce the charging power of some batteries to prevent the spread of faults. Global policy reset: When the risk is high, recalculate the charging strategy to ensure the stability of the overall system. Risk penalty factor: Used to quantify the impact of risk on the charging optimization strategy and dynamically adjust the optimization weight. Improve charging safety: Prevent battery overheating or overload damage through real-time risk assessment. Local adjustment to reduce risk: Avoid system-level failures through local load reduction and improve overall stability. Intelligent feedback mechanism: Dynamically adjust the optimization strategy through risk penalty factors to improve the system's adaptability.

[0025] Specifically, the specific process of constructing the physical connection and thermal conduction topological relationship of the battery pack through the graph neural network is as follows: each single cell in the battery pack is defined as a node of the graph neural network, and the node attributes include real-time voltage and internal resistance parameters; the physical connection relationship and thermal conduction path of adjacent single cells are defined as edges, and the edge weights are dynamically calculated through the voltage fluctuation correlation and the temperature gradient changes of infrared thermal imaging data to generate a topological network of the battery pack coupling characteristics.

[0026] In this implementation scheme, the graph neural network node definition regards each single cell as a node, and its attributes include: Real-time voltage: the instantaneous voltage value of the single cell, reflecting the charge and discharge status. Internal resistance: equivalent AC impedance, measuring the electrochemical impedance characteristics. Edge construction and weight calculation: Physical connection relationship: If the current sampling circuits of two single cells are directly connected, an edge is connected on the graph. Heat conduction path: If there is a main heat exchange path between two single cells (such as shell contact or coolant flow), a heat conduction edge is established. Edge weight calculation: Voltage fluctuation correlation :Measure the coupling degree of monomer voltage during charging and discharging based on Pearson correlation coefficient: ;in: , is the voltage value at time t , is the average voltage in the time window T, temperature gradient correlation :Calculate heat conduction relationship based on infrared thermal imaging data: ;in: , The formula reflects the relative difference of temperature gradient change. A large value indicates weak heat conduction, and a small value indicates close coupling. Topological network generation, learning the relationship between nodes through graph neural network (GNN), defining the adjacency matrix : ; The adjacency matrix is ​​used for subsequent battery pack state estimation and balancing optimization.

[0027] Specifically, the specific process of outputting the monomer aging degree map and cross-regional thermal coupling correlation characteristics is as follows: the aging coefficient of the monomer battery is inverted according to the electrochemical impedance spectroscopy data, and the voltage-internal resistance correlation characteristics extracted from the graph neural network topological network are combined to generate an aging degree distribution heat map with regions as units; the cross-regional thermal coupling correlation matrix is ​​generated by calculating the thermal diffusion intensity of infrared thermal imaging data and the edge weights of the topological network to identify high-risk thermal diffusion paths.

[0028] In this embodiment, the electrochemical impedance spectrum can reflect the aging process of the battery cell, and the health of the battery is evaluated mainly by analyzing the impedance characteristics of the battery. The electrochemical impedance spectrum data can invert the aging coefficient of each battery, thereby providing a basis for the battery status assessment. Generally, the impedance spectrum of the battery contains two main parts: charge transfer impedance and solid electrolyte interface impedance. The calculation of the aging coefficient is based on the electrochemical impedance data of the battery, through the following steps: Obtain impedance spectrum data: by measuring the impedance value of the battery at different frequencies. Extract related parameters: extract charge transfer impedance and solid electrolyte interface impedance from the impedance data. Invert aging coefficient: invert the aging coefficient of each battery cell through frequency response analysis and model fitting of the impedance. The larger the aging coefficient, the worse the health of the battery. Graph neural network is used to extract the correlation features between each single cell in the battery pack. Assume that the battery pack consists of N single cells, where each single cell is regarded as a node in the graph. The attributes of each node include: : Real-time voltage of battery cell k. : The internal resistance of battery cell k. The connection relationship (i.e., adjacency relationship) between batteries is represented by the physical connection and the correlation between the fluctuation of voltage and internal resistance. The edge between nodes ( ) represents the relationship between battery cells m and n, and its edge weight Dynamic calculation based on voltage fluctuation and temperature gradient changes is defined as: ;in, and are the voltages of batteries m and n, and is their internal resistance, and is their temperature, and the function g describes the coupling relationship between these factors. The graph neural network updates the features of each node through a propagation mechanism based on this information. After multiple iterations, the feature vector of the node (battery health status information) is updated as: ;in, is the feature vector of node k after the tth iteration, N(k) is the set of adjacent nodes of node k, and σ is the activation function. Finally, the graph neural network aggregates the features of all nodes into a global feature to represent the health status and aging degree of the battery pack. Generation of cross-region thermal coupling association matrix, the temperature distribution of the battery pack is an important factor affecting battery health. Through infrared thermal imaging technology, the surface temperature data of battery cells can be obtained. In order to evaluate the thermal coupling relationship between batteries, we need to calculate the thermal diffusion intensity between each pair of batteries. Thermal diffusion intensity describes the heat transfer efficiency between battery m and battery n and is defined as: ;in: and are the surface temperatures of battery m and battery n respectively. is the thermal conductivity coefficient between battery m and battery n, indicating the heat transfer efficiency. By calculating the thermal diffusion intensity between every two battery cells, a thermal coupling matrix H can be generated, which reflects the thermal coupling between different areas in the battery pack.

[0029] ; where each element of the matrix Represents the heat diffusion intensity between battery m and battery n. Identification of high-risk heat diffusion paths Based on the heat diffusion intensity matrix H, we can define a heat diffusion risk threshold τ if >τ, it is considered that there is a high-risk heat diffusion path between battery m and battery n. A high-risk heat diffusion path indicates that the temperature difference between these batteries is large, which may cause temperature imbalance or local overheating. By calculating the heat diffusion intensity between all batteries and comparing it with the threshold, the system can automatically identify high-risk heat diffusion paths, thereby providing a reference for optimizing the charging strategy.

[0030] Specifically, according to the monomer aging degree map and cross-regional thermal coupling correlation characteristics, the specific process of multi-regional data aggregation is as follows: cross-regional hierarchical aggregation of the aging degree distribution heat map and the cross-regional thermal coupling correlation matrix: Regional layer: align the aging degree distribution characteristics of each region, and weightedly generate a global aging trend map; Cross-domain layer: extract high-risk paths in the thermal coupling correlation matrix, and construct a cross-regional thermal interference chain relationship.

[0031] In this embodiment, in the management of the battery pack, the aging degree of the battery and the thermal coupling relationship are key factors in evaluating the health status of the battery and optimizing the operation strategy. By performing multi-region data aggregation on the aging degree map and thermal coupling correlation characteristics of the single battery, the overall health trend and potential thermal interference risk areas in the battery pack can be more effectively identified, thereby providing data support for battery management and optimization strategies. Regional layer: Generate a global aging trend map. The task of the regional layer is to generate a global aging trend map by weighted aggregation based on the aging degree distribution heat map of each region. The aging degree of each region has been calculated by inversion of electrochemical impedance spectroscopy data and voltage-internal resistance correlation characteristics of the graph neural network. In order to generate a global aging trend map, it is first necessary to align and weight the aging degree of each region. The specific steps are as follows: Align the aging degree characteristics of each region: Align the aging degree data of each region in the battery pack to ensure the scale consistency between different regions. In some cases, normalization may be required to ensure that the aging degree data of different regions can be compared under the same standard. Weighted generation of global aging trend map: Different weights are given to the aging degree data of each region according to the importance of each region or the number of battery cells. Cross-domain layer: Construct cross-regional thermal interference chain relationships: At the cross-domain layer, the task is to extract high-risk paths based on the thermal coupling association matrix and construct cross-regional thermal interference chain relationships. Each element in the thermal coupling association matrix reflects the thermal coupling strength between battery m and battery n. Through this matrix, high-risk thermal diffusion paths within the battery pack can be identified, that is, battery pairs with relatively strong heat transfer, which may cause local temperature increases and affect battery health. Construct cross-regional thermal interference chain relationships: Through the extracted high-risk paths, a cross-regional thermal interference chain relationship can be constructed. This chain relationship reflects the heat transfer path within the battery pack and how heat diffusion propagates from one area to another. If there are continuous thermal coupling relationships between multiple batteries, they will form a thermal interference chain.

[0032] Specifically, the three optimization goals of integrating battery health decay suppression, temperature gradient balancing and real-time grid demand are as follows: mapping the global aging trend map into a health decay suppression item, mapping the thermal interference chain relationship into a temperature gradient balancing item, and mapping the real-time grid demand signal into a charging and discharging cost item. The three optimization goals are merged into a unified reward function for federal collaborative optimization through a dynamic weight allocation function.

[0033] In this implementation scheme, the three optimization objectives are defined as follows: When optimizing the battery charge and discharge management strategy, we need to convert the three key factors (battery health decay, temperature gradient balance, and real-time grid demand) into optimization objectives and fuse them through appropriate weights. Specifically, we will map these factors in the following ways: Battery health decay inhibition term: The degree of battery aging directly affects its health decay, so it is necessary to reduce the rate of battery decay through a health decay inhibition term. Based on the battery aging information in the global aging trend map, a decay inhibition term can be constructed with the goal of slowing down the battery aging process and increasing the battery life. This inhibition term can be expressed as a function , where adjustments are made based on the global aging trend map (reflecting the health decay of the battery). Temperature gradient balancing term: Uneven battery temperature may lead to the formation of thermal interference chains, thereby increasing the risk of battery failure. In order to balance the temperature gradient of the battery pack, the thermal interference chain relationship (extracted through the cross-region thermal coupling matrix) can be used as a temperature gradient balancing term. The goal of this optimization is to reduce the heat load in the high-temperature area and optimize the temperature distribution of the battery pack. Charging and discharging cost term: The efficiency of the charging and discharging process and the real-time demand of the power grid are another key factor. In the dynamic charging and discharging process, the real-time demand signal of the power grid will affect the charging and discharging decisions of the battery. In order to optimize the charging and discharging process, the real-time demand signal of the power grid must be considered to avoid overcharging or discharging during the peak load period of the power grid, which helps to reduce electricity costs. This optimization goal maps the power grid demand to the charging and discharging cost term. Fusion of objective functions: dynamic weight allocation. In order to integrate these three optimization objectives, a dynamic weight allocation function needs to be designed. The function of this function is to assign different weights to each objective according to the current state, thereby balancing the conflicts and synergies between the three objectives. Function of dynamic weight allocation It can be adjusted in real time according to the health of the battery, the current temperature gradient and the grid load. The design of the weight distribution function needs to consider the following factors: Battery health decay suppression: When the battery ages faster, the weight of health decay suppression should be increased to reduce battery loss. Temperature gradient balance: When the temperature difference is large, the weight of temperature balance should be increased to avoid the impact of thermal interference chain on the battery pack. Real-time grid demand: When the grid demand is high, the battery charging speed should be appropriately reduced to avoid excessive grid load. The weight distribution function w(t) can be expressed as: ; ; ;in: is the weight of health decay suppression, It is the global aging trend, indicating the aging speed of the battery. is the weight of temperature equilibrium, is the temperature difference between regions. is the weight of the charging and discharging cost, It is the real-time load demand of the power grid. , , are the coefficients that control the weight distribution of the three objectives respectively. Adjusting these coefficients can balance the influence between the various optimization objectives. Unified reward function: Ultimately, the three optimization objectives need to be combined through a unified reward function R(t). This reward function will dynamically guide the battery charging strategy by combining the weight of each objective and its corresponding optimization objective. The form of the unified reward function is as follows: ;in: It is the objective function of battery health decay suppression, which represents the efficiency of slowing down battery aging. It is the objective function of temperature balance, which indicates the degree of reduction of temperature difference. It is the objective function of the charging and discharging cost, which indicates the adaptability of the grid load during the charging process.

[0034] Specifically, the specific process of generating the safety constraint parameters of the dynamic charging current and the priority parameters of the inter-regional balancing mode is as follows: based on the unified reward function, through interactive game training of distributed edge nodes and cloud models, the maximum allowable charging current value of each region is output as the safety constraint parameter, and combined with the thermal interference chain relationship and the global aging trend map, the priority score of the inter-regional balancing operation is calculated as the priority parameter.

[0035] In this implementation scheme, in order to ensure the safety, balance and efficiency of the battery pack during charging, it is necessary to generate the safety constraint parameters of the dynamic charging current and the priority parameters of the inter-regional balancing operation according to the health status of the battery, the chain relationship of thermal interference and the demand of the power grid. This process involves the use of a unified reward function, interactive game training of distributed edge nodes and cloud models, and the calculation of safety constraints and balancing priorities. Based on the training process of the unified reward function and the role of the unified reward function, we use the previously mentioned unified reward function R(t) to guide the entire charging management strategy. This function combines the three optimization objectives of battery health decay suppression, temperature balance and charging and discharging cost. Through the interactive game training of the edge computing node and the cloud model, the battery charging strategy is optimized, and the charging strategy is adjusted in real time according to the health status of the battery, the temperature distribution of the region and the power grid load during the charging process. Interactive game training of distributed edge nodes and cloud models, edge computing nodes: The edge nodes in each region collect local data (such as battery voltage, temperature, internal resistance, etc.) and adjust the charging current value according to the model. The edge node calculates the maximum allowable value of the charging current based on the local battery and grid status to meet the safety requirements. These values ​​are transmitted to the cloud through the model. Cloud model: The cloud model aggregates the calculation results from each region, optimizes the model according to global demand, and readjusts the strategy of the edge node. The cloud can not only adjust the charging current according to the global grid demand, but also perform global aging trend analysis and adjust the charging strategy between regions. This interactive game training process outputs a stable charging current distribution plan through continuous optimization and feedback. Through the training process, the edge node can perform safety constraint calculations on the maximum charging current value of each region according to the global optimization guidance given by the cloud. Calculation of safety constraint parameters: Maximum allowable charging current value In order to ensure the safety of the charging process, the maximum allowable charging current value Certain safety constraints are required. These constraints take into account the battery's aging state, thermal interference, and the overall temperature distribution of the battery pack. Safety constraint current calculation formula: ;in: It is the maximum charging current calculated by the battery health attenuation suppression requirement. As the battery ages, the maximum charging current value will decrease. It is the maximum charging current under the condition of temperature balance limit. When the temperature of the battery pack area is too high, the charging current needs to be reduced to prevent runaway. It is the maximum charging current adjusted according to the real-time load demand of the power grid to avoid overcharging during peak load periods of the power grid. In addition to the safety constraints of the charging current, the calculation of the priority parameters of the inter-regional balancing mode also requires the calculation of the priority of the inter-regional balancing operation based on the status of each area of ​​the battery pack. By combining the thermal interference chain relationship and the global aging trend map, the charging priority of different areas is determined to ensure that the heat load is reasonably distributed in the battery pack. Thermal interference chain relationship: Based on the cross-regional thermal coupling matrix, the thermal interference chain relationship between different areas in the battery pack can be extracted. These relationships indicate the heat conduction path and intensity between battery areas. Areas with strong thermal interference should be given priority to control the charging current to avoid overheating. Global aging trend map: Combined with the global aging trend map, the aging degree of the battery in each area can be calculated. Areas with severe aging need to be protected first and their charging current reduced. The higher the aging degree of the area, the lower its priority, because these areas are more sensitive to excessive current input. Priority score, based on the thermal interference chain and the global aging trend, a priority score can be calculated for each area , which is calculated based on the following factors: ;in: It is the thermal interference chain path strength of region a, indicating the strength of thermal coupling between this region and other regions. The higher the path strength, the higher the region needs to be treated first. is the aging coefficient of region a. The higher the aging degree, the lower the priority score. It is the weight coefficient for adjusting thermal interference and aging effects.

[0036] Specifically, according to the inter-regional equilibrium mode priority parameters, the specific process of dynamically switching between active energy transfer and passive risk suppression modes through the fuzzy control algorithm is as follows: the inter-regional equilibrium mode priority parameters, thermal coupling correlation matrix and global aging trend map are used as fuzzy control inputs, the membership functions of active mode and passive mode are defined, and the mode switching threshold is calculated through fuzzy rule base matching to trigger energy transfer path selection or abnormal monomer isolation instructions.

[0037] In this implementation, a fuzzy control algorithm is introduced for dynamic mode switching: during the battery pack charging management process, the system needs to switch between the active energy transfer mode and the passive risk suppression mode according to the health status, thermal coupling and aging degree of different regions. Through the fuzzy control algorithm, dynamic switching between these two modes can be performed to ensure that the battery pack charging process can be charged efficiently while avoiding potential risks. Fuzzy control input parameters: Inter-regional balancing mode priority parameter: Inter-regional balancing mode priority parameter Inter-regional balancing mode priority parameter It represents the charging priority of each area. The area with higher priority needs to be allocated the charging current first. This parameter reflects factors such as heat load, aging degree and thermal coupling relationship between regions. Thermal coupling correlation matrix The thermal coupling correlation matrix describes the thermal conduction relationship between different areas of the battery pack, indicating the intensity of heat transfer from one area to another. If the heat load of a certain area is high and the thermal coupling with other areas is strong, the temperature of the area may be greatly affected, which in turn affects the charging decision. Global aging trend map The global aging trend map reflects the aging degree of each area in the battery pack. Areas with severe aging are more vulnerable to damage from excessive charging current, so the charging current needs to be limited and these areas need to be protected first. Membership function definition: In order to implement fuzzy control, we need to define membership functions for active mode and passive mode. The membership function will determine whether the current state belongs to active mode or passive mode based on the different values ​​of the input parameters. Active mode refers to a mode that maximizes energy transfer and charging efficiency while ensuring safety. When the priority of inter-area balancing is high and the thermal coupling is weak, the system tends to enable active mode. Membership function of active mode It can be defined as: ;in: is a coefficient that adjusts the priority and thermal coupling weight. It is the inter-region balancing mode priority, indicating the charging priority of the region. is the maximum value of the thermal coupling strength, indicating the degree of thermal coupling between regions. This membership function indicates that when the priority is high and the thermal coupling is weak, the membership of the active mode is high, and the system tends to choose the active mode. Passive mode membership function, passive mode means that in order to avoid risks, the system chooses to isolate high-temperature or severely aged single cells to suppress heat diffusion. When the regional aging is serious and the thermal coupling is strong, the system tends to enter the passive mode. The membership function of the passive mode is defined as: ;in: is the coefficient for adjusting the aging degree and thermal coupling weight. It is the aging degree of the battery area, reflecting the aging status of the battery. is the maximum value of thermal coupling intensity. When the area is severely aged and the thermal coupling is strong, the membership of the passive mode is high, and the system will choose to enter the passive mode for safety protection. Fuzzy rule base matching: According to the membership function, a fuzzy rule base can be designed to describe the switching conditions between the active mode and the passive mode. The rules in the fuzzy rule base will determine whether to trigger the mode switch based on the current membership value. Fuzzy rules: Rule 1: If the inter-regional balance priority is high and the thermal coupling intensity is low, the active mode is enabled. Rule 2: If the area is aged and the thermal coupling intensity is high, the passive mode is enabled. Rule 3: If the area is aged and the thermal coupling intensity is high, the active mode is enabled, but the energy transfer rate is limited. Mode switching threshold calculation, the fuzzy control algorithm calculates the threshold of mode switching by calculating the membership function of the input parameters, matching the rules in the fuzzy rule base. If the current conditions meet the triggering conditions of a certain rule, the system will trigger the mode switch. Finally, based on the output of the fuzzy control algorithm, the system decides whether to enter the active energy transfer mode or the passive risk suppression mode. According to the mode switching threshold, the system will: Energy transfer path selection: In active mode, the optimal charging path is selected according to the priority and thermal coupling relationship between regions to maximize the energy transfer efficiency. Abnormal single cell isolation instruction: In passive mode, by isolating severely aged or overheated single cells, the risk of heat diffusion is suppressed to ensure the safety of the battery pack.

[0038] Specifically, based on the safety constraint parameters of the dynamic charging current, the battery cluster-level risk level is comprehensively determined, and the specific process of triggering local load reduction and global strategy reset instructions is as follows: Based on the current over-limit ratio, the number of temperature gradient over-limit areas, and the number of abnormal aging trend areas in the safety constraint parameters, a multi-level risk assessment model is constructed: Level 1 risk: triggering local load reduction to limit the charging current in the abnormal area; Level 2 risk: resetting the global charging strategy and switching to the backup balancing path.

[0039] In this implementation scheme, the process of comprehensively determining the risk level of the battery cluster level according to the safety constraint parameters of the dynamic charging current mainly includes the following steps: Evaluation of safety constraint parameters: Current overlimit ratio: This refers to the ratio of the charging current in the battery cluster that exceeds the set maximum value. If the charging current of some single cells is too large, it may cause overcharging risk. Number of temperature gradient overlimit areas: refers to the number of areas in the battery cluster where the temperature exceeds the set safety range. If the temperature of multiple areas is too high, it may cause thermal runaway or battery damage. Number of abnormal aging trend areas: Indicates the number of areas where the aging rate is too fast. If the battery aging degree in some areas is too high, it may affect the overall health of the battery. Multi-level risk assessment model: Based on the above safety constraint parameters, the system will build a multi-level risk assessment model to determine the current risk level of the battery cluster. The risk level is divided into two levels: Level 1 risk: When the current overlimit ratio, the number of temperature overlimit areas, or the number of abnormal aging areas exceeds the set threshold, the battery cluster faces a higher risk. At this time, the system will trigger a local load reduction strategy to limit the charging current in high-risk areas to reduce the risk of overcharging or overheating. Level 2 risk: If the risk further increases, the system will initiate more stringent measures, reset the global charging strategy and switch to the backup balancing path. The purpose of this measure is to distribute the charging current more evenly to avoid problems in some areas due to overcharging or overheating. Triggering protection measures: Local load reduction: When the level 1 risk is triggered, the system will limit the charging current, especially for those areas with excessive temperature or severe aging, to reduce their charging load. Global strategy reset: When the level 2 risk is triggered, the system will reset the charging strategy of the entire battery cluster and use the backup balancing path to redistribute the current to ensure that all areas within the battery cluster are in a safe charging state. By real-time monitoring of the current, temperature and aging within the battery cluster, the current risk level is evaluated, and corresponding protection measures are taken according to different risk levels to ensure that the battery pack always remains within a safe range during the charging process to avoid potential risks such as overcharging, overheating or rapid aging.

[0040] Specifically, the risk penalty factor is fed back to the federal collaborative optimization module, and the specific process of dynamically adjusting the weight distribution of health decay and temperature balance is as follows: According to the battery cluster-level risk level assessment results, a risk penalty factor that is nonlinearly positively correlated with the risk level is generated; the risk penalty factor is input into the weight distribution function of the federal collaborative optimization module to dynamically increase the weight coefficient of the health decay inhibition item or the temperature gradient balance item; through the gradient feedback mechanism of the federal collaborative optimization module, the weight ratio of health decay and temperature balance in the cloud global model is updated, and the updated model parameters are sent to the edge nodes to achieve adaptive optimization of multi-region collaborative strategies.

[0041] In this implementation, a risk penalty factor related to the risk level is generated: the battery cluster-level risk assessment results reflect the relationship between the health status of the battery cluster and the current temperature, aging, and current. Based on these assessment results, a risk penalty factor is generated, which is nonlinearly positively correlated with the risk level of the battery cluster. In other words, when the risk level of the battery cluster increases, the risk penalty factor will also increase accordingly. This is because more optimization measures are required to prevent failures or overheating in high-risk states. The formula is: ;in: is the risk penalty factor. is the risk level of the battery, which is usually calculated based on battery aging, temperature and other parameters. Function Represents a nonlinear positive correlation, reflecting the relationship between the risk level and the penalty factor. Dynamically adjust the weight distribution function: The risk penalty factor generated during the risk assessment process is input into the weight distribution function of the federated collaborative optimization module. This function dynamically adjusts the weight distribution between the optimization objectives according to the current risk level. Specifically: If the risk level is high, the system will increase the weight of the health decay suppression item and focus on the control of battery decay. If the temperature anomaly is more serious, the system may increase the weight of the temperature gradient balancing item to avoid the risk of overheating. Gradient feedback mechanism and global model update: In the federated collaborative optimization module, the gradient feedback mechanism is used to update the optimization model based on the current risk penalty factor and the adjusted weight. This means that the model calculates the gradient based on real-time data (such as the health status of the battery, temperature distribution, etc.) and transmits it back to the cloud, and then adjusts the parameters of the entire model on the cloud. This process continuously optimizes the weight ratio of the health decay suppression and temperature balance objectives to ensure that the optimization objectives always meet the actual needs of the current battery cluster. The formula represents: ;in: are the parameters of the optimization model, involving health decay and temperature equilibrium. is the learning rate, which controls the step size of the gradient update. It is the gradient of the loss function with respect to the parameters, reflecting the direction of improvement of the model performance. is a loss function, which is calculated based on the health and temperature of the battery cluster. Send updated model parameters to edge nodes: The updated global model parameters are sent to edge nodes through the federated collaborative optimization module. The edge nodes adjust the charging and balancing behavior of the local battery cluster based on the updated weights and optimization strategies to adapt to the real-time risks and needs of each region. This combination of edge computing and global optimization can realize the adaptive optimization strategy of the battery cluster.

[0042] In summary, this application has at least the following effects: A multi-regional battery charging monitoring and regulation system monitors the health status of battery clusters in real time and performs comprehensive optimization based on factors such as battery attenuation, temperature distribution, and grid demand, effectively suppressing battery attenuation, extending battery life, and improving battery efficiency. The system uses cross-region thermal coupling correlation analysis and temperature gradient balancing strategy to optimize the temperature distribution of battery packs in real time to avoid overheating and ensure the safe operation of the battery system. The system uses a federal collaborative optimization framework to dynamically adjust the charging and discharging strategies of each region, and automatically optimizes resource allocation between regions according to risk levels and temperature changes to achieve intelligent collaboration and load balancing among multiple regions. Through the feedback mechanism of risk assessment model and risk penalty factor, the charging parameters of the battery cluster are adjusted in time to prevent safety risks caused by overcharging or excessive temperature, and ensure the stability and safety of system operation. Through gradient feedback and dynamic weight adjustment mechanism, the system can automatically adapt to the health status and demand changes of different battery clusters, optimize charging strategies, and realize adaptive adjustment of global and local strategies, thereby improving the efficiency and intelligence level of overall battery management. The system uses edge computing and cloud collaborative optimization to make the management of battery clusters more intelligent and efficient, and reduces power consumption and improves processing efficiency through optimization algorithms to reduce unnecessary energy consumption. Through the combination of fuzzy control and optimization algorithms, strategies can be flexibly adjusted according to the needs of specific scenarios, making the system highly scalable and able to adapt to battery management needs of different scales.

[0043] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0044] The present invention is described with reference to flowcharts and / or block diagrams of systems, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0045] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0046] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0047] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0048] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A multi-region battery charging monitoring and regulation system, characterized in that: It includes the following modules: multimodal perception module, federated collaborative optimization module, dynamic hybrid equilibrium strategy module, and cross-domain security decision-making module; The multimodal sensing module is used to synchronously collect electrochemical impedance spectroscopy, infrared thermal imaging and electrothermal parameters in the battery partition, and construct the physical connection and thermal conduction topological relationship of the battery pack through the graph neural network, and output the monomer aging degree map and cross-region thermal coupling correlation characteristics; The federal collaborative optimization module is used to aggregate multi-region data according to the monomer aging degree map and cross-region thermal coupling correlation characteristics, integrate the three optimization goals of battery health attenuation suppression, temperature gradient balance and real-time grid demand, and generate safety constraint parameters of dynamic charging current and priority parameters of inter-regional balancing mode; The dynamic hybrid balancing strategy module is used to dynamically switch between active energy transfer and passive risk suppression modes according to the inter-regional balancing mode priority parameters through a fuzzy control algorithm. In the active mode, cross-regional energy redistribution is achieved through an inductor-capacitor network. In the passive mode, abnormal monomers are quickly isolated based on charge state prediction. The cross-domain safety decision module is used to comprehensively determine the battery cluster-level risk level according to the safety constraint parameters of the dynamic charging current, trigger local load reduction and global strategy reset instructions, and feedback the risk penalty factor to the federal collaborative optimization module to dynamically adjust the weight distribution of health attenuation and temperature balance.

2. The multi-region battery charging monitoring and regulation system according to claim 1, characterized in that: The specific process of constructing the physical connection and thermal conduction topological relationship of the battery pack through the graph neural network is as follows: Each single cell in the battery pack is defined as a node of the graph neural network, and the node attributes include real-time voltage and internal resistance parameters; the physical connection relationship and heat conduction path of adjacent single cells are defined as edges, and the edge weights are dynamically calculated through the voltage fluctuation correlation and the temperature gradient changes of infrared thermal imaging data to generate a topological network of the battery pack coupling characteristics.

3. A multi-region battery charging monitoring and regulation system according to claim 2, characterized in that: The specific process of outputting the monomer aging degree map and cross-region thermal coupling correlation characteristics is as follows: The aging coefficient of the single battery is inverted based on the electrochemical impedance spectroscopy data, and the voltage-internal resistance correlation features extracted from the graph neural network topology network are combined to generate a heat map of the aging degree distribution with regions as units; Through the calculation of heat diffusion intensity based on infrared thermal imaging data and topological network edge weights, a cross-regional thermal coupling correlation matrix is ​​generated to identify high-risk heat diffusion paths.

4. The multi-region battery charging monitoring and regulation system according to claim 3 is characterized in that: According to the monomer aging degree map and cross-region thermal coupling correlation characteristics, the specific process of multi-region data aggregation is as follows: Perform cross-regional hierarchical aggregation on the aging degree distribution heat map and the cross-regional thermal coupling correlation matrix: Regional layer: align the aging distribution characteristics of each region and generate a global aging trend map by weighting; Cross-domain layer: extract high-risk paths in the thermal coupling association matrix and construct cross-regional thermal interference chain relationships.

5. The multi-region battery charging monitoring and regulation system according to claim 4, characterized in that: The three optimization goals of integrating battery health attenuation suppression, temperature gradient balancing and real-time grid demand are as follows: The global aging trend map is mapped into the health decay inhibition item, the thermal interference chain relationship is mapped into the temperature gradient balance item, and the real-time demand signal of the power grid is mapped into the charging and discharging cost item. The three optimization objectives are merged into a unified reward function for federal collaborative optimization through a dynamic weight allocation function.

6. A multi-region battery charging monitoring and regulation system according to claim 5, characterized in that: The specific process of generating the safety constraint parameters of the dynamic charging current and the priority parameters of the inter-regional balancing mode is as follows: Based on the unified reward function, through interactive game training between distributed edge nodes and cloud models, the maximum allowable charging current value of each area is output as the safety constraint parameter. Combined with the thermal interference chain relationship and the global aging trend map, the priority score of the inter-regional balancing operation is calculated as the priority parameter.

7. The multi-region battery charging monitoring and regulation system according to claim 6, characterized in that: According to the priority parameters of the inter-regional equilibrium mode, the specific process of dynamically switching between the active energy transfer and passive risk suppression modes through the fuzzy control algorithm is as follows: The priority parameters of the inter-regional equilibrium mode, the thermal coupling correlation matrix and the global aging trend map are used as fuzzy control inputs to define the membership functions of the active mode and the passive mode. The mode switching threshold is calculated through fuzzy rule base matching to trigger the energy transfer path selection or abnormal monomer isolation instructions.

8. The multi-region battery charging monitoring and regulation system according to claim 7, characterized in that: According to the safety constraint parameters of the dynamic charging current, the specific process of comprehensively determining the battery cluster-level risk level and triggering the local load reduction and global strategy reset instructions is as follows: According to the current over-limit ratio, the number of temperature gradient over-limit areas and the number of abnormal aging trend areas in the safety constraint parameters, a multi-level risk assessment model is constructed: Level 1 risk: triggering local load reduction and limiting charging current in abnormal areas; Level 2 risk: Reset the global charging strategy and switch to the backup balancing path.

9. A multi-regional battery charging monitoring and regulation system according to claim 8, characterized in that: The specific process of feeding back the risk penalty factor to the federated collaborative optimization module and dynamically adjusting the weight distribution of health decay and temperature balance is as follows: According to the battery cluster-level risk level assessment results, a risk penalty factor that is nonlinearly positively correlated with the risk level is generated; Input the risk penalty factor into the weight allocation function of the federated collaborative optimization module to dynamically increase the weight coefficient of the health decay suppression item or the temperature gradient balance item; Through the gradient feedback mechanism of the federated collaborative optimization module, the weight ratio of health decay and temperature balance in the cloud global model is updated, and the updated model parameters are sent to the edge nodes to achieve adaptive optimization of multi-region collaborative strategies.

Images