Forced equalizing charging system based on lead-acid storage battery pack of electric vehicle

Abstract

The invention discloses a forced equalizing charging system based on an electric vehicle lead-acid storage battery pack, which relates to the technical field of power supply charging, and comprises a monomer health assessment module for acquiring an initial state of charge (SOC) parameter of a monomer battery from a dynamic operation parameter, correcting a vector influenced by the SOC parameter in an original health feature vector, and determining the state of charge (SOC) of the monomer battery; the current SOH evaluation value of the single battery is calculated based on the corrected health feature vector F, trend fusion is carried out on the current SOH evaluation value and historical SOHp data, and a comprehensive SOHS evaluation result is generated; a battery grouping management module; and a charging mode customization module. According to the method, the comprehensive SOHS of each battery is calculated and divided into five classes according to the preset standard, each class corresponds to different aging degrees and risk levels, a complete life cycle portrait is formed, the system supports early recognition and key monitoring of each single battery, and the service life of each single battery is greatly prolonged. And a decision basis can be provided for controlling the differentiation of subsequent single batteries.

Description

Technical Field

[0001] This invention relates to the field of power supply and charging technology, and in particular to a forced equalization charging system based on lead-acid battery packs for electric vehicles. Background Technology

[0002] For example, patent CN107147197A, entitled "A Flexible Follow-up Intelligent Charging Method and Charging Device," includes, in chronological order, a pre-charging stage, a first constant current charging stage, a constant voltage and constant current pulse charging stage, a first constant voltage charging stage, a second constant current charging stage, a second constant voltage charging stage, and an equalization charging stage. This flexible follow-up intelligent charging method and charging device can monitor the battery status in real time and actively change the charging voltage and current parameters according to the battery's power receiving capacity at each stage, automatically implementing the entire charging process. It solves the common problems of undercharging, overcharging, excessive temperature rise, and excessive gas evolution in secondary batteries, helping to charge more power during the charging process and extending the battery's single-use time and actual service life.

[0003] This application addresses the systematic shift in health characteristics such as gas evolution voltage and response time of lead-acid battery packs for electric vehicles caused by polarization effects during charging at different initial SOCs. Furthermore, SOH estimation often relies on linear models or single parameters, making it difficult to understand the aging behavior of batteries with multiple coexisting mechanisms. It also relies on judgments based on single data points without incorporating historical trends, making it susceptible to interference that can lead to abrupt changes in evaluation results and fail to accurately reflect long-term degradation patterns. Additionally, the charging systems in the aforementioned applications employ a fixed constant current and constant voltage strategy, failing to differentiate control based on the individual health status of batteries. This results in excessive equalization and energy waste in healthy batteries, while degraded batteries face a high risk of overcharging and safety hazards. Therefore, this application provides a forced equalization charging system for lead-acid battery packs for electric vehicles to meet these requirements. Summary of the Invention

[0004] The purpose of this application is to provide a forced equalization charging system based on lead-acid battery packs for electric vehicles, which can effectively solve the problems mentioned in the background art.

[0005] To achieve the above objectives, this application provides the following technical solution: a forced equalization charging system based on a lead-acid battery pack for electric vehicles, comprising:

[0006] The data acquisition module collects the dynamic operating parameters of each individual battery cell in real time.

[0007] The gas evolution initiation detection module identifies the starting point of battery gas evolution behavior based on the collected dynamic operating parameters, and detects and outputs the initiation event parameter vector of battery gas evolution behavior.

[0008] The gas evolution behavior analysis module extracts features and performs physical transformation on the parameter vector of the gas evolution initiation event to construct an original health feature vector that reflects the battery aging mechanism.

[0009] The single-cell health assessment module obtains the initial state of charge (SOC) parameters of individual cells from dynamic operating parameters, corrects the vectors in the original health feature vector affected by the SOC parameters, calculates the current SOH assessment value of the individual cell based on the corrected health feature vector F, and combines the current SOH assessment value with the historical SOH value. p Data is trend-following fusion to generate a comprehensive SOH. S Evaluation results;

[0010] The battery group management module, based on SOH... S The evaluation results are continuously updated to categorize each individual battery cell based on its health profile.

[0011] The charging mode customization module allows you to configure exclusive charging parameters based on the battery type.

[0012] The initial event parameter vector includes the gas evolution initiation time S, the gas evolution initiation voltage V, the gas evolution initiation temperature T, and the current internal resistance R.

[0013] The original health feature vector includes the gas evolution voltage offset ΔV, the constant pressure response delay ΔS, and the internal resistance growth rate R. T ;

[0014] The gas evolution voltage offset ΔV is calculated by comparing the difference between the current gas evolution initiation voltage V of the battery and the standard voltage of a healthy battery at the start of charging. The standard voltage refers to the typical voltage value exhibited by a healthy battery at the beginning of charging.

[0015] The constant voltage response delay ΔS is determined by comparing the gas evolution start time S when the current battery reaches the constant voltage charging state with the standard time point when a healthy battery starts gas evolution under the same conditions. The standard response time refers to the time required for a healthy battery to start constant voltage charging.

[0016] The internal resistance growth rate R T It reflects the trend of battery internal resistance changing with temperature. It is derived by comparing the internal resistance after temperature compensation with the internal resistance of a single cell at the standard reference temperature. The internal resistance after temperature compensation is the percentage change in battery resistance for every 1°C increase, and is used to reflect the changes in battery performance under actual use conditions.

[0017] Among them, the vectors affected by the SOC parameter in the original health feature vector include the gas evolution voltage offset ΔV and the constant pressure response delay ΔS;

[0018] The correction calculation for the gas evolution voltage offset ΔV is as follows:

[0019] ΔVsoc=ΔV-ΔV1+ΔV2;

[0020] ΔV1 is the expected value of a healthy battery under the current SOC;

[0021] ΔV2 is the gas evolution voltage offset ΔV of a healthy battery when SOC = 100%;

[0022] ΔVsos is the corrected gas evolution voltage offset.

[0023] The correction calculation for the constant pressure response delay ΔS is as follows:

[0024] ΔSsoc=ΔS-ΔS1+ΔS2;

[0025] ΔS1 is the expected constant voltage response delay time of a healthy battery at the current SOC.

[0026] ΔS2 is the expected constant voltage response delay time of a healthy battery when SOC = 100%;

[0027] The constant-pressure response delay time after ΔSsoc correction.

[0028] The calculation method for the current SOH evaluation value of a single cell based on the corrected health feature vector F is as follows:

[0029]

[0030] C i This represents the center vector of the i-th hidden layer neuron;

[0031] ||FC i || represents the Euclidean distance between the current battery's health feature vector F and the i-th ci, while exp is the similarity function;

[0032] W i The output weight of the i-th neuron;

[0033] N represents the number of hidden layer neurons.

[0034] Among them, based on the current SOH assessment value and combined with historical SOH values p Data is trend-following and a comprehensive SOH is calculated. S The evaluation results are as follows:

[0035] SOH S =α·SOH + (1-α)·SOH p ;

[0036] SOH pFor the previous comprehensive SOH S The evaluation results show that the SOH (State of Health) is highest when a single cell is being charged for the first time. p =SOH;

[0037] α is the smoothing coefficient.

[0038] Among them, based on the comprehensive SOH S The individual cell health profile obtained from the evaluation results includes basic information about the battery, current state of health (SOH), and other relevant data. S Evaluation results and historical trend data are used to comprehensively record and continuously track the health status of the battery.

[0039] The battery category is based on the comprehensive SOH in the health record. S The assessment results and multidimensional health characteristics were divided into five levels, namely, novel (SOH) S ≥98%), healthy (90% ≤ SOH) S <98%), Warning (80%≤SOH) S <90%), deterioration (70%≤SOH S <80%) and scrap (SOH) S <70%), used for fine-grained classification of the health status of individual cells.

[0040] Among them, dedicated charging parameters are configured based on battery type. The new and healthy types are planned to force equalization charging in the later stage of charging, which is used to correct the differences in battery voltage, SOC and current internal resistance R, and avoid energy loss caused by excessive equalization.

[0041] The warning type is planned to perform active balancing in real time during charging, and to forcibly start the balancing circuit in the middle of charging to actively suppress voltage deviation and prevent overvoltage of a single cell.

[0042] Degradation types are divided into forced active balancing, which means that the balancing circuit is forced to start balancing at the beginning of charging, giving priority to energy discharge of high-voltage individual cells to prevent them from entering dangerous areas, while avoiding overcharging of degraded cells.

[0043] Scrapping-type batteries are those whose capacity has severely degraded and no longer have normal usability. Normal charging should be prohibited, and only maintenance charging at very low power is allowed under controlled conditions. The temperature and voltage changes of individual cells must be monitored in real time.

[0044] In summary, the technical effects and advantages of this invention are as follows:

[0045] 1. The present invention has a reasonable structure. The health feature vector F after correction by the initial state of charge (SOC) parameter is an equivalent representation of all measured values ​​under the same state. It effectively eliminates the evaluation bias caused by different charging start conditions, and makes the health features comparable across cycles and environments. Especially in complex application scenarios such as frequent charging of electric vehicles and multi-stage charging and discharging of energy storage systems, this mechanism ensures the long-term stable output of SOH evaluation results, avoids the problem of misjudgment of different results for the same battery, and improves the intelligence level of the battery management system.

[0046] 2. In this invention, the current SOH evaluation value is achieved through the multi-aging mode matching mechanism of the RBF neural network, realizing the refined identification and quantitative evaluation of the complex aging path of the battery, and the center vector C of the hidden layer neurons is used to... i Defined as the feature center of a specific aging mode, the system inputs the corrected health feature vector F into the network during the evaluation process, calculates its Euclidean distance with each mode center, and transforms it into similarity weights through a Gaussian kernel function. Finally, the current SOH value is output by weighting. This mechanism not only has strong nonlinear fitting ability and significantly improves the physical interpretability and discrimination accuracy of the model, but also, compared with the black box model, this method can effectively distinguish different aging mechanisms and avoid misjudgment.

[0047] 3. This invention achieves high-precision and robust battery health status estimation through a three-level progressive evaluation architecture of "feature correction—pattern recognition—trend fusion". First, health features affected by SOC interference are dynamically corrected to ensure the consistency and comparability of evaluation results under different operating conditions. Second, a multi-aging pattern matching mechanism based on RBF neural network is adopted, using the hidden layer center vector to represent typical aging paths, and comparing the current SOH evaluation value with historical SOH values. p By performing trend fusion, a more stable comprehensive SOH is generated. S The evaluation results, by suppressing the influence of transient noise and outliers, not only improved the accuracy and continuity of SOH estimation; secondly, after each charge, the current SOH evaluation value was compared with the previous comprehensive SOH. p Weighted fusion is performed, where α is a smoothing coefficient used to adjust the contribution weights of new and old data. When the battery is charged for the first time, SOH p Initialized to the current SOH value, and then updated sequentially to give historical data a continuous influence, effectively suppressing transient interference and improving the overall SOH. S The assessment results show a smooth and continuous trend, which more accurately reflects the long-term degradation pattern of the battery.

[0048] 4. This invention calculates the overall SOH of each battery cell. SBased on preset standards, the batteries are divided into five categories, each corresponding to different degrees of aging and risk levels. The classification results are updated in real time to the health record of each battery. The health record includes basic information and historical SOH. S Multidimensional data, including curves, gas evolution characteristics, and internal resistance changes, form a complete lifecycle profile. This system not only supports early identification and key monitoring of each individual cell, but also provides a basis for decision-making for subsequent differentiated control of individual cells. Attached Figure Description

[0049] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0050] Figure 1 A flowchart of a forced equalization charging system for lead-acid battery packs in electric vehicles. Detailed Implementation

[0051] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0052] Example 1, Reference Figure 1 The forced equalization charging system based on lead-acid battery packs for electric vehicles shown includes: a data acquisition module for real-time acquisition of dynamic operating parameters of each individual battery cell;

[0053] The dynamic operating parameters include voltage, current, temperature, timestamp, internal resistance, and SOC (state of charge).

[0054] The gas evolution initiation detection module identifies the starting point of battery gas evolution behavior based on the collected dynamic operating parameters, and detects and outputs the initiation event parameter vector of battery gas evolution behavior.

[0055] The initial event parameter vector includes the gas evolution initiation time S, the gas evolution initiation voltage V, the gas evolution initiation temperature T, and the current internal resistance R.

[0056] Among them, the gas evolution start time S refers to the time point when the single cell first shows obvious gas evolution behavior during the constant voltage charging stage. This parameter is obtained by real-time monitoring of the current change trend during the charging process. In the early stage of constant voltage charging, the battery mainly undergoes electrochemical reactions, and the charging current decreases slowly.

[0057] When the gas evolution threshold is reached, the water decomposition reaction (hydrogen evolution and oxygen evolution) is activated, which leads to a significant acceleration in the rate of current decrease. By setting the inflection point criterion of the current change rate and combining it with voltage plateau stability analysis, the starting moment S of the gas evolution behavior is accurately identified, and its time difference relative to the start of the constant voltage stage is recorded as an important temporal characteristic for assessing the battery aging state.

[0058] Among them, the gas evolution initiation voltage V refers to the terminal voltage value corresponding to the start of significant gas evolution reaction in a single cell during constant voltage charging. This parameter is extracted simultaneously with the detection of the gas evolution initiation time S, that is, the actual measured voltage of the battery at that time is taken. Since the gas evolution voltage V is closely related to the electrolyte concentration and the state of the plates, healthy batteries usually start gas evolution near a specific voltage (about 2.35V / cell for lead-acid batteries), while aged batteries can start gas evolution at a lower voltage due to enhanced polarization or reduced active materials. Therefore, the degree of deviation of the gas evolution initiation voltage V can directly reflect the degradation of the internal chemical state of the battery.

[0059] Among them, the gas evolution initiation temperature T refers to the real-time temperature value of the battery surface or interior when gas evolution begins. This parameter is obtained in real time from the dynamic operating parameters. When the gas evolution initiation time S is identified, the real-time temperature value of the individual battery is obtained simultaneously. The high temperature of the individual battery will reduce the gas evolution voltage threshold and accelerate water decomposition. Therefore, temperature compensation analysis combined with the gas evolution initiation temperature T can eliminate the interference of environmental or operating temperature rise on the judgment of gas evolution characteristics and ensure the accuracy and comparability of health status assessment.

[0060] The current internal resistance R refers to the AC or DC internal resistance value of the battery measured by the pulse discharge method at the initial moment S of gas evolution. The specific method is to apply a short-duration, low-amplitude current pulse during constant voltage charging and simultaneously collect the instantaneous voltage drop value and current. The current internal resistance R is calculated according to Ohm's law and is used to reflect the resistance of ion conduction and charge transfer processes inside the single cell. As the battery ages, the current internal resistance R will increase significantly due to grid corrosion and electrolyte drying. It is used as a synchronous parameter in the gas evolution process to comprehensively evaluate the degree of electrochemical performance degradation of the battery.

[0061] The gas evolution behavior analysis module extracts features and performs physical transformation on the parameter vector of the gas evolution initiation event to construct an original health feature vector that reflects the battery aging mechanism.

[0062] The original health feature vector includes the gas evolution voltage offset ΔV, the constant pressure response delay ΔS, and the internal resistance growth rate R. T ;

[0063] The gassing voltage offset ΔV is calculated by comparing the gassing start voltage V of the current battery with the standard voltage of a healthy battery at the start of charging, where the standard voltage refers to the typical voltage value exhibited by a healthy battery in the initial stage of charging;

[0064] The calculation of the gassing voltage offset ΔV is as follows:

[0065] ΔV = V1 - V;

[0066] where V1 is the standard voltage at the start of charging of a healthy battery. When the battery is in a healthy state, its gassing behavior occurs near the standard voltage V1;

[0067] As the battery ages (such as increased negative sulfate formation, electrolyte stratification, and shedding of active material on the electrode plate), the polarization effect increases, and the water decomposition reaction is activated earlier, resulting in a significant decrease in the gassing start voltage V. Therefore, the larger the value of the gassing voltage offset ΔV, the more severe the battery aging.

[0068] The constant voltage response delay ΔS is determined by comparing the gassing start moment S when the current battery reaches the constant voltage charging state with the standard time point when a healthy battery starts to gas under the same conditions. The standard response time refers to the time required for a healthy battery at the start of constant voltage charging;

[0069] The calculation of the constant voltage response delay ΔS is as follows:

[0070] ΔS = S1 - S;

[0071] where S1 is the standard response time of a healthy battery at the start of constant voltage charging. ΔS is not only a time difference but also a direct manifestation of the degradation of the internal electrochemical kinetic performance of the battery. When the battery is in a healthy state, its electrochemical system responds quickly and stably, and the gassing behavior occurs within a short time after the start of the constant voltage stage. As the battery ages (such as increased negative sulfate formation, electrolyte stratification, and shedding of active material on the electrode plate), the internal polarization of the battery increases, and the ion migration resistance increases, resulting in more charging energy being used to overcome the internal resistance rather than for effective charging, thus delaying the start of the gassing reaction. Therefore, the actual gassing start moment S is earlier than the standard time S1, that is, S < S1, resulting in ΔS > 0.

[0072] Internal resistance growth rate R T reflects the trend of the battery internal resistance changing with temperature and is obtained by comparing the internal resistance after temperature compensation with the internal resistance of the single - cell battery at the standard reference temperature. The internal resistance after temperature compensation is the percentage change in the battery resistance per 1°C increase, which is used to reflect the change in the battery performance under actual use conditions.

[0073] Internal resistance growth rate R T The calculation is as follows:

[0074]

[0075] Where 'a' represents the percentage change in battery resistance for every 1°C increase;

[0076] R1 is the internal resistance after temperature compensation;

[0077] R0 is the standard reference temperature when a single cell is charging, and T1 is the reference temperature when a single cell is charging.

[0078] When the battery is brand new or in a healthy state, its active material structure is intact, the electrolyte is sufficient, the grid contact is good, and the internal resistance is close to the initial value R0. Therefore, the internal resistance growth rate R is high. T ≈1.0;

[0079] As batteries age, grid corrosion narrows the conductive path, electrolyte dehydration hinders ion conduction, and sulfation increases interfacial impedance, causing the internal resistance to continuously rise (R1 > R0), resulting in an internal resistance growth rate R. T >1.

[0080] Example 2: The single-cell health assessment module obtains the initial state of charge (SOC) parameters of a single cell from dynamic operating parameters, corrects the vectors in the original health feature vector affected by the SOC parameters, calculates the current SOH assessment value of the single cell based on the corrected health feature vector F, and combines the current SOH assessment value with the historical SOH value. p Data is trend-following fusion to generate a comprehensive SOH. S Evaluation results;

[0081] The vectors in the original health feature vector that are affected by the SOC parameter include the gas evolution voltage offset ΔV and the constant pressure response delay ΔS;

[0082] The correction calculation for the gas evolution voltage offset ΔV is as follows:

[0083] ΔVsoc=ΔV-ΔV1+ΔV2;

[0084] ΔV1 is the expected value of a healthy battery under the current SOC;

[0085] ΔV2 is the gas evolution voltage offset ΔV of a healthy battery when SOC = 100%;

[0086] ΔVsos is the corrected gas evolution voltage offset.

[0087] The correction for the constant voltage response delay ΔS is calculated as follows:

[0088] ΔSsoc=ΔS-ΔS1+ΔS2;

[0089] ΔS1 is the expected constant voltage response delay time of a healthy battery at the current SOC.

[0090] ΔS2 is the expected constant voltage response delay time of a healthy battery when SOC = 100%;

[0091] The constant-pressure response delay time after ΔSsoc correction.

[0092] It is worth noting that the health feature vector F, corrected by the initial state of charge (SOC) parameter, is an equivalent representation of all measured values ​​under the same state. This effectively eliminates the evaluation bias caused by different charging start conditions, making the health features comparable across cycles and environments. Especially in complex application scenarios such as frequent charging of electric vehicles and multi-stage charging and discharging of energy storage systems, this mechanism ensures the long-term stable output of SOH evaluation results, avoids the problem of misjudging different results for the same battery, and improves the intelligence level of the battery management system.

[0093] The current SOH assessment value of a single cell is calculated based on the corrected health feature vector F, and the calculation method is as follows:

[0094]

[0095] C i This represents the center vector of the i-th hidden layer neuron, specifically used to represent the feature center of a specific aging mode in a single cell, representing a particular aging mode, C. i It is trained based on a large amount of historical data and can characterize the typical features of a battery at a certain aging stage.

[0096] ||FC i || represents the Euclidean distance between the current battery health feature vector F and the i-th "battery-specific aging mode" ci. This distance measures the similarity between the current battery features and the specific aging mode. The smaller the distance, the closer the current battery features are to the aging mode. exp is the similarity function used to convert the Euclidean distance into similarity weights.

[0097] W i The output weight of the i-th neuron represents the contribution of the current mode to SOH. It is a parameter obtained through network training and reflects the importance of each aging mode in assessing battery health status.

[0098] N represents the number of hidden layer neurons, i.e., the total number of aging patterns.

[0099] It is worth noting that the process of calculating SOH can be divided into the following steps:

[0100] S1. Input the corrected health feature vector F into the RBF neural network;

[0101] S2. For each hidden layer neuron i, calculate F and C. i Euclidean distance between them ||FC i ||;

[0102] S3. Use the Gaussian kernel function to convert distance into similarity weights;

[0103] S4. Combine the similarity weights with the corresponding output weights W i Multiplying these values ​​yields the contribution of this aging mode to SOH;

[0104] S5. Sum the contributions of all aging modes to obtain the final current SOH assessment value;

[0105] The current SOH evaluation value is achieved through the multi-aging mode matching mechanism of the RBF neural network, which enables refined identification and quantitative evaluation of complex aging paths of the battery. This is achieved by using the center vector C of the hidden layer neurons. i Defined as the feature center of a specific aging mode, such as typical degradation modes such as positive electrode grid corrosion-dominated type, negative electrode sulfation type, and electrolyte dehydration type, each center vector represents the standard health feature vector under this mode;

[0106] During the evaluation process, the system inputs the corrected health feature vector F into the network, calculates its Euclidean distance to the center of each pattern, and transforms it into similarity weights through a Gaussian kernel function. Finally, it outputs the current SOH value by weighting. This mechanism not only has strong nonlinear fitting ability, but also realizes the integration of "pattern recognition + health assessment", which significantly improves the physical interpretability and discrimination accuracy of the model. Compared with black box models, this method can effectively distinguish different aging mechanisms and avoid misjudgment. For example, a battery with premature gas evolution due to water loss and another battery with increased internal resistance due to grid corrosion may have similar SOH, but their feature vector distributions are different. The model can identify their respective modes and give differentiated evaluation suggestions.

[0107] Based on the current SOH assessment value and combined with historical comprehensive SOH p Data is trend-following and a comprehensive SOH is calculated. S The evaluation results are as follows:

[0108] SOH S =α·SOH + (1-α)·SOH p ;

[0109] SOH p For the previous comprehensive SOH S The evaluation results show that the SOH (State of Health) is highest when a single cell is being charged for the first time. p =SOH;

[0110] α is the smoothing coefficient.

[0111] It is worth noting that a three-tiered progressive evaluation architecture of feature correction, pattern recognition, and trend fusion achieves high-precision and robust battery health state estimation. First, health features affected by SOC interference are dynamically corrected to ensure consistency and comparability of evaluation results under different operating conditions. Second, a multi-aging pattern matching mechanism based on RBF neural networks is employed, utilizing the hidden layer center vector to represent typical aging paths, and comparing the current SOH evaluation value with historical SOH values. p By performing trend fusion, a more stable comprehensive SOH is generated. S The evaluation results, by suppressing the influence of transient noise and outlier data, not only improved the accuracy and continuity of SOH estimation;

[0112] Secondly, after each charge, the current SOH assessment value is compared with the previous comprehensive SOH value. p Weighted fusion is performed, where α is a smoothing coefficient used to adjust the contribution weights of new and old data. When the battery is charged for the first time, SOH p Initialized to the current SOH value, and then updated sequentially to give historical data a continuous influence, effectively suppressing transient interference and improving the overall SOH. S The evaluation results show a smooth and continuous trend, which more accurately reflects the long-term degradation pattern of the battery.

[0113] Furthermore, by adjusting the α value, the response speed and stability can be flexibly balanced. A larger α value results in a faster response, which is suitable for the rapid aging stage, while a smaller α value results in a smoother response, which is suitable for the stable operation period. This method not only improves the credibility of the evaluation results but also provides a reliable data foundation for advanced functions such as life prediction and fault warning. Moreover, after adopting the trend fusion mechanism, the fluctuation amplitude of the SOH evaluation curve is reduced by more than 60%, which is significantly better than the traditional single-point evaluation method, realizing a leap from instantaneous judgment to trend decision-making.

[0114] The battery group management module, based on SOH... S The evaluation results are continuously updated to categorize each individual battery cell based on its health profile.

[0115] Based on comprehensive SOH S The individual cell health profile obtained from the evaluation results includes basic information about the battery, current state of health (SOH), and other relevant data. S Evaluation results and historical trend data are used to comprehensively record and continuously track the health status of the battery.

[0116] Battery category is based on the comprehensive SOH in the health record. S The assessment results and multidimensional health characteristics were divided into five levels, namely, novel (SOH) S ≥98%), healthy (90%≤SOH) S <98%), Warning (80%≤SOH) S<90%), deterioration (70%≤SOH S <80%) and scrap (SOH) S <70%), used for a refined classification of the health status of individual cells, and the classification of individual cells is shown in Table 1.

[0117] Table 1

[0118]

[0119] It is worth noting that by calculating the overall SOH of each battery cell... S The system categorizes batteries into five classes based on preset standards, each corresponding to different aging levels and risk grades. The classification results are updated in real time to the health record of each battery. Historical trend data includes multi-dimensional data such as gas evolution characteristics and internal resistance changes, forming a complete life cycle profile. This system not only supports early identification and key monitoring of each individual battery, but also provides a basis for decision-making for subsequent differentiated control of individual batteries. For example, the system can automatically mark warning batteries and prompt maintenance personnel to pay attention, or initiate protective charging for deteriorated batteries.

[0120] The charging mode customization module allows you to configure exclusive charging parameters based on the battery type.

[0121] Based on the battery type, dedicated charging parameters are configured. New and healthy batteries have good performance, high consistency, and slight aging. The charging strategy adopts the standard constant current-constant voltage mode, which allows for higher charging current to achieve fast and full charging. It is planned to force equalization charging in the later stage of charging to correct the differences in battery voltage, SOC and current internal resistance R, and avoid energy loss caused by excessive equalization.

[0122] Warning type batteries have shown obvious signs of aging, such as premature gas evolution, increased internal resistance, or delayed constant voltage response, which pose a risk of local overcharging or temperature rise. Therefore, the charging parameters need to be moderately conservative, the constant current charging current should be reduced, the constant voltage should be lowered, and the constant voltage stage should be terminated early (reducing the cutoff current) to slow down the aging process. The plan is to perform active balancing in real time during the charging process, and to force the balancing circuit to start in the middle of the charging process to actively suppress voltage deviation and prevent overvoltage of a single cell.

[0123] Degraded batteries exhibit significantly reduced performance, markedly increased internal resistance, severe polarization, and a risk of thermal runaway. To ensure safety, strict protective charging must be implemented, including low-current constant-current charging, further reducing constant-voltage, setting stricter temperature protection thresholds, and shortening charging time. Forced active balancing is employed, which involves forcibly balancing the circuit from the start of charging, prioritizing energy discharge from high-voltage cells to prevent them from entering dangerous areas, while avoiding overcharging of degraded cells.

[0124] The scrapped type is due to the severe degradation of battery capacity, which no longer has normal use value. Normal charging process should be prohibited. Only maintenance charging with extremely low power is allowed under controlled conditions. The temperature and voltage changes of individual cells must be monitored in real time, and an alarm should be triggered immediately. The cells should be marked as "to be replaced" in the health record to prevent them from continuing to participate in the group's work, so as to avoid affecting other normal cells or causing safety accidents.

[0125] Among them, the closed-loop control mechanism that configures exclusive charging parameters based on battery type enables power-dependent policies. It dynamically configures exclusive charging parameters according to the category of individual battery cells, which not only improves charging efficiency but also significantly extends the overall lifespan of the battery pack and reduces the risk of thermal runaway.

[0126] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A forced equalization charging system based on a lead-acid battery pack for electric vehicles, characterized in that, include: The data acquisition module collects the dynamic operating parameters of each individual battery cell in real time. The gas evolution initiation detection module identifies the starting point of battery gas evolution behavior based on the collected dynamic operating parameters, and detects and outputs the initiation event parameter vector of battery gas evolution behavior. The gas evolution behavior analysis module extracts features and performs physical transformation on the parameter vector of the gas evolution initiation event to construct an original health feature vector that reflects the battery aging mechanism. The single-cell health assessment module obtains the initial state of charge (SOC) parameters of individual cells from dynamic operating parameters, corrects the vectors in the original health feature vector affected by the SOC parameters, calculates the current SOH assessment value of the individual cell based on the corrected health feature vector F, and combines the current SOH assessment value with the historical SOH value. p Data is trend-following fusion to generate a comprehensive SOH. S Evaluation results; The battery group management module, based on SOH... S The evaluation results are continuously updated to categorize each individual battery cell based on its health profile. The charging mode customization module allows you to configure exclusive charging parameters based on the battery type.

2. The forced equalization charging system based on a lead-acid battery pack for electric vehicles according to claim 1, characterized in that: The initial event parameter vector includes the gas evolution initiation time S, the gas evolution initiation voltage V, the gas evolution initiation temperature T, and the current internal resistance R.

3. The forced equalization charging system based on a lead-acid battery pack for electric vehicles according to claim 2, characterized in that: The original health feature vector includes the gas evolution voltage offset ΔV, the constant pressure response delay ΔS, and the internal resistance growth rate R. T ; The gas evolution voltage offset ΔV is calculated by comparing the difference between the current gas evolution initiation voltage V of the battery and the standard voltage of a healthy battery at the start of charging. The standard voltage refers to the typical voltage value exhibited by a healthy battery at the beginning of charging. The constant voltage response delay ΔS is determined by comparing the gas evolution start time S when the current battery reaches the constant voltage charging state with the standard time point when a healthy battery starts gas evolution under the same conditions. The standard response time refers to the time required for a healthy battery to start constant voltage charging. The internal resistance growth rate R T It reflects the trend of battery internal resistance changing with temperature. It is derived by comparing the internal resistance after temperature compensation with the internal resistance of a single cell at the standard reference temperature. The internal resistance after temperature compensation is the percentage change in battery resistance for every 1°C increase, and is used to reflect the changes in battery performance under actual use conditions.

4. The forced equalization charging system based on a lead-acid battery pack for electric vehicles according to claim 1, characterized in that: The vectors in the original health feature vector that are affected by the SOC parameter include the gas evolution voltage offset ΔV and the constant pressure response delay ΔS. The correction calculation for the gas evolution voltage offset ΔV is as follows: ΔVsoc=ΔV-ΔV1+ΔV2; ΔV1 is the expected value of a healthy battery under the current SOC; ΔV2 is the gas evolution voltage offset ΔV of a healthy battery when SOC = 100%; ΔVsos is the corrected gas evolution voltage offset.

5. A forced equalization charging system based on a lead-acid battery pack for electric vehicles according to claim 4, characterized in that: The correction calculation for the constant voltage response delay ΔS is as follows: ΔSsoc=ΔS-ΔS1+ΔS2; ΔS1 is the expected constant voltage response delay time of a healthy battery at the current SOC. ΔS2 is the expected constant voltage response delay time of a healthy battery when SOC = 100%; The constant-pressure response delay time after ΔSsoc correction.

6. The forced equalization charging system based on a lead-acid battery pack for electric vehicles according to claim 1, characterized in that: The current SOH assessment value of a single cell is calculated based on the corrected health feature vector F, and the calculation method is as follows: C i This represents the center vector of the i-th hidden layer neuron; ||FC i || represents the Euclidean distance between the current battery's health feature vector F and the i-th ci, while exp is the similarity function; W i The output weight of the i-th neuron; N represents the number of hidden layer neurons.

7. A forced equalization charging system based on a lead-acid battery pack for electric vehicles according to claim 6, characterized in that: Based on the current SOH assessment value and combined with historical SOH values p Data is trend-following and a comprehensive SOH is calculated. S The evaluation results are as follows: SOH S α·SOH+(1-α)·SOH p ; SOH p For the previous comprehensive SOH S The evaluation results show that the SOH (State of Health) is highest when a single cell is being charged for the first time. p =SOH; α is the smoothing coefficient.

8. The forced equalization charging system based on a lead-acid battery pack for electric vehicles according to claim 1, characterized in that: Based on the aforementioned comprehensive SOH S The individual cell health profile obtained from the evaluation results includes basic information about the battery, current state of health (SOH), and other relevant data. S Evaluation results and historical trend data are used to comprehensively record and continuously track the health status of the battery.

9. A forced equalization charging system based on a lead-acid battery pack for electric vehicles according to claim 8, characterized in that: The battery category is based on the comprehensive SOH in the health record. S The assessment results and multidimensional health characteristics are divided into five levels: brand new, healthy, warning, deteriorated, and scrap, which are used for a refined classification of the health status of individual batteries.

10. A forced equalization charging system based on a lead-acid battery pack for electric vehicles according to claim 9, characterized in that: Based on the battery type, dedicated charging parameters are configured. The new and healthy types are planned to force equalization charging in the later stages of charging to correct the differences in battery voltage, SOC and current internal resistance R, and avoid energy loss caused by excessive equalization. The warning type is planned to perform active balancing in real time during the charging process, and to forcibly start the balancing circuit in the middle of the charging process to actively suppress voltage deviation and prevent overvoltage of a single cell. Degradation types are divided into forced active balancing, which means that the balancing circuit is forced to start balancing at the beginning of charging, giving priority to energy discharge of high-voltage individual cells to prevent them from entering dangerous areas, while avoiding overcharging of degraded cells. Scrapping-type batteries are those whose capacity has severely degraded and no longer have normal usability. Normal charging should be prohibited, and only maintenance charging at very low power is allowed under controlled conditions. The temperature and voltage changes of individual cells must be monitored in real time.

Images