Intelligent management method and system for optimizing cycle life of storage and charging equipment

Abstract

The application discloses an intelligent management method and system for optimizing the cycle life of storage and charging equipment, which comprises the following steps: filtering abnormal values according to electrical parameter thresholds and normalizing the historical operation data of the storage and charging equipment group, extracting performance indicators for clustering and grouping, and determining performance levels; matching computing power nodes according to performance levels to allocate real-time operation data, triggering a hierarchical adjustment mechanism if the node load exceeds the overload threshold, and adjusting until the processing rate difference of all nodes is within the preset fluctuation range; extracting specified degradation indicators of the storage and charging equipment of each performance level, weighting and fusing the comprehensive performance value according to the preset weight, constructing an evaluation matrix mapped with the comprehensive performance value, calculating the cycle life prediction value and comparing it with the preset target value, generating a charge-discharge parameter optimization strategy based on the comparison result, setting differentiated charge-discharge parameters combined with a temperature compensation function, and generating a cycle life management strategy. The method improves the accuracy and efficiency of the cycle life management of the storage and charging equipment.

Description

Technical Field

[0001] The present invention relates to the field of new energy technologies, and in particular to an intelligent management method and system for optimizing the cycle life of storage and charging equipment. Background Art

[0002] In the field of cycle life management of storage and charging equipment, improving equipment durability and energy utilization efficiency is the core link in optimizing the economic efficiency of the energy system.

[0003] On the one hand, existing technical solutions rely on the analysis of single device data, ignoring the mutual influence of performance differences between groups, resulting in a lack of integrity in the optimization strategy. Unified parameter settings are difficult to adapt to the characteristics of equipment in different usage environments, and cannot accurately identify the differences in key degradation indicators. On the other hand, the multi-dimensional operating data generated by massive equipment needs to be efficiently integrated, but the uneven distribution of computing resources leads to analysis delays. The response is slow when processing large-scale data, and it cannot meet the real-time needs of dynamic working conditions. In addition, the dual limitations of performance difference identification and data processing capabilities result in a lag in the generation process of charge and discharge parameters. The lack of an environmental adaptability mechanism further reduces the accuracy of cycle life prediction and strategy formulation. Summary of the Invention

[0004] The present invention provides an intelligent management method and system for optimizing the cycle life of storage and charging equipment, so as to improve the accuracy and efficiency of cycle life management of storage and charging equipment.

[0005] In a first aspect, in order to solve the above technical problems, the present invention provides an intelligent management method for optimizing the cycle life of storage and charging equipment, comprising:

[0006] Obtain historical operating data of the storage and charging equipment group, filter outliers based on electrical parameter thresholds and normalize them to form a standard set, extract performance indicators for clustering and grouping, and determine performance levels through group feature consistency verification;

[0007] Match computing power nodes by performance level to distribute real-time operating data, monitor node load in real time, and trigger a hierarchical adjustment mechanism if the load exceeds the threshold until the processing rate differences of all nodes are within the preset fluctuation range;

[0008] When all nodes are in a balanced state, the designated degradation indicators of the storage and charging equipment of each performance level are extracted, and a comprehensive performance value is generated by weighted fusion according to preset weights, and an evaluation matrix mapping the storage and charging equipment to the comprehensive performance value is constructed;

[0009] Based on the comprehensive performance value in the evaluation matrix, a cycle life prediction value is calculated, which is compared with a preset target value. Based on the comparison result, a charge and discharge parameter optimization strategy is generated. Differentiated charge and discharge parameters are set in combination with a temperature compensation function to generate a cycle life management strategy.

[0010] As an optional implementation, the acquisition of historical operating data of the storage and charging equipment group, filtering outliers according to electrical parameter thresholds and normalizing the data to form a standard set includes:

[0011] Collect historical operating data of the storage and charging equipment group through a preset data interface, wherein the operating data includes charge and discharge current, voltage fluctuation, temperature change and cycle number;

[0012] Setting operating condition threshold boundaries for the charge and discharge current and the voltage fluctuation, and in response to detecting that the collected charge and discharge current historical data and / or voltage fluctuation historical data exceeds the operating condition threshold boundaries, marking the data as abnormal data and removing the data, thereby obtaining a cleaned data set;

[0013] The temperature changes and cycle times in the cleaning data set are dimensionally normalized to form a standard set.

[0014] As an optional implementation, the extracted performance indicators are clustered and grouped, and the performance level is determined by performing a group feature consistency check, including:

[0015] Extracting key performance indicator vectors from the standard set, calculating the indicator similarity between devices using an unsupervised clustering algorithm, aggregating devices with indicator similarity higher than a cutoff value into similar clusters, and separating devices with indicator similarity lower than the cutoff value into different clusters;

[0016] Extract group common labels based on the indicator distribution characteristics of similar cluster devices, and generate initial level labels according to preset performance level mapping rules;

[0017] Detect the discreteness of device indicators within the same cluster. In response to the discreteness exceeding the tolerance threshold, trigger the cluster re-division mechanism until the discrete permission conditions are met, and output the device stratification results with performance level labels to determine the performance level of each device in the storage and charging device group.

[0018] As an optional implementation, the method of matching computing power nodes by performance level to allocate real-time operation data, monitoring node load in real time, and triggering a hierarchical adjustment mechanism if the load exceeds a threshold until the processing rate differences of all nodes are within a preset fluctuation range includes:

[0019] Establish a mapping rule between performance level and node computing power, and allocate real-time operation data of different performance levels to corresponding computing power nodes;

[0020] Real-time collection of resource occupancy and task processing rate of each node;

[0021] In response to detecting that a resource occupancy rate of a node exceeds a dynamic load threshold, performing a data migration operation;

[0022] After migration, the task processing rate differences between nodes are re-detected. If the difference exceeds the permitted fluctuation range, a proportional fine-tuning operation is performed.

[0023] The data migration operation and the ratio fine-tuning operation are executed cyclically until the task processing rate differences of all nodes are within a preset fluctuation range, and the current resource allocation status of all nodes is locked.

[0024] As an optional implementation, the establishing of a mapping rule between performance level and node computing power includes: establishing a mapping rule that a higher performance level corresponds to a higher node computing power;

[0025] The performing of the data migration operation includes: transferring a preset proportion of data in the overloaded node to a non-overloaded node;

[0026] The execution ratio fine-tuning operation includes: adjusting the data distribution ratio transferred to each of the non-overloaded nodes according to a preset range.

[0027] As an optional implementation manner, when all nodes are in a balanced state, extracting the specified degradation index of the storage device of each performance level includes:

[0028] In response to receiving a lock signal for the resource allocation status of all nodes, confirming that all nodes are in a load-balanced state;

[0029] The storage device corresponding to the performance level classification index is accessed to the storage area of ​​the corresponding node to extract the specified decay indicator data set;

[0030] The designated decay indicator data set includes capacity decay rate, internal resistance change slope, and charge and discharge efficiency.

[0031] As an optional implementation manner, the weighted fusion according to preset weights to generate a comprehensive performance value and construct an evaluation matrix mapping the storage and charging equipment to the comprehensive performance value include:

[0032] Performing dimension normalization conversion on each indicator of the designated recession indicator data set to form a standardized indicator vector;

[0033] Perform weighted fusion calculation on the standardized index vector according to the preset weight coefficient to generate the comprehensive performance value of each storage and charging device;

[0034] Based on the mapping relationship between equipment numbers and comprehensive performance values, a determinant matrix structure is constructed to form an evaluation matrix;

[0035] The row index of the determinant-column matrix structure is the unique device number, and the column data is the corresponding comprehensive performance value.

[0036] As an optional embodiment, the cycle life prediction value is calculated based on the comprehensive performance value in the evaluation matrix, the predicted value is compared with the preset target value, and a charge and discharge parameter optimization strategy is generated based on the comparison result, including:

[0037] Analyzing the mapping relationship between the equipment number and the comprehensive performance value in the evaluation matrix, and inputting the comprehensive performance value into the life prediction model to calculate the corresponding cycle life prediction value;

[0038] The preset cycle life target value library is retrieved for comparison. If the cycle life prediction value is higher than the target value, a positive charge and discharge power adjustment instruction is generated. If the cycle life prediction value is not higher than the target value, a negative charge and discharge power adjustment instruction is generated.

[0039] As an optional implementation, the step of setting differentiated charge and discharge parameters in combination with a temperature compensation function to generate a cycle life management strategy includes:

[0040] Collect ambient temperature data in real time and dynamically calculate the compensation coefficient through the temperature compensation function;

[0041] Determine a current ratio reference value according to the equipment performance level, dynamically modify the current ratio reference value by applying a power positive / negative adjustment instruction, and generate a final current limit value in combination with the compensation coefficient;

[0042] Configure discharge depth thresholds according to performance level;

[0043] The final current limit is dynamically fused with the discharge depth threshold, and if it is verified that the fusion result complies with the electrical safety operation boundary, a cycle life management strategy is generated.

[0044] In a second aspect, the present invention provides an intelligent management system for optimizing the cycle life of storage and charging equipment, comprising:

[0045] The data acquisition unit is used to obtain the historical operating data of the storage and charging equipment group, filter outliers according to the electrical parameter threshold and normalize them to form a standard set, extract performance indicators for clustering and grouping, and determine the performance level through group feature consistency verification;

[0046] The processing unit is used to match computing nodes according to performance levels to allocate real-time operating data, monitor node load in real time, and trigger a hierarchical adjustment mechanism if the load exceeds a threshold until the processing rate differences of all nodes are within a preset fluctuation range; when all nodes are in a balanced state, extract the specified degradation indicators of storage and charging equipment of each performance level, weightedly fuse them according to preset weights to generate a comprehensive performance value, and construct an evaluation matrix mapping the storage and charging equipment to the comprehensive performance value;

[0047] A management strategy generation unit is used to calculate a cycle life prediction value based on the comprehensive performance value in the evaluation matrix, compare it with a preset target value, generate a charge and discharge parameter optimization strategy based on the comparison result, set differentiated charge and discharge parameters in combination with a temperature compensation function, and generate a cycle life management strategy.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] The present invention provides an intelligent management method for optimizing the cycle life of storage and charging equipment. It adopts clustering grouping and group feature consistency verification mechanism to accurately divide the performance level of equipment and overcome the problem of lack of integrity of the strategy caused by single equipment analysis. Based on the performance level, computing power nodes are dynamically allocated and hierarchical load adjustment is implemented. Data migration is fine-tuned to achieve millisecond-level load balancing and eliminate the bottleneck of uneven distribution of computing resources. An evaluation matrix driven by the fusion of decay indicators is constructed to drive life prediction. The temperature compensation function and environmental adaptive calibration are combined to generate differentiated charging and discharging parameters. The electrical safety boundary verification is simultaneously passed to solve the technical problems of hysteresis and insufficient environmental adaptability of unified parameter settings. This solution forms a full-process closed loop of precise classification-efficient calculation-dynamic optimization, which significantly improves the accuracy and efficiency of cycle life management of storage and charging equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a flow chart of an intelligent management method for optimizing the cycle life of storage and charging equipment provided by an embodiment of the present invention;

[0051] Figure 2 1 is a flow chart of a method for allocating real-time operation data provided by an embodiment of the present invention;

[0052] Figure 3 1. It is a flow chart of a method for extracting designated degradation indicators of storage and charging equipment of various performance levels provided by an embodiment of the present invention;

[0053] Figure 4 1 is a flow chart of a method for constructing an evaluation matrix provided by an embodiment of the present invention;

[0054] Figure 5 This is a flow chart of a method for generating a charge and discharge parameter optimization strategy provided by an embodiment of the present invention;

[0055] Figure 6 This is a flow chart of a method for generating a cycle life management strategy provided by an embodiment of the present invention;

[0056] Figure 7 This is a structural diagram of an intelligent management system for optimizing the cycle life of storage and charging equipment provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0058] In the field of cycle life management of storage and charging equipment, improving equipment durability and energy utilization efficiency is the core link in optimizing the economic efficiency of the energy system.

[0059] On the one hand, existing technical solutions rely on the analysis of single device data, ignoring the mutual influence of performance differences between groups, resulting in a lack of integrity in the optimization strategy. Unified parameter settings are difficult to adapt to the characteristics of equipment in different usage environments, and cannot accurately identify the differences in key degradation indicators. On the other hand, the multi-dimensional operating data generated by massive equipment needs to be efficiently integrated, but the uneven distribution of computing resources leads to analysis delays. The response is slow when processing large-scale data, and it cannot meet the real-time needs of dynamic working conditions. In addition, the dual limitations of performance difference identification and data processing capabilities result in a lag in the generation process of charge and discharge parameters. The lack of an environmental adaptability mechanism further reduces the accuracy of cycle life prediction and strategy formulation.

[0060] In order to solve the above problems, an intelligent management method for optimizing the cycle life of storage and charging equipment provided in an embodiment of the present application will be introduced and explained in detail through the following specific embodiments.

[0061] Reference Figure 1 The first embodiment of the present invention provides an intelligent management method for optimizing the cycle life of storage and charging equipment, comprising the following steps:

[0062] S1: Obtain historical operating data of the storage and charging equipment group, filter outliers based on electrical parameter thresholds and normalize them to form a standard set, extract performance indicators for clustering, and determine performance levels through group feature consistency verification;

[0063] S2 matches computing nodes by performance level to distribute real-time operating data, monitors node load in real time, and triggers a hierarchical adjustment mechanism if the load exceeds the threshold until the processing rate differences of all nodes are within the preset fluctuation range;

[0064] S3, when all nodes are in a balanced state, extracts the designated degradation indicators of storage and charging equipment at each performance level, generates a comprehensive performance value by weighted fusion according to preset weights, and constructs an evaluation matrix mapping storage and charging equipment to the comprehensive performance value;

[0065] S4, calculates the cycle life prediction value based on the comprehensive performance value in the evaluation matrix, compares it with the preset target value, generates a charge and discharge parameter optimization strategy based on the comparison result, sets differentiated charge and discharge parameters in combination with the temperature compensation function, and generates a cycle life management strategy.

[0066] It should be noted that the standard set refers to a structured data set that has been filtered and normalized for outliers, and is used to characterize the performance status of the equipment. The standard set includes at least: charging and discharging current, in amperes; voltage fluctuation, in volts; temperature change, in degrees Celsius; and number of cycles, dimensionless. Group feature consistency check refers to a closed-loop process that verifies whether the performance dispersion of the equipment in the group meets the technical tolerance requirements by calculating the standard deviation of the key performance indicators of the same cluster equipment. The standard deviation is used for dispersion calculation. The hierarchical adjustment mechanism refers to a load balancing control process that includes a two-stage linkage of data migration and proportion fine-tuning. The key operations of the hierarchical adjustment mechanism include: data migration, which transfers more than 10% of the data volume in the overloaded node, in MB; proportion fine-tuning, which adjusts the data distribution ratio by no more than 5%. The preset fluctuation range refers to the maximum allowable deviation range of the task processing rate difference between nodes. The rate difference rate is defined by the following formula,

[0067]

[0068] in, is the rate difference rate, is the rate of node A, is the Node B rate, is the average rate of all nodes.

[0069] It should be noted that the designated degradation indicators include: capacity attenuation rate, which refers to the percentage difference between the rated capacity and actual capacity of the device. The internal resistance change slope refers to the resistance change per unit time, and the unit is milliohm / year, mΩ / year. The charge and discharge efficiency refers to the output energy / input energy ratio. The evaluation matrix refers to a two-dimensional data structure with the unique device number as the row index and the comprehensive performance value as the column element. The comprehensive performance value is a dimensionless scalar, which is the result of normalized weighted calculation. The temperature compensation function is defined as the temperature compensation coefficient, which is calculated by the following formula,

[0070]

[0071] in, is the temperature compensation coefficient, is the temperature influence coefficient, the unit is % / ℃, the default value is 0.02, is the real-time ambient temperature in °C. The reference temperature is in °C and the default value is 25.

[0072] In some embodiments, the method is executed by the server and includes the following steps: Obtaining historical operating data of a group of storage and charging equipment, the historical operating data including multi-dimensional parameters such as charging and discharging current, voltage fluctuation, temperature change and number of cycles. Filtering outliers on the charging and discharging current and voltage fluctuation data through a data preprocessing module: If the parameter value exceeds the preset electrical parameter threshold, for example, the current threshold is 5-60 amperes and the voltage fluctuation threshold is 2-4 volts, it is marked as abnormal and eliminated. Normalizing parameters such as temperature change and number of cycles, for example, mapping the temperature of 20-35 degrees Celsius to the range of 0-1 to form a standard set. Extracting performance indicators from the standard set, and using a cluster analysis algorithm to perform similarity comparison on the device groups: If the similarity of performance indicators between devices exceeds a preset threshold, for example, 80%, they are classified into the same performance level; if it is lower than the threshold, they are divided into different levels. After checking the consistency of group features, for example, verifying the logical correlation between the performance level and the temperature change rate, the final performance level stratification result is determined.

[0073] As a feasible embodiment, a multi-level data processing architecture is established according to performance levels. The real-time operating data of devices of different levels are distributed to matching computing nodes through an adaptive load balancing algorithm. For example, the data of high-performance level devices are distributed to nodes 1-2, and the data of medium and low performance levels are distributed to nodes 3-5. The processing load of each node is monitored in real time. If the load of a node exceeds the preset load threshold, such as 80%, the hierarchical adjustment mechanism is triggered: part of the data is dynamically reallocated to the low-load node, for example, the data is transferred from the 85% load node to the 30% load node. Repeat the adjustment until the processing rate differences of all nodes are within the preset fluctuation range, for example, the difference does not exceed ±5%.

[0074] As an optional embodiment, after all nodes reach equilibrium, specific degradation indicators are extracted for each performance level of storage and charging equipment, including capacity decay rate and internal resistance change trend. A weight allocator is used to assign weight coefficients to these indicators. For example, a weight of 0.5 for capacity decay rate and 0.3 for internal resistance change trend. After normalization and dimensionality unification, these indicators are weighted and fused to generate a comprehensive performance value, for example, a weight of 0.545. An evaluation matrix is ​​constructed, mapping storage and charging equipment to comprehensive performance values, using device numbers as row indexes and comprehensive performance values ​​as column data.

[0075] As an optional embodiment, a cycle life prediction value is calculated based on an evaluation matrix. For example, if a device has a predicted life of 3500 cycles, it is compared with a preset target value, such as 3000 cycles. If the predicted value is higher than the target value, a charge and discharge power increase plan is generated, such as increasing it from 50kW to 60kW; if it is lower than the target value, a charge and discharge intensity reduction plan is generated, such as reducing it from 65kW to 55kW. Combined with the temperature compensation function, the compensation coefficient is calculated using the aforementioned formula, and differentiated charge and discharge parameters are set. For example, in a high temperature environment, the charging current is adjusted from 2500mA to 2000mA, generating a cycle life management strategy that includes charging current limit and discharge depth control.

[0076] As a specific example, during the data preprocessing phase, sensors collect real-time charging and discharging current data from a group of storage and charging devices. When a current value exceeds an electrical parameter threshold, such as 70 amps exceeding the upper limit of 60 amps, the data preprocessing module automatically flags and removes the abnormal data. For example, temperature parameters are normalized, converting the actual temperature value of 25 degrees Celsius to a standard value of 0.2, eliminating the impact of dimensional differences on cluster analysis.

[0077] In some embodiments, during the performance grading phase, a cluster analysis algorithm is used to calculate the parameter similarity between two devices. If the similarity calculated based on the number of cycles and the temperature change rate is 85%, exceeding the 80% threshold, they are classified into the same performance grade. For example, a group feature consistency check revealed that a device was mistakenly classified as excellent: its temperature change rate of 0.6 was far higher than the average of 0.3 for similar devices. Based on this, it was adjusted to good, ensuring that the grade classification is consistent with the actual status of the device.

[0078] As a feasible embodiment, during the load balancing process, for example, when the load of node 1 reaches 85%, 10% of the data processing tasks are automatically migrated to node 5, which has a load of only 20%, reducing the node load to 75% and 50%, respectively. The computing node configuration is dynamically adjusted to temporarily increase the memory allocation of the high-load node by 20%, while reducing the priority of non-urgent tasks on other nodes.

[0079] As a feasible embodiment, during the parameter optimization phase, for example, for devices whose predicted cycle life is lower than the target value, the charge and discharge power adjustment rule generator uses the intensity control logic to reduce the discharge depth from 80% to 70%, while also setting a lower capacity threshold, such as no less than 800mAh. For example, when the ambient temperature rises to 35 degrees Celsius, the temperature compensation function automatically generates a compensation coefficient of 0.8, adjusting the charging current of high-performance devices from 2500mA to 2000mA to prevent battery overheating damage.

[0080] As a feasible implementation, differentiated parameters are integrated to form a personalized set of charge and discharge control instructions. For example, a charging current limit of 2500mA and a depth of discharge of 80% are set for high-performance devices, while a limit of 1500mA and a depth of discharge of 70% are set for medium-performance devices. For example, when the device's State of Health (SOH) drops to 80%, the system automatically triggers a maintenance mechanism: the charging current limit is reduced to 70% of the original value, such as from 2500mA to 1750mA, and a maintenance alert is sent to the management platform. The resulting cycle life management strategy is linked to the device's operating status in real time, forming a closed-loop control system.

[0081] The above-mentioned intelligent management method for optimizing the cycle life of storage and charging equipment adopts a clustering grouping and group feature consistency verification mechanism to accurately divide the performance level of equipment and overcome the problem of lack of overall strategy caused by single equipment analysis. Based on the performance level, computing power nodes are dynamically allocated and hierarchical load adjustment is implemented. Data migration is fine-tuned to achieve millisecond-level load balancing and eliminate the bottleneck of uneven distribution of computing resources. An evaluation matrix driven by the fusion of decay indicators is constructed to drive life prediction. The temperature compensation function and environmental adaptive calibration are combined to generate differentiated charging and discharging parameters. The electrical safety boundary verification is simultaneously passed to solve the technical problems of hysteresis and insufficient environmental adaptability of unified parameter settings. This solution forms a closed loop of the entire process of precise classification-efficient calculation-dynamic optimization, which significantly improves the accuracy and efficiency of the cycle life management of storage and charging equipment.

[0082] According to some embodiments of the present invention, in step S1, historical operating data of a group of storage and charging devices is obtained, abnormal values ​​are filtered and normalized according to electrical parameter thresholds to form a standard set, including:

[0083] S11, collecting historical operating data of the storage and charging equipment group through a preset data interface, wherein the operating data includes charging and discharging current, voltage fluctuation, temperature change and cycle number;

[0084] S12, setting operating condition threshold boundaries for charge and discharge current and voltage fluctuations, and in response to detecting that the collected charge and discharge current historical data and / or voltage fluctuation historical data exceeds the operating condition threshold boundaries, marking them as abnormal data and removing them to obtain a cleaned data set;

[0085] S13, performing dimension normalization conversion on the temperature change and the number of cycles in the cleaning data set to form a standard set.

[0086] In some embodiments, the method for optimizing the cycle life of storage and charging equipment implements data standardization preprocessing through the following steps: Historical operating data of a group of storage and charging equipment is collected through a pre-set data interface, including charge and discharge current, voltage fluctuation, temperature change, and cycle count. The operating thresholds for charge and discharge current are set between 5 and 60A, and the voltage fluctuation threshold is set between 2 and 4V. When a current value of 70A or a voltage value below 2V is detected, it is automatically marked as abnormal data and removed to generate a cleaned data set.

[0087] As a feasible implementation, the temperature variations and cycle counts in the cleaning dataset are dimensionally normalized. For example, if the temperature range is [20°C, 35°C], they are linearly mapped to the interval [0, 1]. For example, 25°C-30°C is converted to 0.2-0.4. The cycle count is also normalized based on the maximum design life. For example, if the design life is 1000 cycles, 500 cycles is mapped to 0.5. This ultimately forms a standard set, eliminating parameter dimensional differences.

[0088] As a specific embodiment, data reliability is achieved through a dual protection mechanism. Abnormal filtering mechanism: Taking current overlimit as an example, when a 70A current record is detected, which exceeds the 60A threshold, the data point is immediately removed; taking voltage abnormality as an example, when a 1.8V voltage record is detected, which is lower than the 2V threshold, it is automatically marked for removal. Correlation verification mechanism: For example, the deviation tolerance rate is set to ±5%. When the current is 40A, the theoretical voltage should be 3.2V. If the actual record is 2.8V, it is corrected to 3.0V through linear interpolation; after verification, the data deviation rate is controlled within 2% to ensure the credibility of the standard set.

[0089] This method for forming a standard set precisely filters out abnormal current and voltage data using operating condition thresholds, eliminating noise interference. Combined with dimensional normalization of temperature and cycle count, it addresses analytical obstacles caused by inconsistent unit values ​​for multiple source parameters. The resulting standard set provides a consistent and comparable data foundation for precise equipment performance grading, addressing technical shortcomings such as inefficient multidimensional data integration and insufficient parameter consistency.

[0090] According to some embodiments of the present invention, in step S1, the performance indicators are extracted for clustering and grouping, and the performance level is determined by performing a group feature consistency check, including:

[0091] S14, extracting key performance indicator vectors from the standard set, using an unsupervised clustering algorithm to calculate the indicator similarity between devices, clustering devices with indicator similarity higher than a cutoff value into similar clusters, and separating devices with indicator similarity lower than the cutoff value into different clusters;

[0092] S15, extracting group common labels based on the indicator distribution characteristics of similar cluster devices, and generating initial level labels according to preset performance level mapping rules;

[0093] S16, detects the dispersion of device indicators within the same cluster. In response to the dispersion exceeding the tolerance threshold, triggers the cluster re-division mechanism until the discrete permission conditions are met, and outputs the device stratification results with performance level labels to determine the performance level of each device in the storage and charging device group.

[0094] It is important to understand that discrete licensing conditions refer to the standard deviation of key device indicators within a cluster being less than or equal to a preset tolerance threshold. For example, the standard deviation threshold for the number of cycles is less than or equal to 50.

[0095] In some embodiments, a method for determining the performance level of storage and charging equipment includes the following steps: extracting key performance indicator vectors from a standard set, including current stability and voltage range, and calculating the similarity of indicators between devices using an unsupervised clustering algorithm. A cutoff value is set at 80%, and devices with similarity above the cutoff value are clustered into similar clusters, while devices with similarity below the cutoff value are separated into different clusters.

[0096] As a feasible embodiment, common labels are extracted for groups based on the distribution characteristics of device indicators in similar clusters. For example, if the number of cycles of devices in cluster A is greater than 1000, they are labeled "High Cycle Durability." For example, if the temperature change rate of devices in cluster B is less than 5% per month, they are labeled "Low Thermal Sensitivity." Based on the preset performance level mapping rules, excellent is defined as cycles greater than 1000, and poor is defined as cycles less than 500, generating initial level labels.

[0097] As a specific example, the dispersion verification and reclassification mechanism performs the following: Dispersion detection calculates the standard deviation of key indicators for devices in the same cluster, such as the standard deviation of cycle counts, with a tolerance threshold of 50. If the standard deviation of cycle counts within a cluster reaches 80, such as 1200 for device 1 and 400 for device 2, the dispersion is determined to have exceeded the threshold. Cluster reclassification removes abnormal devices, such as those with a cycle count of 400, from the excellent cluster and reclassifies them into the poor cluster. Cluster indicator dispersion is recalculated until the standard deviation is ≤ 50. The device stratification results are output with performance level labels (Excellent / Good / Poor).

[0098] This performance grading method ensures consistent performance across similar devices through an unsupervised clustering algorithm and similarity cutoffs, eliminating blind spots in identifying differences between groups. It also incorporates a group feature consistency verification mechanism, discrete tolerance threshold control, and cluster re-division to dynamically correct for grading deviations. The resulting device stratification provides a reliable basis for subsequent resource scheduling and personalized parameter optimization, avoiding technical issues such as inability to identify device differences and a lack of integrated strategy.

[0099] like Figure 2As shown, in step S2, computing power nodes are matched according to performance levels to allocate real-time operation data, and node loads are monitored in real time. If the load exceeds the threshold, a hierarchical adjustment mechanism is triggered until the processing rate differences of all nodes are within the preset fluctuation range, including:

[0100] S21, establish a mapping rule between performance level and node computing power, and allocate real-time operation data of different performance levels to corresponding computing power nodes;

[0101] S22, real-time collection of resource occupancy and task processing rate of each node;

[0102] S23, in response to detecting that the resource occupancy of a node exceeds a dynamic load threshold, performing a data migration operation;

[0103] S24, after the migration, re-detect the difference in task processing rates between the nodes. If the difference exceeds the permitted fluctuation range, perform a proportional fine-tuning operation.

[0104] S25, cyclically executing the data migration operation and the ratio fine-tuning operation until the task processing rate differences of all nodes are within a preset fluctuation range, and locking the current resource allocation status of all nodes.

[0105] It's important to understand that the dynamic load threshold refers to the upper limit of resource utilization, which is dynamically calculated based on the node's hardware configuration. It's measured in terms of CPU (Central Processing Unit) utilization or memory utilization.

[0106] In some embodiments, a method for matching computing nodes by performance level to distribute real-time operating data includes: establishing a mapping rule between performance level and node computing power, wherein higher performance levels are matched to higher node computing power. Real-time operating data of different performance levels is distributed to corresponding computing power nodes, for example, data for high-performance devices is distributed to the first through third nodes, and data for medium- and low-performance devices is distributed to the fourth through tenth nodes. Resource utilization and task processing rates of each node are collected in real time, and a dynamic load threshold is set at 75%. In response to detecting that the resource utilization of a node exceeds the dynamic load threshold, for example, the utilization of the first node reaches 80%, a data migration operation is performed: a preset proportion (e.g., 10%) of data from the overloaded node is transferred to a fifth node with a resource utilization below the threshold (e.g., 20%). After the migration, the task processing rate differences between the nodes are re-detected. If the difference exceeds an allowable fluctuation range (±5%), a ratio fine-tuning operation is performed: the data allocation proportion transferred to each non-overloaded node is adjusted by 5%. The data migration and ratio fine-tuning operations are repeated until the task processing rate differences of all nodes are within the preset fluctuation range, locking the current resource allocation state.

[0107] As a specific embodiment, the dynamic adjustment process of this method includes the following: in the initial allocation plan, high-level device data is allocated to the first through third nodes. During operation, it is detected that the resource utilization rate of the second node has reached 85%, triggering a data migration operation to transfer 15% of the data to the previously idle tenth node. After the migration, re-testing reveals that the processing rate difference between the first node (70%) and the tenth node (45%) has reached 10%, triggering a ratio fine-tuning operation to adjust 5% of the high-level data from the first node to the tenth node, ultimately stabilizing the processing rate difference of all nodes within a 3% range.

[0108] As a specific example, during the load redistribution phase, it was discovered that the fifth node's resource utilization was consistently high due to processing complex tasks. Through proportional fine-tuning, 8% of the data from mid-level devices was transferred to the sixth node, and 15% of the memory resources were dynamically increased for the fifth node, reducing the difference in task processing rates between nodes to 4%.

[0109] As a specific example, after completing multiple rounds of data migration and fine-tuning, the system locks in resource allocation: high-level device data is fixedly processed by the first and second nodes (50% each), mid-level device data is dynamically allocated to the third through fifth nodes, adjusted in a 35%:35%:30% ratio based on real-time load, and low-level device data is processed by the sixth through tenth nodes in a round-robin fashion. At this point, resource utilization across all nodes remains between 60% and 75%, with a 4% or less variance in task processing rates.

[0110] This method for allocating real-time data significantly improves system resource utilization by establishing a dynamic migration and cyclic adjustment mechanism. Data migration is automatically triggered when node resource utilization exceeds capacity, effectively avoiding processing delays caused by localized node overload. By continuously detecting differences in task processing rates and performing proportional adjustments, the load on all nodes quickly converges to a balanced state. This closed-loop control mechanism enhances the system's adaptability to real-time data fluctuations and provides stable computing power for efficient management of device groups.

[0111] According to several embodiments of the present invention, in step S21, a mapping rule between performance level and node computing power is established, including: establishing a mapping rule that the higher the performance level, the higher the node computing power matched; in step S23, a data migration operation is performed, including: transferring a preset proportion of data in the overloaded node to the non-overloaded node; in step S24, a proportion fine-tuning operation is performed, including: adjusting the data distribution ratio transferred to each non-overloaded node according to a preset amplitude.

[0112] As a feasible implementation, a mapping rule between performance level and node computing power is established. For example, energy storage device groups can be divided into three performance levels: high, medium, and low. A corresponding relationship between these levels and node computing power is established: high-level devices are matched with the top 30% of nodes in terms of processing power, medium-level devices are matched with the middle 40% of nodes, and low-level devices are matched with the bottom 30% of nodes. Given the high frequency and small batch size of data from high-level devices, nodes with real-time processing capabilities are assigned to them. Low-level devices, with their low frequency and large batch size, are assigned to batch processing nodes.

[0113] In some embodiments, a data migration operation is performed. For example, when it is detected that the resource utilization of the third node reaches 80%, which exceeds the 75% threshold, 20% of the data on the node is automatically migrated to the eighth node, which has a utilization rate of only 25%. After the migration, the utilization rate of the third node drops to 65%, while the utilization rate of the eighth node rises to 45%, achieving load balancing.

[0114] In some embodiments, a proportional fine-tuning operation is performed. For example, after migration, it is detected that the processing rate difference between nodes is 8%, which exceeds the 5% permitted range. The high-level device data is fine-tuned from the first node to the seventh node by 3%, so that the processing rate difference is reduced to 4%.

[0115] The aforementioned mapping rule-building method achieves precise matching of data processing requirements with computing resources through a positive correlation between performance level and node computing power. A data migration mechanism with preset ratios can quickly alleviate node overloads while avoiding system turbulence caused by large-scale data transfers. Combined with fine-tuning of data allocation ratios based on amplitude, this approach ensures both responsiveness and precision in resource scheduling. This hierarchical control strategy significantly improves the processing efficiency of heterogeneous device data and lays the foundation for differentiated performance management.

[0116] like Figure 3 As shown, in step S3, when all nodes are in a balanced state, the designated degradation index of the storage and charging equipment of each performance level is extracted, including:

[0117] S31, in response to receiving a lock signal for the resource allocation status of all nodes, confirming that all nodes are in a load balancing state;

[0118] S32, classifying and indexing the corresponding storage device according to the performance level, accessing the storage area of ​​the corresponding node to extract the specified decay indicator data set;

[0119] The designated decay indicator data sets include capacity decay rate, internal resistance change slope, and charge and discharge efficiency.

[0120] In some embodiments, in response to receiving a resource allocation status lock signal from all nodes, it is confirmed that all computing nodes are in a load-balanced state. Storage and charging devices are indexed by performance level, and a storage area of ​​a corresponding node is accessed to extract a specified decay indicator dataset, which includes a capacity decay rate, an internal resistance change slope, and a charge-discharge efficiency.

[0121] For example, when extracting capacity decay rate data, a high-frequency acquisition mode is set for high-performance devices, and a low-frequency acquisition mode is set for medium and low-performance devices. When it is detected that the capacity decay rate of a device exceeds the preset risk threshold, it is automatically marked as a high-risk device and an alarm is triggered. When obtaining the slope of the internal resistance change, a time series curve is generated through continuous monitoring. If the curve shows a continuous upward trend, it is determined that the health status of the device has declined, and the slope value is stored in the data set as a key decline indicator. When calculating the charge and discharge efficiency indicators, the charging efficiency (stored energy / input energy) and the discharge efficiency (output energy / stored energy) are calculated separately, and the product of the two is included in the data set as the comprehensive efficiency value.

[0122] The above method for extracting specific degradation indicators for storage and charging devices at each performance level ensures the stability of the data extraction process through a load balancing state lock mechanism, effectively preventing interruptions in indicator collection due to node fluctuations. Accurately locating device storage locations by performance level classification index significantly improves the efficiency of acquiring degradation indicator datasets. Through the multi-dimensional collaborative collection of capacity decay rate, internal resistance change slope, and charge and discharge efficiency, a complete status profile is provided for performance evaluation, significantly enhancing the reliability of device health analysis.

[0123] like Figure 4 As shown, in step S3, a comprehensive performance value is generated by weighted fusion according to preset weights, and an evaluation matrix mapping storage and charging equipment to the comprehensive performance value is constructed, including:

[0124] S33, performing dimension normalization conversion on each indicator of the designated recession indicator data set to form a standardized indicator vector;

[0125] S34, performing weighted fusion calculation on the standardized indicator vector according to a preset weight coefficient to generate a comprehensive performance value of each storage and charging device;

[0126] S35, based on the mapping relationship between the equipment number and the comprehensive performance value, construct a determinant matrix structure to form an evaluation matrix;

[0127] The row index of the determinant-column matrix structure is the unique device number, and the column data is the corresponding comprehensive performance value.

[0128] As a feasible embodiment, a dimension normalization conversion is performed on a specified decay indicator dataset: the capacity decay rate percentage value, the internal resistance change slope milliohm value, and the charge-discharge efficiency percentage value are uniformly mapped to a standardized range from 0 to 1 to form a standardized indicator vector. The standardized indicator vector is then weighted and fused according to preset weight coefficients, for example, the capacity decay rate is given the highest weight and the charge-discharge efficiency is given the lowest weight, to generate a comprehensive performance value for each storage and charging device.

[0129] As a specific embodiment, a deterministic matrix structure is constructed based on the mapping relationship between device numbers and comprehensive performance values. The evaluation matrix is ​​formed with the device unique number as the row index and the corresponding comprehensive performance value as the column data. A sorting algorithm is used to sort the evaluation matrix in descending order of comprehensive performance value, placing the best-performing device in the first row of the matrix.

[0130] For example, when constructing an evaluation matrix for a specific energy storage cluster, the comprehensive performance value of 0.85 for device A001 corresponds to the top row of the matrix, the value of 0.62 for device B203 is in the middle, and the value of 0.41 for device C307 is in the bottom row. This matrix sorting allows managers to quickly identify devices requiring maintenance. For example, if a device's comprehensive performance value falls below a preset acceptable threshold, it is automatically classified as requiring maintenance. The evaluation matrix also updates the maintenance status indicator, providing a basis for resource scheduling decisions.

[0131] The aforementioned evaluation matrix construction method eliminates bias in the evaluation of multi-source indicators through dimensional normalization, making parameters of different dimensions comparable. Weighted fusion calculations enhance the influence of key degradation indicators, generating comprehensive performance values ​​that objectively reflect the true state of the equipment. A determinant matrix structure constructed based on equipment numbers enables visual priority management of evaluation results, providing structured data support for equipment maintenance decisions and significantly improving the scientific nature of resource allocation.

[0132] like Figure 5 As shown, in step S4, a cycle life prediction value is calculated based on the comprehensive performance value in the evaluation matrix, compared with the preset target value, and a charge and discharge parameter optimization strategy is generated based on the comparison result, including:

[0133] S41, analyzing the mapping relationship between the device number and the comprehensive performance value in the evaluation matrix, and inputting the comprehensive performance value into the life prediction model to calculate the corresponding cycle life prediction value;

[0134] S42, retrieve the preset cycle life target value library for comparison, if the cycle life prediction value is higher than the target value, generate a charge and discharge power positive adjustment instruction, if the cycle life prediction value is not higher than the target value, generate a charge and discharge power negative adjustment instruction.

[0135] In some embodiments, the mapping relationship between the device number and the comprehensive performance value in the evaluation matrix is ​​analyzed, and the comprehensive performance value is input into the life prediction model to calculate the corresponding cycle life prediction value. A preset cycle life target value library is retrieved for comparison. If the cycle life prediction value is higher than the target value, a positive charge and discharge power adjustment instruction is generated; if the cycle life prediction value is not higher than the target value, a negative charge and discharge power adjustment instruction is generated.

[0136] As a feasible embodiment, when executing charge and discharge power adjustment instructions, the power adjustment rule generator determines the adjustment range in combination with the device operating status. For positive adjustment instructions, the charge and discharge power is increased according to the preset rules; for negative adjustment instructions, the charge and discharge intensity is reduced according to the intensity control logic.

[0137] As a specific embodiment, after generating a personalized charge and discharge plan, compliance is verified. If the parameter optimization strategy improves the comprehensive performance value of the equipment to exceed the performance matrix qualification line, the strategy is adopted; otherwise, the charge and discharge intensity parameters are readjusted until they meet the requirements.

[0138] For example, the life prediction model analyzes the overall performance value of a device, 0.85, and calculates a predicted cycle life of 3,800 cycles. The target value of 3,000 cycles for the corresponding model is retrieved from the target value library. Because the predicted value is higher than the target, a positive charge and discharge power adjustment instruction is automatically generated. When the predicted cycle life of another device, 2,800 cycles, is lower than the target value of 3,000 cycles, a negative charge and discharge power adjustment instruction is generated. Considering the device's real-time high-load operation, the power adjustment rule generator reduces the depth of discharge from 80% to 70% according to the intensity control logic.

[0139] As an optional embodiment, when executing a forward regulation instruction, the device power is increased from 50kW to 65kW. If it is detected that the upper limit of the safety range of 60kW is exceeded, the safety mechanism is triggered, and the power is adjusted back to 55kW through the intensity control calculator, and the lower limit of the discharge capacity is set. For example, after completing parameter optimization, the device performance is verified: before adjustment, the comprehensive performance value of 0.58 is lower than the qualified line of 0.6. After limiting the deep discharge and optimizing the charging current, the comprehensive performance value is improved to 0.63. The strategy is determined to be effective and the configuration is locked.

[0140] This method for generating charge and discharge parameter optimization strategies accurately predicts device lifespan through a dynamic analytical evaluation matrix. Combined with an intelligent comparison mechanism using a preset target value library, it automatically generates differentiated charge and discharge power adjustment commands. Positive adjustment commands fully unleash the potential of high-performance devices, while negative adjustment commands effectively suppress the degradation risk of low-performance devices, forming a bidirectional adaptive optimization capability. This closed-loop control strategy, based on real-time linkage between lifespan predictions and target values, significantly improves the targeted nature of parameter optimization and the refinement of device management, providing a key guarantee for extending the overall lifespan of energy storage systems.

[0141] like Figure 6 As shown, in step S4, differentiated charge and discharge parameters are set in combination with the temperature compensation function to generate a cycle life management strategy, including:

[0142] S43, collecting ambient temperature data in real time and dynamically calculating the compensation coefficient through the temperature compensation function;

[0143] S44, determining a current ratio reference value according to the equipment performance level, dynamically correcting the current ratio reference value by applying a power positive / negative adjustment instruction, and generating a final current limit value in combination with a compensation coefficient;

[0144] S45, configure the discharge depth threshold according to the performance level ladder;

[0145] S46, dynamically fuse the final current limit and the discharge depth threshold, and if the fusion result is verified to be consistent with the electrical safety operation boundary, generate a cycle life management strategy.

[0146] It's important to understand that the electrical safety operating boundary refers to the extreme range of electrical parameters within which the device is allowed to operate safely. Key boundary dimensions include: charging current limit ≤ device rated capacity * safety factor, where the safety factor is dimensionless, and depth of discharge threshold ≤ the maximum allowable value (in percent) corresponding to the performance level.

[0147] In some embodiments, ambient temperature data is collected in real time, and a compensation coefficient is dynamically calculated using a temperature compensation function. A current ratio baseline value is determined based on the device's performance level, for example, 0.5 times the capacity value for high-performance devices. This current ratio baseline value is dynamically corrected using positive or negative power adjustment instructions, and the final current limit is generated in combination with the compensation coefficient. Depth of discharge thresholds are configured based on performance levels, for example, high-performance devices are configured with the highest depth of discharge threshold. The final current limit is dynamically merged with the depth of discharge threshold. If the fusion result is verified to meet the electrical safety operating boundaries, a cycle life management strategy is generated.

[0148] As a feasible embodiment, the temperature compensation function is calculated, for example, when the ambient temperature exceeds the standard range, based on the temperature difference, a compensation coefficient is calculated according to a preset function. If the detected ambient temperature is 35 degrees Celsius and the standard temperature is 25 degrees Celsius, the compensation coefficient is calculated as 0.8. For example, when applying a power forward regulation instruction: the current ratio of the high-performance device is increased by 0.5 times the baseline value, for example, to 0.55 times, and then multiplied by the compensation coefficient to generate the final current limit.

[0149] As a specific example, the dynamic fusion process integrates a final current limit of 2000mA and a depth of discharge threshold of 80% for high-performance devices. When verifying the electrical safety operating boundary, if the current value is detected to be below the fuse protection threshold and the depth of discharge is within the battery cell tolerance, it is determined to meet safety requirements. For example, after generating a cycle life management strategy, health monitoring is simultaneously initiated: when the device health drops to a preset threshold, the current limit is automatically reduced proportionally and the depth of discharge threshold is updated, forming a closed-loop maintenance mechanism.

[0150] The above-mentioned method for generating a cycle life management strategy significantly improves the environmental adaptability of charging and discharging parameters through a real-time linkage mechanism between temperature compensation functions and environmental parameters. A dynamic correction strategy for the current ratio reference value based on the device performance level achieves the dual goals of tapping the potential of high-performance devices and protecting the safety of low-performance devices. The step-by-step configuration of the discharge depth threshold, combined with intelligent verification of the electrical safety boundary, ensures that the parameter fusion process is both flexible and reliable. The resulting cycle life management strategy effectively balances device performance output and life decay rate through a multi-dimensional dynamic calibration mechanism, providing full-cycle adaptive management capabilities for the energy storage system.

[0151] Reference Figure 7 The second embodiment of the present invention provides an intelligent management system for optimizing the cycle life of storage and charging equipment, including:

[0152] The data acquisition unit is used to obtain the historical operating data of the storage and charging equipment group, filter outliers according to the electrical parameter threshold and normalize them to form a standard set, extract performance indicators for clustering and grouping, and determine the performance level through group feature consistency verification;

[0153] The processing unit is used to match computing nodes by performance level to allocate real-time operating data, monitor node load in real time, and trigger a hierarchical adjustment mechanism if the load exceeds the threshold until the processing rate differences of all nodes are within the preset fluctuation range. When all nodes are in a balanced state, the specified degradation indicators of storage and charging equipment at each performance level are extracted, and a weighted fusion is generated according to the preset weights to generate a comprehensive performance value, and an evaluation matrix is ​​constructed to map the storage and charging equipment to the comprehensive performance value.

[0154] The management strategy generation unit is used to calculate the cycle life prediction value based on the comprehensive performance value in the evaluation matrix, compare it with the preset target value, generate a charge and discharge parameter optimization strategy based on the comparison result, set differentiated charge and discharge parameters in combination with the temperature compensation function, and generate a cycle life management strategy.

[0155] It should be noted that the intelligent management system for optimizing the cycle life of storage and charging equipment provided in an embodiment of the present invention is used to execute all the process steps of the intelligent management method for optimizing the cycle life of storage and charging equipment in the above embodiment. The working principles and beneficial effects of the two correspond one to one, so they will not be repeated here.

[0156] It should be noted that the system embodiment described above is merely illustrative, wherein the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the system embodiment provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive work.

[0157] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. An intelligent management method for optimizing the cycle life of storage and charging equipment, characterized in that: Executed by the server, including: Obtain historical operating data of the storage and charging equipment group, filter outliers based on electrical parameter thresholds and normalize them to form a standard set, extract performance indicators for clustering and grouping, and determine performance levels through group feature consistency verification; Match computing power nodes by performance level to distribute real-time operating data, monitor node load in real time, and trigger a hierarchical adjustment mechanism if the load exceeds the threshold until the processing rate differences of all nodes are within the preset fluctuation range; When all nodes are in a balanced state, the designated degradation indicators of the storage and charging equipment of each performance level are extracted, and a comprehensive performance value is generated by weighted fusion according to preset weights, and an evaluation matrix mapping the storage and charging equipment to the comprehensive performance value is constructed; Calculating a cycle life prediction value based on the comprehensive performance value in the evaluation matrix, comparing it with a preset target value, generating a charge and discharge parameter optimization strategy based on the comparison result, setting differentiated charge and discharge parameters in combination with a temperature compensation function, and generating a cycle life management strategy; The extraction of performance indicators for clustering and grouping, and determination of performance levels through group feature consistency verification, include: Extracting key performance indicator vectors from the standard set, calculating the indicator similarity between devices using an unsupervised clustering algorithm, aggregating devices with indicator similarity higher than a cutoff value into similar clusters, and separating devices with indicator similarity lower than the cutoff value into different clusters; Extract group common labels based on the indicator distribution characteristics of similar cluster devices, and generate initial level labels according to preset performance level mapping rules; Detecting the dispersion of device indicators within the same cluster, and in response to the dispersion exceeding a tolerance threshold, triggering a cluster re-division mechanism until the discrete permission condition is met, and outputting a device stratification result with a performance level label to determine the performance level of each device in the storage and charging device group; The method of matching computing nodes by performance level to allocate real-time operation data, monitoring node load in real time, and triggering a hierarchical adjustment mechanism if the load exceeds the threshold until the processing rate differences of all nodes are within a preset fluctuation range includes: Establish a mapping rule between performance level and node computing power, and allocate real-time operation data of different performance levels to corresponding computing power nodes; Real-time collection of resource occupancy and task processing rate of each node; In response to detecting that a resource occupancy rate of a node exceeds a dynamic load threshold, performing a data migration operation; After migration, the task processing rate differences between nodes are re-detected. If the difference exceeds the permitted fluctuation range, a proportional fine-tuning operation is performed. cyclically executing the data migration operation and the ratio fine-tuning operation until the task processing rate differences of all nodes are within a preset fluctuation range, and locking the current resource allocation status of all nodes; The step of generating a comprehensive performance value by weighted fusion according to preset weights and constructing an evaluation matrix mapping the storage and charging equipment to the comprehensive performance value includes: Perform dimension normalization transformation on each indicator of the specified recession indicator data set to form a standardized indicator vector; Perform weighted fusion calculation on the standardized index vector according to the preset weight coefficient to generate the comprehensive performance value of each storage and charging device; Based on the mapping relationship between equipment numbers and comprehensive performance values, a determinant matrix structure is constructed to form an evaluation matrix; The row index of the determinant-column matrix structure is the unique device number, and the column data is the corresponding comprehensive performance value.

2. The method according to claim 1, characterized in that The method of obtaining historical operating data of the storage and charging equipment group, filtering abnormal values ​​according to electrical parameter thresholds and normalizing the data to form a standard set includes: Collect historical operating data of the storage and charging equipment group through a preset data interface, wherein the operating data includes charge and discharge current, voltage fluctuation, temperature change and cycle number; Setting operating condition threshold boundaries for the charge and discharge current and the voltage fluctuation, and in response to detecting that the collected charge and discharge current historical data and / or voltage fluctuation historical data exceeds the operating condition threshold boundaries, marking the data as abnormal data and removing the data, thereby obtaining a cleaned data set; The temperature changes and cycle times in the cleaning data set are dimensionally normalized to form a standard set.

3. The method according to claim 1, characterized in that The establishing of a mapping rule between performance level and node computing power includes: establishing a mapping rule that a higher performance level corresponds to a higher node computing power; The performing of the data migration operation includes: transferring a preset proportion of data in the overloaded node to a non-overloaded node; The execution ratio fine-tuning operation includes: adjusting the data distribution ratio transferred to each of the non-overloaded nodes according to a preset range.

4. The method according to claim 1, wherein The method of extracting the designated degradation index of the storage and charging equipment of each performance level when all nodes are in a balanced state includes: In response to receiving a lock signal for the resource allocation status of all nodes, confirming that all nodes are in a load-balanced state; The storage device corresponding to the performance level classification index is accessed to the storage area of ​​the corresponding node to extract the specified decay indicator data set; The designated decay indicator data set includes capacity decay rate, internal resistance change slope, and charge and discharge efficiency.

5. The method according to claim 1, wherein The method of calculating a cycle life prediction value based on the comprehensive performance value in the evaluation matrix, comparing the cycle life prediction value with a preset target value, and generating a charge and discharge parameter optimization strategy based on the comparison result includes: Analyzing the mapping relationship between the equipment number and the comprehensive performance value in the evaluation matrix, and inputting the comprehensive performance value into the life prediction model to calculate the corresponding cycle life prediction value; The preset cycle life target value library is retrieved for comparison. If the cycle life prediction value is higher than the target value, a positive charge and discharge power adjustment instruction is generated. If the cycle life prediction value is not higher than the target value, a negative charge and discharge power adjustment instruction is generated.

6. The method according to claim 5, characterized in that The method of setting differentiated charge and discharge parameters in combination with the temperature compensation function to generate a cycle life management strategy includes: Collect ambient temperature data in real time and dynamically calculate the compensation coefficient through the temperature compensation function; Determine a current ratio reference value according to the equipment performance level, dynamically modify the current ratio reference value by applying a power positive / negative adjustment instruction, and generate a final current limit value in combination with the compensation coefficient; Configure discharge depth thresholds according to performance level; The final current limit is dynamically fused with the discharge depth threshold, and if it is verified that the fusion result complies with the electrical safety operation boundary, a cycle life management strategy is generated.

7. An intelligent management system for optimizing the cycle life of storage and charging equipment, characterized in that: Used to implement the method according to any one of claims 1 to 6, comprising: The data acquisition unit is used to obtain the historical operating data of the storage and charging equipment group, filter outliers according to the electrical parameter threshold and normalize them to form a standard set, extract performance indicators for clustering and grouping, and determine the performance level through group feature consistency verification; The processing unit is used to match computing nodes according to performance levels to allocate real-time operating data, monitor node load in real time, and trigger a hierarchical adjustment mechanism if the load exceeds a threshold until the processing rate differences of all nodes are within a preset fluctuation range; when all nodes are in a balanced state, extract the specified degradation indicators of storage and charging equipment of each performance level, weightedly fuse them according to preset weights to generate a comprehensive performance value, and construct an evaluation matrix mapping the storage and charging equipment to the comprehensive performance value; A management strategy generation unit is used to calculate a cycle life prediction value based on the comprehensive performance value in the evaluation matrix, compare it with a preset target value, generate a charge and discharge parameter optimization strategy based on the comparison result, set differentiated charge and discharge parameters in combination with a temperature compensation function, and generate a cycle life management strategy.

Images