A medium-large electrochemical energy storage power station battery health and residual life early warning method

Abstract

A medium and large-sized electrochemical energy storage power station battery health degree and residual life early warning method belongs to the field of battery health evaluation, and the method is: firstly, a battery health state grading system is established, threshold ranges of key parameters such as voltage, internal resistance and charge-discharge capacity are determined, a discrimination model is constructed by training historical data through various algorithms such as support vector machine, and the health grade is accurately evaluated. The health detection system monitors the data in real time, and the evaluation model is constructed through preprocessing and correlation analysis, and the health grade is judged and the report is generated. For sub-healthy battery units, the isolation priority is determined according to fuzzy comprehensive judgment, the deterioration trend is predicted, and the maintenance plan is made. During discharging, the battery is selected according to the health grade and the power is adjusted. The application utilizes machine learning algorithm to cluster analysis historical data to construct evaluation model to predict new data grade, and dynamically adjusts the operation strategy of the energy storage system based on the health grade, thereby improving the economy and reliability.

Description

Technical Field

[0001] This invention belongs to the field of battery health assessment technology, and specifically relates to a method for early warning of battery health and remaining life in medium and large-scale electrochemical energy storage power stations. Background Technology

[0002] Medium to large-scale electrochemical energy storage power stations typically consist of a large number of battery cells, whose operating status directly affects the performance, safety, and lifespan of the energy storage system. During long-term operation, battery cells inevitably experience performance degradation, such as capacity decay, increased internal resistance, and voltage fluctuations. These problems not only reduce the charging and discharging efficiency of the energy storage power station but may also lead to safety hazards, such as serious accidents like thermal runaway.

[0003] Currently, traditional battery management systems have limitations in monitoring and warning of battery health. Most systems can only monitor and judge simple parameters such as voltage and current, lacking the ability to comprehensively and accurately assess the battery's health status. They struggle to effectively identify and provide early warnings of sub-healthy battery conditions, making it difficult to take timely and targeted measures to prevent further deterioration. Furthermore, in terms of energy storage system operation strategies, they often fail to dynamically optimize and adjust based on the real-time health status of battery cells. This leads to an imbalance in the use of healthy, sub-healthy, and unhealthy battery cells, accelerating the overall performance degradation of the battery pack, shortening the lifespan of battery cells, and also affecting the economic efficiency and reliability of energy storage power stations in meeting grid dispatch requirements.

[0004] Based on the above analysis, the main technical problem solved by this invention is: how to classify the health status of batteries in energy storage power stations and make corresponding decisions according to the classification. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a method for early warning of battery health and remaining life of medium and large-scale electrochemical energy storage power stations. It establishes an accurate battery health status classification system for energy storage power stations. By comprehensively considering key parameters such as voltage, internal resistance and charge / discharge capacity, it accurately classifies battery cells into three levels: healthy, sub-healthy and unhealthy, and sets corresponding operating parameter standards and threshold ranges to improve the accuracy of battery health status assessment.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0007] A method for early warning of battery health and remaining life in medium-to-large-scale electrochemical energy storage power stations, comprising the following steps:

[0008] S1. Based on the pre-established battery health status classification system for energy storage power stations, the battery cells are divided into three levels: healthy, sub-healthy, and unhealthy. For each level, corresponding operating parameter standards are set, including the threshold ranges of voltage, internal resistance, and charge / discharge capacity.

[0009] S2. A health monitoring system is used to monitor the operating status information of each battery cell in real time, obtain real-time data on voltage, internal resistance, and charge / discharge capacity, and determine the health level of each battery cell through internal comparison and analysis.

[0010] S3. For the detected sub-healthy battery cells, the control system automatically generates a high-priority unloading command. By cutting off the output of the battery cell, the sub-healthy battery cell is isolated from the energy storage system to prevent the sub-healthy battery cell from deteriorating further.

[0011] S4. While unloading the sub-healthy battery units, dynamically adjust the working status of other healthy battery units. Based on the real-time health rating results, determine the unloading order of different battery units to ensure the overall output performance of the energy storage system.

[0012] S5. When the energy storage system needs to perform a discharge operation, it prioritizes the selection of battery cells with higher health levels to participate in the discharge. For battery cells that are assessed as sub-healthy or unhealthy, the control system gradually reduces their discharge power until the discharge is completely stopped.

[0013] S6. Machine learning algorithms are used to analyze the historical operating data of battery cells, and clustering algorithms are used to identify the characteristic patterns of different health levels. A health status assessment model is established to predict and classify the health level of newly collected battery cell data.

[0014] S7. The operation strategy of the energy storage system is dynamically adjusted based on the health level of the battery cells. While meeting the grid dispatch requirements, the service life of the battery cells is extended to the maximum extent, thereby improving the overall economy and reliability of the energy storage power station.

[0015] Preferably, the sub-step of step S1 is:

[0016] S1.1. Based on the pre-established battery health status classification system of the energy storage power station, obtain the real-time operating parameters of the battery cells. The real-time operating parameters include key parameters such as voltage, internal resistance and charge / discharge capacity.

[0017] S1.2. Compare the acquired real-time operating parameters of the battery cells with the preset health level threshold range to determine the health status of each battery cell;

[0018] S1.3 If the real-time operating parameters of the battery cell are all within the threshold range of the health level, then the health status of the battery cell is determined as the health level.

[0019] S1.4 If one or more of the real-time operating parameters of the battery cell exceed the threshold range of the health level, but are all within the threshold range of the sub-health level, then the health status of the battery cell is determined to be sub-health level.

[0020] S1.5 If one or more of the real-time operating parameters of the battery cell exceed the threshold range of the sub-health level, the health status of the battery cell shall be determined as unhealthy.

[0021] S1.6. Based on the determined health status of the battery cells, adopt the corresponding operating parameter standards to control and optimize the charging and discharging process of the battery cells to ensure the safe and stable operation of the energy storage power station.

[0022] S1.7. The historical operating data of the battery cells are trained using the support vector machine algorithm to obtain a discrimination model for health status classification. This model is used to predict and classify the health status of new battery cells, thereby improving the accuracy and efficiency of battery health status assessment.

[0023] The training process of the support vector machine algorithm on the historical operating data of the battery cells is as follows:

[0024] Let the training dataset be... ,in The voltage of the battery cells Internal resistance and charge / discharge capacity The eigenvectors, i.e. ;

[0025] These correspond to three categories: unhealthy, sub-healthy, and healthy.

[0026] Its objective function is: ,in , These are the weighted vector components corresponding to the characteristics of voltage, internal resistance, and charge / discharge capacity. It is the hyperplane intercept. G You can first set it to an experience value. It is a slack variable;

[0027] The constraints are: .

[0028] Preferably, the sub-step of step S2 is: S2.1, acquiring real-time operating status data of each group of battery cells, including key parameters such as voltage, internal resistance and charge / discharge capacity;

[0029] S2.2 Transmit the acquired real-time data to the health monitoring system and store it in the corresponding database;

[0030] S2.3 Preprocess the stored real-time data, including data cleaning and normalization operations, to improve data quality;

[0031] S2.4 Based on the preprocessed real-time data, a correlation analysis algorithm is used to calculate the correlation coefficient between each key parameter and determine the degree of correlation between the parameters;

[0032] S2.5. Based on the correlation coefficients of each key parameter, construct a battery health assessment model, and comprehensively consider the impact of each key parameter on the battery health status.

[0033] S2.6 Using the constructed battery health assessment model, the health status of each group of battery cells is assessed to obtain the health level of each battery cell.

[0034] S2.7. Based on the battery cell health level obtained from the assessment, generate a battery health report and send the report to relevant management personnel so as to keep abreast of the battery health status and formulate operation and maintenance strategies in advance.

[0035] Preferably, the sub-step of step S3 is: S3.1, according to the preset battery health status assessment model, monitor each battery cell in the energy storage system in real time and obtain key parameters reflecting the health status of each battery cell;

[0036] S3.2 If the health status parameters of a certain battery cell are detected to exceed the predetermined threshold range, the battery cell is determined to be in a sub-healthy state and is marked as a sub-healthy battery cell.

[0037] S3.3. Based on the severity of the sub-healthy battery cell's health status, a fuzzy comprehensive evaluation method is used to determine its isolation priority and generate corresponding isolation instructions.

[0038] The analytical method for fuzzy comprehensive evaluation is as follows:

[0039] Setting Factors Set ,in For the degree of voltage deviation , Rate of change of internal resistance , The charge / discharge capacity decay rate ;

[0040] Evaluation Set , Indicates low-priority isolation. Indicates medium priority isolation. Indicates high-priority isolation;

[0041] Determine the fuzzy relation matrix M: Indicates the degree of voltage deviation low-priority isolation membership degree, when hour, To build in this way ;

[0042] Determine the weight set The degree of voltage deviation was determined using the analytic hierarchy process (AHP). rate of change of internal resistance Charge and discharge capacity decay rate ;

[0043] Fuzzy comprehensive evaluation results , , ,according to The maximum membership principle is used to determine the isolation priority of sub-healthy battery cells;

[0044] S3.4. Send the isolation command to the control unit of the energy storage system. The control unit cuts off the output port of the sub-healthy battery unit according to the command, so as to realize the electrical isolation between the sub-healthy battery unit and the energy storage system.

[0045] S3.5. Through the battery management system, the isolated sub-healthy battery cells are removed from the logical structure of the energy storage system, and the working state of the remaining healthy battery cells is adjusted to ensure the overall performance of the energy storage system.

[0046] S3.6 For the isolated sub-healthy battery cells, a degradation prediction algorithm is used, combined with their historical monitoring data, to predict the deterioration trend of their health status and generate health status early warning information.

[0047] S3.7 Based on the health status warning information and the design life of the battery unit, determine the maintenance or replacement time of the sub-healthy battery unit, form a battery maintenance and replacement plan, and avoid its continued deterioration from affecting the system.

[0048] Preferably, the sub-step of step S4 is as follows:

[0049] S4.1 Obtain real-time health data of each battery cell in the energy storage system, analyze the data through machine learning algorithms, and obtain the health rating result of each battery cell;

[0050] S4.2. Based on the health rating results, a decision tree algorithm is used to determine the health status of each battery cell and mark battery cells with a health status below a preset threshold as sub-healthy.

[0051] S4.3 For battery cells marked as being in a sub-healthy state, use a control circuit to switch them off from the working circuit of the energy storage system and stop their charging and discharging operations.

[0052] S4.4 For battery cells in good health, a clustering algorithm is used to classify the battery cells into different health levels based on their health rating results.

[0053] S4.5 During the operation of the energy storage system, the working status of battery cells with different health levels is dynamically adjusted, and battery cells with higher health levels are given priority for charging and discharging operations.

[0054] S4.6 When it is necessary to unload sub-healthy battery cells, the optimal unloading sequence is determined by comprehensively considering the health level and state of charge of the battery cells through reinforcement learning algorithm.

[0055] S4.7. Based on the optimized unloading sequence, control the unloading process of the battery cells, and dynamically adjust the working status of healthy battery cells to ensure the overall output performance of the energy storage system is stable.

[0056] Preferably, the sub-step of step S5 is as follows:

[0057] S5.1 Collect the initial battery charge data and battery capacity data of all battery cells in the energy storage system;

[0058] S5.2. Use measurement algorithms to acquire the individual cell voltage data, battery internal resistance data, and operating temperature data of all battery cells, and synchronously transmit all the obtained data to the health model.

[0059] S5.3 Determine battery health level data;

[0060] Based on battery health level data, battery capacity data, and battery state of charge data;

[0061] The formula SOH=(Q) is used for calculation. actual / Q rated )100%, where SOH represents the battery health status, Q actual Q represents the maximum currently available capacity of the battery. rated This represents the battery's rated capacity.

[0062] Use the obtained Q actual The Q of each battery is obtained. rated The data is mapped using a recurrent neural network algorithm to a pre-established health status relationship mapping library. The obtained data is then input into an expert system to obtain the initial health level data of the battery cells.

[0063] S5.4 Obtain a sorted list of all battery cell health levels, and obtain the set of battery cell sequences with the highest health level by comparing the health levels;

[0064] S5.5. By sorting the values, determine the discharge order list of all battery cells in the battery cell sequence set with the highest health level value.

[0065] S5.6 According to the discharge sequence list, discharge is performed using the first discharge depth through the battery cell sequence set with the highest health level;

[0066] S5.7 Set a preset threshold V for each battery cell in the energy storage system. min ;

[0067] The monitoring module detects the actual voltage value of individual battery cells during discharge. If the voltage value of any individual battery cell is lower than the set threshold V, the monitoring module will detect the actual voltage value of the individual battery cells. min If so, the voltage of the battery cell is determined to be abnormal;

[0068] Collect ambient temperature sensor data and operating temperature sensor data, and use a Kalman filter fusion algorithm to obtain the fused value of ambient temperature and operating temperature;

[0069] If the fusion value is greater than the preset temperature value X, the temperature management module is activated to adjust the operating temperature of the energy storage system; if the fusion value is less than the preset value Y, the real-time information is stored in the database.

[0070] S5.8. Repeatedly determine the temperature status;

[0071] For battery cell voltage below threshold V min When this state occurs, the preset rule base is updated based on the charge / discharge rate, cycle count, and battery internal resistance information.

[0072] S5.9. Use logistic regression algorithm to obtain the health level update data of each battery cell, and store the update results in the operation log;

[0073] The weighting of the health level update data of different battery cells is obtained by the number of cycles and the charge / discharge rate. When the battery cell with the second-lowest health level is discharged, a command to reduce the discharge power will be obtained according to the weighting data of the battery cell.

[0074] S5.10: For battery cells deemed unhealthy, a stop discharge operation command is obtained.

[0075] Preferably, the sub-step of step S6 is as follows:

[0076] S6.1. Obtain historical operating data of the battery cells, preprocess the data, including data cleaning, feature extraction and data standardization, to obtain a dataset suitable for machine learning algorithm analysis.

[0077] S6.2. The K-means clustering algorithm is used to perform cluster analysis on the preprocessed historical running data. The data is divided into different categories according to the health status of the battery cells, and the feature patterns of each category are obtained.

[0078] S6.3. Based on the feature patterns of different health levels obtained from cluster analysis, construct a support vector machine classification model, use the feature patterns as the input of the model and the health level as the output of the model, and train the health status assessment model.

[0079] S6.4. Acquire newly collected battery cell operating data, perform the same preprocessing operations as historical data on the data, and obtain new data that conforms to the model input format;

[0080] S6.5 Input the preprocessed new data into the trained health status assessment model. The model predicts the health level of the battery cell based on the feature patterns of the input data.

[0081] S6.6 If the predicted health level is lower than the preset threshold, the battery cell is determined to be in an unhealthy state and needs to be maintained or replaced.

[0082] S6.7 If the predicted health level is higher than the threshold, the battery cell is determined to be in a healthy state and can continue to be used.

[0083] S6.8 Based on the health status prediction results of the battery cells, classify and manage the battery cells, formulate corresponding maintenance strategies, and ensure the reliable operation of the battery system; at the same time, add the newly collected data to the historical dataset, and regularly update and optimize the health status assessment model to improve the prediction accuracy of the model.

[0084] Preferably, in step S6.2, the K-means clustering algorithm process is as follows:

[0085] Given a dataset ,in To aggregate the dataset The categories correspond to three levels: healthy, sub-healthy, and unhealthy.

[0086] Random initialization Cluster centers , These represent three levels: healthy, sub-healthy, and unhealthy. ,in , , These are the estimated values ​​of voltage, internal resistance, and charge / discharge capacity for the cluster centers corresponding to the initial health levels;

[0087] Repeat the following steps until convergence:

[0088] For each data point Calculate it to each cluster center distance Euclidean distance ,Will Assign to the class of the nearest cluster center For each class Update cluster centers ,in Is class The number of data points in the data.

[0089] Preferably, the sub-step of step S7 is: S7.1, obtain the health level data of each battery cell in the energy storage power station, input the health level data into the pre-built battery health level assessment model, and obtain the health level score of each battery cell;

[0090] S7.2. Based on the health rating of each battery cell, a clustering algorithm is used to group the battery cells, and battery cells with similar health ratings are grouped together.

[0091] S7.3. For each group of battery cells, determine a scheduling priority parameter. The higher the health rating of the battery cell group, the greater its scheduling priority parameter.

[0092] S7.4 Obtain grid dispatch demand data, combine the dispatch priority parameters of each group of battery units, and solve the optimal energy storage system operation strategy through optimization algorithm. While meeting the grid dispatch demand, prioritize the dispatch of battery unit groups with high health levels.

[0093] S7.5 During the actual operation of the energy storage system, the health status data of each battery cell is acquired in real time. When it is determined that the health level of a certain battery cell has dropped significantly, the scheduling priority parameter of the group to which the battery cell belongs is dynamically adjusted to reduce the usage frequency of the battery cells in that group.

[0094] S7.6 Regularly assess the overall operation of the energy storage system, statistically analyze the health level change trend of each group of battery cells, determine the effect of the optimized operation strategy on extending the service life of the battery cells, and dynamically update the scheduling priority parameters accordingly.

[0095] S7.7 Continuously optimize the operation strategy of the energy storage system. Under the premise of ensuring that the grid dispatch needs are met, maximize the balance of the usage intensity of each group of battery units and slow down the degradation rate of the battery unit health level, thereby improving the overall economy and reliability of the energy storage power station.

[0096] The present invention can achieve the following beneficial effects:

[0097] 1. This invention establishes a precise battery health status classification system for energy storage power stations. By comprehensively considering key parameters such as voltage, internal resistance, and charge / discharge capacity, it accurately classifies battery cells into three levels: healthy, sub-healthy, and unhealthy, and sets corresponding operating parameter standards and threshold ranges to improve the accuracy of battery health status assessment.

[0098] 2. This invention develops a high-efficiency health detection system and algorithm model, such as support vector machine algorithm, fuzzy comprehensive evaluation method, clustering algorithm, recurrent neural network algorithm and logistic regression algorithm, etc. These algorithms are used to conduct in-depth analysis of the historical operation data and real-time monitoring data of battery cells, so as to realize real-time and accurate prediction and classification of battery health status, and timely detection of sub-healthy and unhealthy battery cells.

[0099] 3. This invention addresses sub-healthy battery cells by designing a reasonable isolation and processing mechanism. Based on the severity of their health status, it automatically generates high-priority unloading instructions to isolate them from the energy storage system. Furthermore, it employs a degradation prediction algorithm to forecast the deterioration trend of their health status and formulate effective maintenance or replacement plans to avoid negative impacts on the overall performance of the energy storage system and prevent safety accidents.

[0100] 3. During the operation of the energy storage system, especially during discharge, the present invention dynamically adjusts the working state according to the health level of the battery cells, prioritizes the selection of battery cells with higher health levels to participate in discharge, and reasonably controls the discharge power of sub-healthy and unhealthy battery cells. At the same time, the optimal battery cell unloading sequence is determined through reinforcement learning algorithm to ensure the overall output performance of the energy storage system is stable.

[0101] 4. This invention dynamically adjusts the operation strategy of the energy storage system based on the health level of the battery cells. Combined with the grid dispatch requirements, the optimal operation strategy is solved through optimization algorithms to balance the usage intensity of each group of battery cells, maximize the service life of the battery cells, and improve the overall economy and reliability of the energy storage power station. Attached Figure Description

[0102] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0103] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0104] Preferred solutions include Figure 1 As shown, a method for early warning of battery health and remaining life in medium-to-large-scale electrochemical energy storage power stations includes the following steps:

[0105] S1. Based on the pre-established battery health status classification system for energy storage power stations, battery cells are divided into three levels: healthy, sub-healthy, and unhealthy. Corresponding operating parameter standards are set for each level, including threshold ranges for key parameters such as voltage, internal resistance, and charge / discharge capacity; specifically:

[0106] S1.1. Based on the pre-established battery health status classification system for energy storage power stations, obtain the real-time operating parameters of the battery cells, including key parameters such as voltage, internal resistance, and charge / discharge capacity.

[0107] S1.2. Compare the acquired real-time operating parameters of the battery cells with the preset health level threshold range to determine the health status of each battery cell;

[0108] S1.3 If the real-time operating parameters of the battery cell are all within the threshold range of the health level, then the health status of the battery cell is determined as the health level.

[0109] S1.4 If one or more of the real-time operating parameters of the battery cell exceed the threshold range of the health level, but are all within the threshold range of the sub-health level, then the health status of the battery cell is determined to be sub-health level.

[0110] S1.5 If one or more of the real-time operating parameters of the battery cell exceed the threshold range of the sub-health level, the health status of the battery cell shall be determined as unhealthy.

[0111] S1.6. Based on the determined health status of the battery cells, adopt the corresponding operating parameter standards to control and optimize the charging and discharging process of the battery cells to ensure the safe and stable operation of the energy storage power station.

[0112] S1.7. The historical operating data of the battery cells are trained using the support vector machine algorithm to obtain a discrimination model for health status classification. This model is used to predict and classify the health status of new battery cells, thereby improving the accuracy and efficiency of battery health status assessment.

[0113] The training process of the support vector machine algorithm on the historical operating data of the battery cells is as follows:

[0114] Let the training dataset be... ,in The voltage of the battery cells Internal resistance and charge / discharge capacity Feature vectors with key parameters, i.e. ;

[0115] These correspond to three categories: unhealthy, sub-healthy, and healthy.

[0116] Its objective function is: ,in , These are the weighted vector components corresponding to the characteristics of voltage, internal resistance, and charge / discharge capacity. It is the hyperplane intercept. G You can first set it to an experience value, such as (This can be adjusted based on the actual classification results.) It is a slack variable.

[0117] The constraints are: ,

[0118] For example, for health-related categories ( ),but S2. A health monitoring system is used to monitor the operating status of each battery cell in real time, acquiring real-time data on key parameters such as voltage, internal resistance, and charge / discharge capacity. Through internal comparison and analysis, the health level of each battery cell is determined. Specifically:

[0119] S2.1 Obtain real-time operating status data for each battery cell, including key parameters such as voltage, internal resistance, and charge / discharge capacity;

[0120] S2.2 Transmit the acquired real-time data to the health monitoring system and store it in the corresponding database;

[0121] S2.3 Preprocess the stored real-time data, including data cleaning and normalization, to improve data quality;

[0122] S2.4 Based on the preprocessed real-time data, a correlation analysis algorithm is used to calculate the correlation coefficient between each key parameter and determine the degree of correlation between the parameters;

[0123] S2.5. Based on the correlation coefficients of each key parameter, construct a battery health assessment model, and comprehensively consider the impact of each parameter on the battery health status.

[0124] S2.6 Using the constructed battery health assessment model, the health status of each group of battery cells is assessed to obtain the health level of each battery cell.

[0125] S2.7. Based on the assessed battery cell health level, generate a battery health report and send it to relevant management personnel to promptly understand battery health status and formulate maintenance strategies in advance. S3. For detected sub-healthy battery cells, the control system automatically generates a high-priority unloading command, cutting off the battery cell's output to isolate it from the energy storage system and prevent further deterioration; specifically:

[0126] S3.1 Based on the preset battery health status assessment model, monitor each battery cell in the energy storage system in real time to obtain key parameters reflecting its health status.

[0127] S3.2 If the health status parameters of a certain battery cell are detected to exceed the predetermined threshold range, the battery cell is determined to be in a sub-healthy state and is marked as a sub-healthy battery cell.

[0128] S3.3. Based on the severity of the sub-healthy battery cell's health status, a fuzzy comprehensive evaluation method is used to determine its isolation priority and generate corresponding isolation instructions.

[0129] The analytical method for fuzzy comprehensive evaluation is as follows:

[0130] Setting Factors Set ,in For the degree of voltage deviation (The percentage of deviation relative to the health level voltage threshold range). Rate of change of internal resistance (The percentage change compared to the initial internal resistance). The charge / discharge capacity decay rate .

[0131] Evaluation Set , Indicates low-priority isolation. Indicates medium priority isolation. Indicates high-priority isolation;

[0132] Determine the fuzzy relation matrix M : Indicates the degree of voltage deviation low-priority isolation membership degree, when hour, To build in this way ;

[0133] Determine the weight set The degree of voltage deviation was determined using the analytic hierarchy process (AHP). rate of change of internal resistance Charge and discharge capacity decay rate ;

[0134] Fuzzy comprehensive evaluation results , , ,according to The maximum membership principle is used to determine the isolation priority of sub-healthy battery cells;

[0135] S3.4. Send the isolation command to the control unit of the energy storage system. The control unit cuts off the output port of the sub-healthy battery unit according to the command to achieve electrical isolation between it and the energy storage system.

[0136] S3.5. Through the battery management system, the isolated sub-healthy battery cells are removed from the logical structure of the energy storage system, and the working state of the remaining healthy battery cells is adjusted to ensure the overall performance of the energy storage system.

[0137] S3.6 For the isolated sub-healthy battery cells, a degradation prediction algorithm is used, combined with their historical monitoring data, to predict the deterioration trend of their health status and generate health status early warning information.

[0138] S3.7. Based on health status early warning information and the design life of the battery cells, determine the maintenance or replacement time for sub-healthy battery cells, and formulate a battery maintenance and replacement plan to prevent their continued deterioration from affecting the system. S4. While unloading sub-healthy battery cells, dynamically adjust the operating status of other healthy battery cells. Based on real-time health rating results, determine the unloading sequence of different battery cells to ensure the overall output performance of the energy storage system; specifically:

[0139] S4.1 Obtain real-time health data of each battery cell in the energy storage system, analyze the data through machine learning algorithms, and obtain the health rating result of each battery cell;

[0140] S4.2. Based on the health rating results, a decision tree algorithm is used to determine the health status of each battery cell and mark battery cells with a health status below a preset threshold as sub-healthy.

[0141] S4.3 For battery cells marked as being in a sub-healthy state, use a control circuit to switch them off from the working circuit of the energy storage system and stop their charging and discharging operations.

[0142] S4.4 For battery cells in good health, a clustering algorithm is used to classify the battery cells into different health levels based on their health rating results.

[0143] S4.5 During the operation of the energy storage system, the working status of battery cells with different health levels is dynamically adjusted, and battery cells with higher health levels are given priority for charging and discharging operations.

[0144] S4.6 When it is necessary to unload sub-healthy battery cells, the optimal unloading sequence is determined by comprehensively considering factors such as the health level and state of charge of the battery cells through reinforcement learning algorithm.

[0145] S4.7. Based on the optimized unloading sequence, control the unloading process of the battery cells, and dynamically adjust the working state of healthy battery cells to ensure the overall output performance of the energy storage system remains stable. S5. When the energy storage system needs to discharge, prioritize battery cells with higher health levels for discharge. For battery cells assessed as sub-healthy or unhealthy, the control system gradually reduces their discharge power until discharge is completely stopped; specifically:

[0146] S5.1 Collect the initial battery charge data and battery capacity data of all battery cells in the energy storage system;

[0147] S5.2. Use measurement algorithms to acquire the individual cell voltage data, battery internal resistance data, and operating temperature data of all battery cells, and synchronously transmit all the obtained data to the health model.

[0148] S5.3 Determine battery health level data;

[0149] Based on battery health level data, battery capacity data, and battery state of charge data;

[0150] The formula SOH=(Q) is used for calculation. actual / Q rated )100%, where SOH represents the battery health status, Q actual Q represents the maximum currently available capacity of the battery. rated This represents the battery's rated capacity.

[0151] Use the obtained Q actual The Q of each battery is obtained. rated The data is mapped using a recurrent neural network algorithm to a pre-established health status relationship mapping library. The obtained data is then input into an expert system to obtain the initial health level data of the battery cells.

[0152] The monitoring module detects the actual voltage value of individual battery cells during discharge. If the voltage value of any individual battery cell is lower than the set threshold V, the monitoring module will detect the actual voltage value of the individual battery cells. min If so, the voltage of the battery cell is determined to be abnormal;

[0153] Collect ambient temperature sensor data and operating temperature sensor data, and use a Kalman filter fusion algorithm to obtain the fused value of ambient temperature and operating temperature;

[0154] If the fusion value is greater than the preset temperature value X, the temperature management module is activated to adjust the operating temperature of the energy storage system; if the fusion value is less than the preset value Y, the real-time information is stored in the database.

[0155] S5.8. Repeatedly determine the temperature status;

[0156] For battery cell voltage below threshold V min When this state occurs, the preset rule base is updated based on the charge / discharge rate, cycle count, and battery internal resistance information.

[0157] S5.9. Use logistic regression algorithm to obtain the health level update data of each battery cell, and store the update results in the operation log;

[0158] The weighting of the health level update data of different battery cells is obtained by the number of cycles and the charge / discharge rate. When the battery cell with the second-lowest health level is discharged, a command to reduce the discharge power will be obtained according to the weighting data of the battery cell.

[0159] S5.10. For battery cells deemed unhealthy, a stop-discharge operation command is obtained;

[0160] S6. Machine learning algorithms are used to analyze historical operating data of battery cells. Clustering algorithms are used to identify characteristic patterns of different health levels, and a health status assessment model is established to predict and classify the health level of newly collected battery cell data; specifically:

[0161] S6.1. Obtain historical operating data of the battery cells, preprocess the data, including data cleaning, feature extraction and data standardization, to obtain a dataset suitable for machine learning algorithm analysis;

[0162] S6.2. The K-means clustering algorithm is used to perform cluster analysis on the preprocessed historical running data. The data is divided into different categories according to the health status of the battery cells, and the feature patterns of each category are obtained.

[0163] The K-means clustering algorithm process is as follows:

[0164] Given a dataset ,in (These are the voltage, internal resistance, and charge / discharge capacity of the battery cells, respectively). The data needs to be aggregated into... Categories (corresponding to three levels: healthy, sub-healthy, and unhealthy).

[0165] Random initialization Cluster centers ,For example ,in , , These are the estimated values ​​of voltage, internal resistance, and charge / discharge capacity for the corresponding health level cluster centers (obtainable from the mean value of a large amount of healthy battery cell data), respectively. Similarly, initialize... (corresponding to sub-health) and (Corresponds to unhealthy).

[0166] Repeat the following steps until convergence:

[0167] For each data point Calculate it to each cluster center distance For example, Euclidean distance ,Will Assign to the class of the nearest cluster center - For each class Update cluster centers ,in Is class The number of data points in the data.

[0168] S6.3. Based on the feature patterns of different health levels obtained from cluster analysis, construct a support vector machine classification model, use the feature patterns as the input of the model and the health level as the output of the model, and train the health status assessment model.

[0169] S6.4. Acquire newly collected battery cell operating data, perform the same preprocessing operations as historical data on the data, and obtain new data that conforms to the model input format;

[0170] S6.5 Input the preprocessed new data into the trained health status assessment model. The model predicts the health level of the battery cell based on the feature patterns of the input data.

[0171] S6.6 If the predicted health level is lower than the preset threshold, the battery cell is determined to be in an unhealthy state and needs to be maintained or replaced.

[0172] S6.7 If the predicted health level is higher than the threshold, the battery cell is determined to be in a healthy state and can continue to be used.

[0173] S6.8 Based on the health status prediction results of the battery cells, classify and manage the battery cells, formulate corresponding maintenance strategies, and ensure the reliable operation of the battery system.

[0174] Simultaneously, newly collected data is added to the historical dataset, and the health status assessment model is regularly updated and optimized to improve the model's prediction accuracy. S7. The operation strategy of the energy storage system is dynamically adjusted based on the health level of the battery cells, maximizing the lifespan of the battery cells while meeting grid dispatch requirements, thereby improving the overall economy and reliability of the energy storage power station. Specifically:

[0175] S7.1 Obtain the health level data of each battery cell in the energy storage power station, input the health level data into the pre-built battery health level assessment model, and obtain the health level score of each battery cell;

[0176] S7.2. Based on the health rating of each battery cell, a clustering algorithm is used to group the battery cells, and battery cells with similar health ratings are grouped together.

[0177] S7.3. For each group of battery cells, determine a scheduling priority parameter. The higher the health rating of the battery cell group, the greater its scheduling priority parameter.

[0178] S7.4 Obtain grid dispatch demand data, combine the dispatch priority parameters of each group of battery units, and solve the optimal energy storage system operation strategy through optimization algorithm. While meeting the grid dispatch demand, prioritize the dispatch of battery unit groups with high health levels.

[0179] S7.5 During the actual operation of the energy storage system, the health status data of each battery cell is acquired in real time. When it is determined that the health level of a certain battery cell has dropped significantly, the scheduling priority parameter of the group to which the battery cell belongs is dynamically adjusted to reduce the usage frequency of the battery cells in that group.

[0180] S7.6 Regularly assess the overall operation of the energy storage system, statistically analyze the health level change trend of each group of battery cells, determine the effect of the optimized operation strategy on extending the service life of the battery cells, and dynamically update the scheduling priority parameters accordingly.

[0181] S7.7 Continuously optimize the operation strategy of the energy storage system. Under the premise of ensuring that the grid dispatch needs are met, maximize the balance of the usage intensity of each group of battery units and slow down the degradation rate of the battery unit health level, thereby improving the overall economy and reliability of the energy storage power station.

[0182] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A method for early warning of battery health and remaining life in medium-to-large-scale electrochemical energy storage power stations, characterized in that... Includes the following steps: S1. Based on the pre-established battery health status classification system for energy storage power stations, the battery cells are divided into three levels: healthy, sub-healthy, and unhealthy. For each level, corresponding operating parameter standards are set, including the threshold ranges of voltage, internal resistance, and charge / discharge capacity. S2. A health monitoring system is used to monitor the operating status information of each battery cell in real time, obtain real-time data on voltage, internal resistance, and charge / discharge capacity, and determine the health level of each battery cell through internal comparison and analysis. S3. For the detected sub-healthy battery cells, the control system automatically generates a high-priority unloading command. By cutting off the output of the battery cell, the sub-healthy battery cell is isolated from the energy storage system to prevent the sub-healthy battery cell from deteriorating further. S4. While unloading the sub-healthy battery units, dynamically adjust the working status of other healthy battery units. Based on the real-time health rating results, determine the unloading order of different battery units to ensure the overall output performance of the energy storage system. S5. When the energy storage system needs to perform a discharge operation, it prioritizes the selection of battery cells with higher health levels to participate in the discharge. For battery cells that are assessed as sub-healthy or unhealthy, the control system gradually reduces their discharge power until the discharge is completely stopped. S6. Machine learning algorithms are used to analyze the historical operating data of battery cells, and clustering algorithms are used to identify the characteristic patterns of different health levels. A health status assessment model is established to predict and classify the health level of newly collected battery cell data. The sub-steps of step S6 are as follows: S6.

1. Obtain historical operating data of the battery cells, preprocess the data, including data cleaning, feature extraction and data standardization, to obtain a dataset suitable for machine learning algorithm analysis. S6.

2. The K-means clustering algorithm is used to perform cluster analysis on the preprocessed historical running data. The data is divided into different categories according to the health status of the battery cells, and the feature patterns of each category are obtained. S6.

3. Based on the feature patterns of different health levels obtained from cluster analysis, construct a support vector machine classification model, use the feature patterns as the input of the model and the health level as the output of the model, and train the health status assessment model. S6.

4. Acquire newly collected battery cell operating data, perform the same preprocessing operations as historical data on the data, and obtain new data that conforms to the model input format; S6.5 Input the preprocessed new data into the trained health status assessment model. The model predicts the health level of the battery cell based on the feature patterns of the input data. S6.6 If the predicted health level is lower than the preset threshold, the battery cell is determined to be in an unhealthy state and needs to be maintained or replaced. S6.7 If the predicted health level is higher than the threshold, the battery cell is determined to be in a healthy state and can continue to be used. S6.8 Based on the health status prediction results of the battery cells, classify and manage the battery cells, formulate corresponding maintenance strategies, and ensure the reliable operation of the battery system; at the same time, add the newly collected data to the historical dataset, and regularly update and optimize the health status assessment model to improve the prediction accuracy of the model. S7. The operation strategy of the energy storage system is dynamically adjusted based on the health level of the battery cells. While meeting the grid dispatch requirements, the service life of the battery cells is extended to the maximum extent, thereby improving the overall economy and reliability of the energy storage power station.

2. The method for early warning of battery health and remaining life in a medium-to-large-scale electrochemical energy storage power station according to claim 1, characterized in that: The sub-steps of step S1 are: S1.

1. Based on the pre-established battery health status classification system of the energy storage power station, obtain the real-time operating parameters of the battery cells. The real-time operating parameters include key parameters such as voltage, internal resistance and charge / discharge capacity. S1.

2. Compare the acquired real-time operating parameters of the battery cells with the preset health level threshold range to determine the health status of each battery cell; S1.3 If the real-time operating parameters of the battery cell are all within the threshold range of the health level, then the health status of the battery cell is determined as the health level. S1.4 If one or more of the real-time operating parameters of the battery cell exceed the threshold range of the health level, but are all within the threshold range of the sub-health level, then the health status of the battery cell is determined to be sub-health level. S1.5 If one or more of the real-time operating parameters of the battery cell exceed the threshold range of the sub-health level, the health status of the battery cell shall be determined as unhealthy. S1.

6. Based on the determined health status of the battery cells, adopt the corresponding operating parameter standards to control and optimize the charging and discharging process of the battery cells to ensure the safe and stable operation of the energy storage power station. S1.

7. The historical operating data of the battery cells are trained using the support vector machine algorithm to obtain a discrimination model for health status classification. This model is used to predict and classify the health status of new battery cells, thereby improving the accuracy and efficiency of battery health status assessment. The training process of the support vector machine algorithm on the historical operating data of the battery cells is as follows: Let the training dataset be... ,in The voltage of the battery cells Internal resistance and charge / discharge capacity The eigenvectors, i.e. ; These correspond to three categories: unhealthy, sub-healthy, and healthy. Its objective function is: ,in , These are the weighted vector components corresponding to the characteristics of voltage, internal resistance, and charge / discharge capacity. It is the hyperplane intercept. G For an experience value, It is a slack variable; The constraints are: .

3. The method for early warning of battery health and remaining life in a medium-to-large-scale electrochemical energy storage power station according to claim 1, characterized in that: The sub-steps of step S2 are: S2.1 Obtain real-time operating status data for each battery cell group, including key parameters such as voltage, internal resistance, and charge / discharge capacity; S2.2 Transmit the acquired real-time data to the health monitoring system and store it in the corresponding database; S2.3 Preprocess the stored real-time data, including data cleaning and normalization operations, to improve data quality; S2.4 Based on the preprocessed real-time data, a correlation analysis algorithm is used to calculate the correlation coefficient between each key parameter and determine the degree of correlation between the parameters; S2.

5. Based on the correlation coefficients of each key parameter, construct a battery health assessment model, and comprehensively consider the impact of each key parameter on the battery health status. S2.6 Using the constructed battery health assessment model, the health status of each group of battery cells is assessed to obtain the health level of each battery cell. S2.

7. Based on the battery cell health level obtained from the assessment, generate a battery health report and send the report to relevant management personnel so as to keep abreast of the battery health status and formulate operation and maintenance strategies in advance.

4. The method for early warning of battery health and remaining life in a medium-to-large-scale electrochemical energy storage power station according to claim 1, characterized in that: The sub-steps of step S3 are as follows: S3.1 Based on the preset battery health status assessment model, monitor each battery cell in the energy storage system in real time and obtain key parameters reflecting the health status of each battery cell. S3.2 If the health status parameters of a certain battery cell are detected to exceed the predetermined threshold range, the battery cell is determined to be in a sub-healthy state and is marked as a sub-healthy battery cell. S3.

3. Based on the severity of the sub-healthy battery cell's health status, a fuzzy comprehensive evaluation method is used to determine its isolation priority and generate corresponding isolation instructions. The analytical method for fuzzy comprehensive evaluation is as follows: Setting Factors Set ,in Voltage deviation degree , Rate of change of internal resistance , The charge / discharge capacity decay rate ; Evaluation Set , Indicates low-priority isolation. Indicates medium priority isolation. Indicates high-priority isolation; Determine the fuzzy relation matrix M : Indicates the degree of voltage deviation low-priority isolation membership degree, when hour, To build in this way ; Determine the weight set The degree of voltage deviation was determined using the analytic hierarchy process (AHP). rate of change of internal resistance Charge and discharge capacity decay rate ; Fuzzy comprehensive evaluation results , , ,according to The maximum membership principle is used to determine the isolation priority of sub-healthy battery cells; S3.

4. Send the isolation command to the control unit of the energy storage system. The control unit cuts off the output port of the sub-healthy battery unit according to the command, so as to realize the electrical isolation between the sub-healthy battery unit and the energy storage system. S3.

5. Through the battery management system, the isolated sub-healthy battery cells are removed from the logical structure of the energy storage system, and the working state of the remaining healthy battery cells is adjusted to ensure the overall performance of the energy storage system. S3.6 For the isolated sub-healthy battery cells, a degradation prediction algorithm is used, combined with their historical monitoring data, to predict the deterioration trend of their health status and generate health status early warning information. S3.7 Based on the health status warning information and the design life of the battery unit, determine the maintenance or replacement time of the sub-healthy battery unit, form a battery maintenance and replacement plan, and avoid its continued deterioration from affecting the system.

5. The method for early warning of battery health and remaining life in a medium-to-large-scale electrochemical energy storage power station according to claim 1, characterized in that: The sub-steps of step S4 are as follows: S4.1 Obtain real-time health data of each battery cell in the energy storage system, analyze the data through machine learning algorithms, and obtain the health rating result of each battery cell; S4.

2. Based on the health rating results, a decision tree algorithm is used to determine the health status of each battery cell, and battery cells with a health rating below a preset threshold are marked as sub-healthy. S4.3 For battery cells marked as being in a sub-healthy state, the control circuit switches off the battery cells from the working circuit of the energy storage system to stop their charging and discharging operations. S4.4 For battery cells in good health, a clustering algorithm is used to classify the battery cells into different health levels based on their health rating results. S4.5 During the operation of the energy storage system, the working status of battery cells with different health levels is dynamically adjusted, and battery cells with higher health levels are given priority for charging and discharging operations. S4.6 When it is necessary to unload sub-healthy battery cells, the optimal unloading sequence is determined by comprehensively considering the health level and state of charge of the battery cells through reinforcement learning algorithm. S4.

7. Based on the optimized unloading sequence, control the unloading process of the battery cells, and dynamically adjust the working status of healthy battery cells to ensure the overall output performance of the energy storage system is stable.

6. The method for early warning of battery health and remaining life in a medium-to-large-scale electrochemical energy storage power station according to claim 1, characterized in that: The sub-steps of step S5 are as follows: S5.1 Collect the initial battery charge data and battery capacity data of all battery cells in the energy storage system; S5.

2. Use measurement algorithms to acquire the individual cell voltage data, battery internal resistance data, and operating temperature data of all battery cells, and synchronously transmit all the obtained data to the health model. S5.3 Determine battery health level data; Based on battery health level data, battery capacity data, and battery state of charge data; The formula SOH=(Q) is used for calculation. actual / Q rated )100%, where SOH represents the battery health status, Q actual Q represents the maximum available capacity of the battery at present. rated This represents the battery's rated capacity. Use the obtained Q actual The Q of each battery is obtained. rated The data is mapped using a recurrent neural network algorithm to a pre-established health status relationship mapping library. The obtained data is then input into an expert system to obtain the initial health level data of the battery cells. S5.4 Obtain a sorted list of all battery cell health levels, and obtain the set of battery cell sequences with the highest health level by comparing the health levels; S5.

5. By sorting the values, determine the discharge order list of all battery cells in the battery cell sequence set with the highest health level value. S5.6 According to the discharge sequence list, discharge is performed using the first discharge depth through the battery cell sequence set with the highest health level; S5.7 Set a preset threshold V for each battery cell in the energy storage system. min ; The monitoring module detects the actual voltage value of individual battery cells during discharge. If the voltage value of any individual battery cell is lower than the set threshold V, the monitoring module will detect the actual voltage value of the individual battery cells. min If so, the voltage of the battery cell is determined to be abnormal; Collect ambient temperature sensor data and operating temperature sensor data, and use a Kalman filter fusion algorithm to obtain the fused value of ambient temperature and operating temperature; If the fusion value is greater than the preset temperature value X, the temperature management module is activated to adjust the operating temperature of the energy storage system; if the fusion value is less than the preset value Y, the real-time information is stored in the database. S5.8, Repeatedly determine the temperature status; For battery cell voltage below threshold V min When this state occurs, the preset rule base is updated based on the charge / discharge rate, cycle count, and battery internal resistance information. S5.

9. Use logistic regression algorithm to obtain the health level update data of each battery cell, and store the update results in the operation log; The weighting of the health level update data of different battery cells is obtained by the number of cycles and the charge / discharge rate. When the battery cell with the second-lowest health level is discharged, a command to reduce the discharge power will be obtained according to the weighting data of the battery cell. S5.10: For battery cells deemed unhealthy, a stop discharge operation command is obtained.

7. The method for early warning of battery health and remaining life in a medium-to-large-scale electrochemical energy storage power station according to claim 1, characterized in that: The sub-steps of step S7 are as follows: S7.1 Obtain the health level data of each battery cell in the energy storage power station, input the health level data into the pre-built battery health level assessment model, and obtain the health level score of each battery cell; S7.

2. Based on the health rating of each battery cell, a clustering algorithm is used to group the battery cells, and battery cells with similar health ratings are grouped together. S7.

3. For each group of battery cells, determine a scheduling priority parameter. The higher the health rating of the battery cell group, the greater its scheduling priority parameter. S7.4 Obtain grid dispatch demand data, combine the dispatch priority parameters of each group of battery units, and solve the optimal energy storage system operation strategy through optimization algorithm. While meeting the grid dispatch demand, prioritize the dispatch of battery unit groups with high health levels. S7.5 During the actual operation of the energy storage system, the health status data of each battery cell is acquired in real time. When it is determined that the health level of a certain battery cell has dropped significantly, the scheduling priority parameter of the group to which the battery cell belongs is dynamically adjusted to reduce the usage frequency of the battery cells in that group. S7.6 Regularly assess the overall operation of the energy storage system, statistically analyze the health level change trend of each group of battery cells, determine the effect of the optimized operation strategy on extending the service life of the battery cells, and dynamically update the scheduling priority parameters accordingly. S7.7 Continuously optimize the operation strategy of the energy storage system. Under the premise of ensuring that the grid dispatch needs are met, maximize the balance of the usage intensity of each group of battery units and slow down the degradation rate of the battery unit health level, thereby improving the overall economy and reliability of the energy storage power station.

Images