Battery pack state evaluation method and system

Abstract

The invention relates to the technical field of battery state evaluation, and discloses a battery pack state evaluation method and system, and the method comprises the steps: building a first corresponding relation between a dQ / dV-V curve data characteristic value and the temperature, the charging and discharging current, and the cycle index, and a second corresponding relation between the battery capacity and the dQ / dV-V curve data characteristic value through a machine learning model; if the dQ / dV-V curve data of charging and discharging in the standby state can be obtained, substituting the characteristic value of the dQ / dV-V curve data into the second corresponding relation to obtain the predicted battery capacity; if not, acquiring cycle index data under each charging and discharging current at each temperature in the historical charging and discharging data, predicting a dQ / dV-V curve data characteristic value based on the first corresponding relationship, and combining with a second corresponding relationship to obtain the predicted battery capacity; and the state parameters of the battery pack are evaluated based on the dQ / dV-V curve data characteristic values and the predicted battery capacity, so that the charging and discharging strategy is corrected, the use condition limitation is adjusted, and the service life of the battery is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of battery state evaluation, and particularly to a method and system for evaluating the state of a battery pack. Background Art

[0002] Lithium-ion batteries, sodium-ion batteries, etc. are widely used in power batteries, energy storage batteries, 3C consumer fields, etc. However, due to the complex usage conditions and performance attenuation, it is difficult to obtain the true battery capacity of the battery pack. The actual usage rate will be higher than the battery's own capacity in the middle and late stages of the cycle, resulting in accelerated battery attenuation and affecting the actual service life. Moreover, the current voltage monitoring is limited by the characteristics of the voltage platform of lithium iron phosphate batteries and is not sensitive to voltage changes during the platform period, which has certain limitations for identifying abnormal battery cells. And the temperature acquisition has a certain lag due to the limitations of the battery acquisition points and the number, and the battery management system cannot adjust the usage strategy of the battery pack in time, resulting in a reduction in the service life of the battery. Summary of the Invention

[0003] In view of this, the present invention provides a method and system for evaluating the state of a battery pack, which solves the problem of how to accurately evaluate the state of the battery pack, so that it can timely feedback to the battery management system for correcting the charge and discharge strategy and adjusting the usage condition limit, and extending the service life of the battery.

[0004] In a first aspect, the present invention provides a method for evaluating the state of a battery pack, the method comprising:

[0005] Using a preset machine learning model to fit the first corresponding relationship between the dQ / dV-V curve data eigenvalues at different temperatures, different charge and discharge currents, and different cycle numbers and the temperature, charge and discharge current, and cycle number, and the second corresponding relationship between the battery capacity and the dQ / dV-V curve data eigenvalues;

[0006] Determining whether the dQ / dV-V curve data of the battery pack to be measured after a preset running time can be obtained under a preset small rate charge and discharge in the standby state;

[0007] If the dQ / dV-V curve data of the preset small rate charge and discharge in the standby state can be obtained, then using the obtained dQ / dV-V curve data eigenvalues of the preset small rate charge and discharge and substituting them into the second corresponding relationship to obtain the corresponding predicted battery capacity;

[0008] If the dQ / dV-V curve data of the preset small rate charge and discharge in the standby state cannot be obtained, then obtaining the cycle number data at each temperature and each charge and discharge current in the battery historical charge and discharge data, predicting the dQ / dV-V curve data eigenvalues based on the first corresponding relationship, and combining with the second corresponding relationship to obtain the predicted battery capacity;

[0009] Evaluate the state parameters of the battery pack based on the dQ / dV-V curve data eigenvalues and the predicted battery capacity.

[0010] The actual usage conditions of the battery are complex and variable. Temperature, charge and discharge current, and cycle number will all have a significant impact on battery performance. The battery pack state evaluation method provided by the embodiments of the present invention can more accurately grasp the internal electrochemical process of the battery by comprehensively considering these factors, thereby improving the accuracy of battery capacity prediction, and ultimately making the battery pack state evaluation more accurate. For whether the dQ / dV-V curve data of preset small-rate charge and discharge in the standby state can be obtained, different strategies are adopted to predict the battery capacity, so that the evaluation method can adapt to different actual scenarios, enhancing the flexibility and applicability of the evaluation. Accurate dQ / dV-V curve data eigenvalues and battery capacity prediction can enable the battery management system to timely detect the decline trend of battery performance, thereby correcting the charge and discharge strategy, adjusting the usage condition limit, extending the service life of the battery, reducing the failure risk within the product life cycle, and improving customer satisfaction.

[0011] In an alternative embodiment, the step of using a preset machine learning model to fit the first corresponding relationship between the dQ / dV-V curve data eigenvalues and temperature, charge and discharge current, and cycle number, and the second corresponding relationship between the battery capacity and the dQ / dV-V curve data eigenvalues under different temperatures, different charge and discharge currents, and different cycle numbers includes:

[0012] Obtain the dQ / dV-V curves under different cycle numbers, different temperatures, and different charge and discharge currents as the first sample data;

[0013] Based on the first sample data, use the XGboost model to fit the relationship between the dQ / dV-V curve data eigenvalues and the charge and discharge current, temperature, and cycle number under different charge and discharge currents, temperatures, and cycle numbers as the first corresponding relationship;

[0014] Obtain the dQ / dV-V curve data eigenvalues under different battery capacities as the second sample data;

[0015] Based on the first sample data or the second sample data, use the XGboost model to fit the corresponding relationship between the battery capacity and the dQ / dV-V curve data eigenvalues as the second corresponding relationship.

[0016] In the embodiments of the present invention, the first corresponding relationships between the data characteristic values of the dQ / dV-V curve and these factors are established respectively under different charge and discharge currents, temperatures, and number of cycles, and the second corresponding relationship between the battery capacity and the data characteristic values of the dQ / dV-V curve is established, covering the information of the battery under various actual usage scenarios, realizing the multi-dimensional analysis of the battery state, being able to more comprehensively reflect the actual state of the battery, avoiding the limitations brought by single-factor or simple linear relationship analysis, thereby improving the reliability of the battery pack state assessment. By virtue of the strong adaptability of the XGboost model, it can accurately fit the corresponding relationships under different working conditions, providing a reliable basis for the battery pack state assessment and ensuring that the assessment results have high accuracy and credibility in various actual situations.

[0017] In an alternative embodiment, the data characteristic values of the dQ / dV-V curve include: peak value, peak area, peak time, lower endpoint of the plateau voltage interval, and upper endpoint of the plateau voltage interval.

[0018] In the embodiments of the present invention, multiple characteristic values such as peak value, peak area, peak time, lower endpoint and upper endpoint of the plateau voltage interval are combined for analysis, which can provide more comprehensive and accurate battery state information. Different types of batteries (such as lithium-ion batteries, lead-acid batteries, etc.) and different usage conditions (such as different charge and discharge currents, temperatures, etc.) will cause changes in the data characteristic values of the dQ / dV-V curve. By selecting multiple representative characteristic values, it is possible to better adapt to the characteristics of different battery systems and working conditions, and improve the generality and adaptability of the battery state assessment method.

[0019] In an alternative embodiment, based on the predicted battery capacity to evaluate the battery pack state parameters, including:

[0020] Based on the calculation models of the predicted battery capacity and the SOC and SOH of the battery pack, the SOC value and SOH value corresponding to the predicted battery capacity are obtained. When the SOC value and SOH value corresponding to the predicted battery capacity are closer to the actual values than the SOC value and SOH value provided by the battery management system, the SOC value and SOH value provided by the battery management system are corrected, and the preset charge and discharge current table in the battery management system is corrected based on the predicted battery capacity.

[0021] Embodiments of the present invention help the battery management system to more precisely control the charging and discharging process of the battery through more accurate SOC and SOH values, avoiding overcharging and over-discharging. Overcharging and over-discharging can cause irreversible damage to the electrode materials, electrolytes, etc. of the battery, shortening the battery life. The corrected SOC and SOH values can provide a more reliable basis for the battery management system to stop the charging or discharging operation in time and protect the battery safety. By improving the battery usage safety and correcting the preset charging and discharging ammeter in the battery management system based on the predicted battery capacity, the charging and discharging process can be made more reasonable. The optimal charging and discharging currents of batteries with different health states and battery capacities are different. By correcting the preset charging and discharging ammeter, damage to the battery caused by excessive or too small charging and discharging currents can be avoided.

[0022] In an alternative embodiment, the method further includes: adjusting the SOC voltage range and the battery balancing voltage range according to the lower endpoint and the upper endpoint of the plateau voltage range of the predicted dQ / dV-V curve data.

[0023] Embodiments of the present invention can enable the battery management system to more precisely control the charging and discharging process by adjusting the SOC voltage range and the battery balancing voltage range. During charging, it can ensure that the battery is charged within a suitable voltage range, avoiding poor system consistency caused by overcharging or excessive end voltage difference, improving the charging efficiency and reducing the charging time. During the discharging process, it can also reasonably control the discharging depth according to the accurate SOC and balancing state, avoiding damage to the battery caused by over-discharging, thereby extending the battery life.

[0024] In an alternative embodiment, the method further includes: obtaining the peak voltage displacement change amount and the peak area change amount of the dQ / dV curve of each single battery in the battery pack, and when the peak voltage displacement change amount and / or the peak area change amount exceed the corresponding change threshold, adjusting the used DOD range and the upper and lower limit protection thresholds.

[0025] When the peak voltage displacement change amount and the peak area change amount exceed the threshold in embodiments of the present invention, it may mean that a large change has occurred in the internal chemical state of the battery, and there is a risk of overcharging or over-discharging. At this time, adjusting the upper and lower limit protection thresholds can timely limit the charging and discharging range of the battery, avoid the battery working in an extreme state, and irreversible damage to the battery caused by over-discharging, extend the battery life, and ensure the use safety. Adjusting the DOD range according to the peak voltage displacement change amount and the peak area change amount can enable each single battery to work within the discharging depth range suitable for its current state, avoiding premature attenuation of the battery capacity or performance degradation caused by unreasonable DOD, thereby improving the overall performance and energy utilization rate of the battery.

[0026] In a second aspect, the present invention provides a battery pack state evaluation system, including:

[0027] A correspondence establishing module, configured to use a preset machine learning model to fit a first correspondence between the characteristic values of the dQ / dV-V curve data and the temperature, charge-discharge current, and number of cycles at different temperatures, different charge-discharge currents, and different numbers of cycles, and a second correspondence between the battery capacity and the characteristic values of the dQ / dV-V curve data;

[0028] A data acquisition judgment module, configured to judge whether the dQ / dV-V curve data of the battery pack to be measured after a preset running time can be obtained under a preset small rate charge-discharge in the standby state;

[0029] A first prediction module, configured to, if the dQ / dV-V curve data of the preset small rate charge-discharge in the standby state can be obtained, use the obtained characteristic values of the dQ / dV-V curve data of the preset small rate charge-discharge and substitute them into the second correspondence to obtain the corresponding predicted battery capacity;

[0030] A second prediction module, if the dQ / dV-V curve data of the preset small rate charge-discharge in the standby state cannot be obtained, obtain the number of cycle data at each temperature and each charge-discharge current in the historical charge-discharge data of the battery, predict the characteristic values of the dQ / dV-V curve data based on the first correspondence, and combine the second correspondence to obtain the predicted battery capacity;

[0031] A state evaluation module, configured to evaluate the state parameters of the battery pack based on the characteristic values of the dQ / dV-V curve data and the predicted battery capacity.

[0032] In a third aspect, the present invention provides a computer device, including: a memory and a processor, which are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to execute the battery pack state evaluation method according to the first aspect or any corresponding embodiment thereof.

[0033] In a fourth aspect, the present invention provides a computer-readable storage medium, on which computer instructions are stored, and the computer instructions are used to cause a computer to execute the battery pack state evaluation method according to the first aspect or any corresponding embodiment thereof.

[0034] In a fifth aspect, the present invention provides a computer program product, including computer instructions, and the computer instructions are used to cause a computer to execute the battery pack state evaluation method according to the first aspect or any corresponding embodiment thereof. Description of the Drawings

[0035] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0036] Figure 1 is a schematic flowchart of a battery pack state evaluation method according to an embodiment of the present invention;

[0037] Figure 2 is a flowchart of constructing a battery feature correspondence based on a machine learning model according to an embodiment of the present invention;

[0038] Figure 3 is a dQ / dV-V curve graph under different aging states according to an embodiment of the present invention;

[0039] Figure 4 is a schematic flowchart of another battery pack state evaluation method according to an embodiment of the present invention;

[0040] Figure 5 is a schematic diagram of the platform voltage range under different aging states according to an embodiment of the present invention;

[0041] Figure 6 is a structural block diagram of a battery pack state evaluation system according to an embodiment of the present invention;

[0042] Figure 7 is a schematic hardware structure diagram of a computer device according to an embodiment of the present invention. Specific Embodiments

[0043] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0044] To overcome the deficiencies of the prior art, a battery pack state evaluation method is provided in this embodiment. Figure 1 is a flowchart of a battery pack state evaluation method according to an embodiment of the present invention. As Figure 1 shown, the process includes the following steps:

[0045] S11. Use a preset machine learning model to fit the first corresponding relationship between the eigenvalue of the dQ / dV-V curve data and temperature, charge-discharge current, and number of cycles at different temperatures, different charge-discharge currents, and different numbers of cycles, as well as the second corresponding relationship between battery capacity and the eigenvalue of the dQ / dV-V curve data.

[0046] Specifically, the dQ / dV-V curve, i.e., the capacity increment curve, is a curve plotted with the battery voltage (V) as the abscissa and the differential of the capacity with respect to the voltage (dQ / dV) as the ordinate. The eigenvalue of the dQ / dV-V curve data contains rich information inside the battery and can reflect key characteristics such as the phase change of the battery electrode material and the process of lithium-ion insertion / extraction. In the embodiment of the present invention, based on the machine learning model, a connection is established between the eigenvalue of the dQ / dV-V curve data and the battery capacity, which can more directly reflect the actual state inside the battery, avoiding errors that may occur when evaluating only based on external parameters (such as voltage, current, etc.), thereby improving the accuracy of the evaluation. The actual usage conditions of the battery are complex and variable, and temperature, charge-discharge current, and number of cycles will all have a significant impact on the battery performance. By comprehensively considering these factors, the electrochemical process inside the battery can be grasped more accurately, thereby improving the accuracy of predicting the battery capacity and ultimately making the evaluation of the battery pack state more precise.

[0047] S12. Determine whether the dQ / dV-V curve data of the battery pack to be tested after a preset running time can be obtained under a preset small charge-discharge rate in the standby state.

[0048] Specifically, the range of small charge-discharge rate data may vary depending on different battery types, application scenarios, and research purposes. For example, in some scientific research experiments with high requirements for battery performance, a charge-discharge rate less than 0.05C may also be regarded as a small charge-discharge rate; while in some ordinary application scenarios, 0.3C - 0.8C may also be considered a relatively small charge-discharge rate. In actual applications, the operating states of the battery pack are diverse, and sometimes it may not be possible to obtain the small charge-discharge rate data in the standby state. This way of handling different situations enables the evaluation method to adapt to different actual scenarios and enhances the flexibility and applicability of the evaluation.

[0049] S13. If the dQ / dV-V curve data of the preset small charge-discharge rate in the standby state can be obtained, then use the eigenvalue of the obtained dQ / dV-V curve data of the preset small charge-discharge rate and substitute it into the second corresponding relationship to obtain the corresponding predicted battery capacity.

[0050] Specifically, in the embodiments of the present invention, by obtaining the dQ / dV-V curve data and characteristic values of small-rate charge and discharge in the standby state to calculate and predict the battery capacity, the actual available battery capacity of the battery can be understood more accurately. Because the small-rate charge and discharge process is relatively closer to the static real state of the battery and is less affected by factors such as polarization during large-current charge and discharge (for example, within a certain period of time, the data obtained in each segment is accumulated, and it can be considered that the battery state is quite stable within the set time period), the predicted battery capacity obtained based on this can better reflect the true remaining power of the battery. At the same time, the predicted battery capacity is an important basis for evaluating the battery state.

[0051] S14. If the dQ / dV-V curve data of preset small-rate charge and discharge in the standby state cannot be obtained, obtain the cycle number data at each temperature and each charge and discharge current in the battery historical charge and discharge data, predict the characteristic values of the dQ / dV-V curve data based on the first correspondence, and obtain the predicted battery capacity in combination with the second correspondence.

[0052] Specifically, when the dQ / dV-V curve data in the standby state cannot be obtained, the cycle number data at each temperature and each charge and discharge current in the battery historical charge and discharge data can be used (for example, the system allows statistics after a period of time, 5 cycles at 0.3C at 5°C, 10 cycles at 0.5C at 10°C, 20 cycles at 0.6C at 15°C, etc.). Based on the first correspondence, the characteristic values of the dQ / dV-V curve data are predicted. This makes full use of the historical operation information of the battery, and can predict the battery capacity even in the absence of real-time small-rate charge and discharge data, further improving the flexibility of the evaluation method.

[0053] S15. Evaluate the state parameters of the battery pack based on the characteristic values of the dQ / dV-V curve data and the predicted battery capacity.

[0054] During the use of the battery, its battery capacity will gradually decay due to factors such as the increase in the number of charge and discharge cycles, the passage of use time, and the use environment. According to the accurate characteristic values of the dQ / dV-V curve data and the predicted battery capacity, accurate data can be provided for the battery management system (BMS) to formulate more optimized charge and discharge strategies for the battery, enabling the battery to always operate in the best working state, improving the charge and discharge efficiency of the battery, reducing the charging time, and enhancing the overall use performance of the battery.

[0055] Specifically, the SOC calculation model preset in the battery management system Among them, Q is the initial battery capacity, and SOC0 is obtained by looking up the static SOC-OCV curve in a table; as the battery is used, the battery capacity will undergo irreversible attenuation. The SOC′ calculated by predicting the battery capacity Q′ is compared with the SOC value calculated under the model of the battery management system itself to determine which one is closer to the actual value in terms of accuracy. In one embodiment, verification can be performed according to the relationship between the difference in SOC before and after battery use and the available power time. For example, the accuracy of the two SOCs is compared within one charge or discharge cycle. For instance, if the last displayed SOC is 30%, theoretically, discharging at 1C to 0% SOC can last for 18 minutes. The difference δt between the theoretical sustainable time t′ and t obtained from SOC′ and the displayed SOC and the true sustainable discharge time treal is used to compare the accuracy of the two SOC displays. The smaller Δt is, the closer the estimated SOC is to the true SOC of the battery.

[0056] Since the predicted battery capacity in the embodiments of the present invention takes into account the dQ / dV-V curve data characteristic values under different temperatures, different charge and discharge currents, and different cycle numbers, it can more accurately reflect the actual battery capacity of the battery. The SOC (State of Charge) value and SOH (State of Health, SOH = Q′ / Q) value obtained based on this are more accurate in reflecting the current power and health status of the battery compared to the original calculation results of the battery management system (BMS). For example, during the battery aging process, the BMS may cause deviations in the calculation of SOC and SOH due to insufficient consideration of the impact of complex working conditions on the battery capacity. However, the corrected value using the predicted battery capacity can correct this deviation and make the evaluation result closer to the actual situation.

[0057] The embodiments of the present invention also input the predicted Q′ into the charge and discharge ammeter preset in the battery management system to correct the charge and discharge current map. For example, for a 100Ah cell, the initial 1C charging current capacity is 100A; later, when the battery capacity decays to 90Ah, the charging current capacity is corrected to 90A according to 1C charging, reducing the relative rate of the battery and improving the cycle stability.

[0058] The battery pack state evaluation method provided by the embodiments of the present invention can more accurately grasp the electrochemical process inside the battery by comprehensively considering these factors, thereby improving the accuracy of battery capacity prediction, and ultimately making the battery pack state evaluation more precise. For whether the dQ / dV-V curve data of preset small-rate charge and discharge in the standby state can be obtained, different strategies are respectively adopted to predict the battery capacity, so that the evaluation method can adapt to different actual scenarios, enhancing the flexibility and applicability of the evaluation. Accurate battery capacity prediction can enable the battery management system to timely detect the decline trend of battery performance, thereby correcting the charge and discharge strategies, adjusting the usage condition limits, extending the service life of the battery, reducing the failure risk within the product life cycle, and improving customer satisfaction.

[0059] In one embodiment, step S11 is executed, as Figure 2 shown, and it includes the following steps:

[0060] S111, Obtain the dQ / dV-V curves under different cycle numbers, different temperatures, and different charge and discharge currents as the first sample data.

[0061] S112, Based on the first sample data, use the XGboost model to fit the relationship between the dQ / dV-V curve data eigenvalues and the charge and discharge current, temperature, and cycle number under different charge and discharge currents, temperatures, and cycle numbers as the first corresponding relationship;

[0062] S113, Obtain the dQ / dV-V curve data eigenvalues under different battery capacities as the second sample data.

[0063] S114, Based on the first sample data or the second sample data, use the XGboost model to fit the corresponding relationship between the battery capacity and the dQ / dV-V curve data eigenvalues as the second corresponding relationship.

[0064] Specifically, the dQ / dV-V curves under different aging states are as Figure 3 shown. The embodiments of the present invention obtain the dQ / dV-V curves under different cycle numbers X, different temperatures T, and different charge and discharge currents I as the first sample data, and the dQ / dV-V curve data eigenvalues under different battery capacities as the second sample data, covering the information of the battery under various actual usage scenarios. In actual applications, the corresponding sample data type is selected according to the obtained data. The rich and diverse data enables the model to learn more extensive features and patterns, thereby more accurately capturing the internal relationship between battery performance and various influencing factors.

[0065] The XGboost machine learning model adopted in the embodiments of the present invention is a powerful ensemble learning algorithm. It adopts the gradient boosting framework and forms a strong classifier by iteratively training multiple weak classifiers (decision trees) and combining them. This algorithm can automatically handle the complex interactions between features and has good fitting ability for non-linear relationships. When fitting the first corresponding relationship and the second corresponding relationship, XGboost can effectively mine the potential relationships between temperature, charge and discharge current, number of cycles, dQ / dV-V curve data feature values, and battery capacity, improve the fitting accuracy and generalization ability, and enable the model to have good prediction performance when facing new data.

[0066] Specifically: Let (xi, yi), i = 1, 2, 3... n be the modeling samples, is the predicted value of the k-th round, which is equal to the sum of the predicted values of the previous (k - 1) rounds plus the predicted value f k (x i ), where xi is the i-th input data. Through stacked training:

[0067]

[0068] To make the predicted value more accurate, the difference between and the true value yi and the complexity of the model should be small, that is, to minimize the objective function Obj;

[0069]

[0070] Furthermore, the approximate objective function is obtained through second-order Taylor expansion. By optimizing the objective function, the optimal solution can be obtained when the prediction tree structure is known. The parameters of the prediction tree model can be solved using the greedy algorithm or other algorithms. According to the training model, use R 2 or other evaluation metrics to evaluate the prediction accuracy of the model, and gradually optimize the position points of the sample values in the prediction tree until R 2 approaches an acceptable level, where:

[0071]

[0072] Among them, the predicted value is the true value is yi, and the expectation of the true value is

[0073] The eigenvalue of the dQ / dV-V curve data in the embodiments of the present invention includes: peak value M, peak area S, peak time t, upper endpoint v1 of the plateau voltage range, and lower endpoint v2 of the plateau voltage range; the magnitude of the peak value M directly reflects the maximum value of the rate of change of the capacity with respect to the voltage at a specific voltage during the charge and discharge process of the battery. The larger the peak value, the more intense the electrochemical reaction inside the battery at this voltage; the peak area S is directly related to the charge and discharge capacity of the battery within the corresponding voltage range. By calculating the peak area, the capacity contribution of the battery within different voltage ranges can be intuitively understood, and the change in the peak area S can reflect the attenuation of the battery capacity; the peak time t reflects the time required for the battery to reach the peak value M, which is closely related to kinetic factors such as the ion diffusion rate and charge transfer process inside the battery. A shorter peak time t means that the battery has a faster charge and discharge speed and good kinetic performance. According to the peak time t, the charge and discharge rate and time of the battery can be reasonably adjusted to achieve the best charge and discharge effect; the width of the plateau voltage range can reflect the stability of the battery performance. Determining v1 and v2 can provide an accurate voltage reference for the application of the battery, ensuring that the battery operates within a suitable voltage range and avoiding battery damage or performance degradation caused by too high or too low voltage.

[0074] The first corresponding relationships obtained by fitting based on the machine learning model in the embodiments of the present invention are: peak value M = f1(I, T, X), peak area S = f2(I, T, X); peak time t = f3(I, T, X); upper endpoint f(v1) of the plateau voltage range = f(I, T, X); lower endpoint f(v2) of the plateau voltage range = f(I, T, X); the second corresponding relationship is Q = F(v1, v2, M, S, t).

[0075] The method provided by the embodiments of the present invention, as Figure 4 shown, further includes:

[0076] S16, adjusting the SOC voltage range and the battery equalization voltage range according to the lower endpoint and the upper endpoint of the predicted plateau voltage range of the dQ / dV-V curve data.

[0077] Specifically, the plateau voltage ranges under different aging states are as Figure 5 shown. The voltage plateau range endpoints v1 and v2 can be used to correct the control thresholds of the battery management system. Generally, due to little change in the voltage during the plateau period and large SOC estimation errors, it is difficult to perform SOC calibration and battery equalization. By using the predicted voltage plateau range endpoints v1' and v2' to determine the voltage plateau range after aging and adjusting the SOC calibration and battery equalization voltage ranges, the accuracy of SOC calibration and battery equalization can be improved.

[0078] Based on the accurate SOC voltage range, the battery management system can formulate more reasonable charge and discharge strategies. For example, during the charging process, according to the current SOC state of the battery and the corresponding voltage range, the charging current and voltage can be precisely controlled to avoid poor system consistency caused by overcharging or excessive terminal voltage difference, and improve the charging efficiency; during the discharging process, according to the relationship between SOC and voltage, the load power can be reasonably adjusted to ensure that the battery operates within a safe voltage range and prevent irreversible damage caused by over-discharging of the battery; by adjusting the battery equalization voltage range, the equalization target voltage range of each single battery can be determined more accurately. Through reasonable charge and discharge control of single batteries in different voltage states, the voltages of each single battery gradually tend to be consistent, realizing the equalization management of the battery pack and improving the overall performance and lifespan of the battery pack.

[0079] S17. Obtain the peak voltage displacement change and peak area change of the dQ / dV curve of each single battery in the battery pack. When the peak voltage displacement change and / or peak area change exceed the corresponding change threshold, adjust the used DOD interval and upper and lower limit protection thresholds.

[0080] Specifically, during the charge and discharge process of the battery, the electrode material will undergo various chemical reactions and physical changes. The peak voltage displacement change and peak area change can intuitively reflect the internal electrochemical change situation of the battery. When these change amounts exceed the preset threshold, it means that the battery may be in an abnormal state, such as the structural change of the internal electrode material, the imbalance of chemical reactions, etc., and there may be risks of overcharging or over-discharging. At this time, adjusting the upper and lower limit protection thresholds can timely limit the charge and discharge range of the battery to avoid the battery continuing to work in a dangerous state. Adjusting the DOD (Depth of Discharge) interval can enable each single battery to work within the discharge depth range suitable for its current state, avoiding premature attenuation of battery capacity or performance degradation caused by unreasonable DOD.

[0081] According to the consistency of dQ / dV data in the embodiments of the present invention, the consistency of the internal electrochemical reactions and material phase transitions of the battery monomers in the battery pack can be judged; by comparing the dQ / dV-V data of each single battery in the battery pack, according to the comparison of the peak displacement change δM and peak area change δS of the dQ / dV of each battery monomer, for the SOC range (corresponding voltage range) in which either δM or δS changes significantly (exceeds the corresponding change threshold), appropriate exclusion is performed, that is, by narrowing the system voltage range and slowing down the material phase transition and electrochemical reaction process, the battery system can operate in a more stable state and extend the service life of the battery pack.

[0082] In this embodiment, a battery pack state evaluation system is also provided. This system is used to implement the above-mentioned embodiments and preferred implementation manners, and those that have been described will not be elaborated again. As used hereinafter, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the system described in the following embodiments is preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and contemplated.

[0083] This embodiment provides a battery pack state evaluation system, as Figure 6 shown, including:

[0084] A correspondence establishment module 601, configured to use a preset machine learning model to fit the first correspondence between the dQ / dV-V curve data eigenvalues at different temperatures, different charge and discharge currents, and different cycle numbers and the temperature, charge and discharge current, and cycle number, and the second correspondence between the battery capacity and the dQ / dV-V curve data eigenvalues;

[0085] A data acquisition judgment module 602, configured to judge whether the dQ / dV-V curve data of the battery pack to be measured after a preset running time can be obtained under a preset small rate charge and discharge in the standby state;

[0086] A first prediction module 603, configured to, if the dQ / dV-V curve data of the preset small rate charge and discharge in the standby state can be obtained, use the obtained dQ / dV-V curve data eigenvalues of the preset small rate charge and discharge and substitute them into the second correspondence to obtain the corresponding predicted battery capacity;

[0087] A second prediction module 604, configured to, if the dQ / dV-V curve data of the preset small rate charge and discharge in the standby state cannot be obtained, obtain the cycle number data at each temperature and each charge and discharge current in the battery historical charge and discharge data, predict the dQ / dV-V curve data eigenvalues based on the first correspondence, and combine the second correspondence to obtain the predicted battery capacity;

[0088] A state evaluation module 605, configured to evaluate the battery pack state parameters based on the dQ / dV-V curve data eigenvalues and the predicted battery capacity.

[0089] In some optional implementation manners, the correspondence establishment module 601 includes:

[0090] A first sample data acquisition unit, configured to acquire the dQ / dV-V curves at different cycle numbers, different temperatures, and different charge and discharge currents as the first sample data;

[0091] The first corresponding relationship obtaining unit is configured to, based on the first sample data, fit the relationship between the eigenvalue of the dQ / dV-V curve data and the charge-discharge current, temperature, and number of cycles under different charge-discharge currents, temperatures, and number of cycles through an XGboost model, as the first corresponding relationship;

[0092] The second sample data obtaining unit is configured to obtain the eigenvalue of the dQ / dV-V curve data under different battery capacities as the second sample data;

[0093] The second corresponding relationship obtaining unit is configured to, based on the first sample data or the second sample data, fit the corresponding relationship between the battery capacity and the eigenvalue of the dQ / dV-V curve data through an XGboost model, as the second corresponding relationship.

[0094] In some optional embodiments, the eigenvalue of the dQ / dV-V curve data includes: peak value, peak area, peak time, lower endpoint of the platform voltage range, and upper endpoint of the platform voltage range.

[0095] In some optional embodiments, the state evaluation module 605 includes a calculation model based on the predicted battery capacity and the SOC and SOH of the battery pack to obtain the SOC value and SOH value corresponding to the predicted battery capacity. When the SOC value and SOH value corresponding to the predicted battery capacity are closer to the actual values than the SOC value and SOH value provided by the battery management system, the SOC value and SOH value provided by the battery management system are corrected, and the preset charge-discharge current table in the battery management system is corrected based on the predicted battery capacity.

[0096] In some optional embodiments, the above system further includes: a first adjustment module configured to adjust the SOC voltage range and the battery balancing voltage range according to the lower endpoint of the platform voltage range and the upper endpoint of the platform voltage range of the predicted dQ / dV-V curve data.

[0097] In some optional embodiments, the above system further includes: a second adjustment module configured to obtain the peak voltage displacement change amount and the peak area change amount of the dQ / dV curve of each single battery in the battery pack, and when the peak voltage displacement change amount and / or the peak area change amount exceeds the corresponding change threshold, adjust the DOD range and the upper and lower limit protection thresholds for use.

[0098] The further function descriptions of the above-mentioned modules and units are the same as those in the corresponding embodiments above, and will not be repeated here.

[0099] The battery pack state evaluation system in this embodiment is presented in the form of functional units. Here, the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0100] An embodiment of the present invention also provides a computer device having the above Figure 6 shown battery pack state evaluation system.

[0101] Please refer to Figure 7 , Figure 7 which is a schematic structural diagram of the computer device provided by an optional embodiment of the present invention. As Figure 7 shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including a high-speed interface and a low-speed interface. Each component communicates with each other using different buses and can be installed on a common motherboard or installed in other ways as needed. The processor can process instructions executed within the computer device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (such as a server array, a set of blade servers, or a multi-processor system). Figure 7 In

[0102] Figure, one processor 10 is taken as an example.

[0103] The memory 20 stores instructions executable by at least one processor 10, so that at least one processor 10 executes the method shown in the above embodiment.

[0104] The memory 20 may include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the computer device and the like. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 may optionally include a memory remotely disposed relative to the processor 10, and these remote memories can be connected to the computer device through a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0105] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk, or a solid-state drive; the memory 20 may further include a combination of the above-mentioned types of memories. The computer device further includes a communication interface 30 for communicating the computer device with other devices or communication networks.

[0106] The embodiments of the present invention also provide a computer-readable storage medium. The method according to the embodiments of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented as computer code originally stored in a remote storage medium or a non-transitory machine-readable storage medium and downloaded through a network and to be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memories. It can be understood that a computer, a processor, a microprocessor central control system, or programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the method shown in the above embodiments is implemented.

[0107] A part of the present invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the present invention through the operations of the computer. Those skilled in the art should understand that the forms of existence of computer program instructions in a computer-readable medium include but are not limited to source files, executable files, installation package files, etc. Correspondingly, the ways for a computer to execute computer program instructions include but are not limited to: the computer directly executes the instructions, or the computer compiles the instructions and then executes the corresponding compiled program, or the computer reads and executes the instructions, or the computer reads and installs the instructions and then executes the corresponding installed program. Herein, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to the computer.

[0108] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for evaluating the state of a battery pack, characterized in that, Including: Using a preset machine learning model to fit the first corresponding relationship between the eigenvalue of the dQ / dV-V curve data and temperature, charge-discharge current, and number of cycles at different temperatures, different charge-discharge currents, and different numbers of cycles, as well as the second corresponding relationship between battery capacity and the eigenvalue of the dQ / dV-V curve data; Judging whether the battery pack to be tested after a preset running time can obtain the dQ / dV-V curve data of preset small-rate charge and discharge in the standby state; If the dQ / dV-V curve data of preset small-rate charge and discharge in the standby state can be obtained, then using the eigenvalue of the obtained dQ / dV-V curve data of preset small-rate charge and discharge and substituting it into the second corresponding relationship to obtain the corresponding predicted battery capacity; If the dQ / dV-V curve data of preset small-rate charge and discharge in the standby state cannot be obtained, then obtaining the number of cycle data at each temperature and each charge-discharge current in the battery historical charge-discharge data, predicting the eigenvalue of the dQ / dV-V curve data based on the first corresponding relationship, and combining the second corresponding relationship to obtain the predicted battery capacity; Evaluating the state parameters of the battery pack based on the eigenvalue of the dQ / dV-V curve data and the predicted battery capacity.

2. The method according to claim 1, characterized in that The using a preset machine learning model to fit the first corresponding relationship between the eigenvalue of the dQ / dV-V curve data and temperature, charge-discharge current, and number of cycles at different temperatures, different charge-discharge currents, and different numbers of cycles, as well as the second corresponding relationship between battery capacity and the eigenvalue of the dQ / dV-V curve data includes: Obtaining the dQ / dV-V curves at different numbers of cycles, different temperatures, and different charge-discharge currents as the first sample data; Based on the first sample data, fitting the relationship between the eigenvalue of the dQ / dV-V curve data and the charge-discharge current, temperature, and number of cycles at different charge-discharge currents, temperatures, and numbers of cycles through the XGboost model as the first corresponding relationship; Obtaining the eigenvalue of the dQ / dV-V curve data at different battery capacities as the second sample data; Based on the first sample data or the second sample data, fitting the corresponding relationship between the battery capacity and the eigenvalue of the dQ / dV-V curve data through the XGboost model as the second corresponding relationship.

3. The method according to claim 1 or 2, characterized in that, The eigenvalue of the dQ / dV-V curve data includes: peak value, peak area, peak time, lower endpoint of the platform voltage interval, and upper endpoint of the platform voltage interval.

4. The method according to claim 3, wherein The evaluating the state parameters of the battery pack based on the predicted battery capacity includes: Based on the predicted battery capacity and the calculation models of the SOC and SOH of the battery pack, obtaining the SOC value and SOH value corresponding to the predicted battery capacity. When the SOC value and SOH value corresponding to the predicted battery capacity are closer to the actual values than the SOC value and SOH value provided by the battery management system, then correcting the SOC value and SOH value provided by the battery management system, and correcting the preset charge-discharge ammeter in the battery management system based on the predicted battery capacity.

5. The method according to claim 4, characterized in that Also including: Adjusting the SOC voltage interval and the battery equalization voltage interval according to the lower endpoint of the platform voltage interval and the upper endpoint of the predicted dQ / dV-V curve data.

6. The method according to claim 5, characterized in that, It further includes: Obtaining the peak voltage displacement change amount and the peak area change amount of the dQ / dV curve of each single battery in the battery pack, and when the peak voltage displacement change amount and / or the peak area change amount exceed the corresponding change thresholds, adjusting the used DOD interval and the upper and lower limit protection thresholds.

7. A battery pack state evaluation system, characterized in that, It includes: A corresponding relationship establishment module, configured to use a preset machine learning model to fit the first corresponding relationship between the dQ / dV-V curve data eigenvalues at different temperatures, different charge and discharge currents, and different cycle numbers and the temperature, charge and discharge current, and cycle number, and the second corresponding relationship between the battery capacity and the dQ / dV-V curve data eigenvalues; A data acquisition and judgment module, configured to judge whether the dQ / dV-V curve data of the battery pack to be measured after a preset running time can be obtained during preset small-rate charge and discharge in the standby state; A first prediction module, configured to, if the dQ / dV-V curve data of the preset small-rate charge and discharge in the standby state can be obtained, use the obtained dQ / dV-V curve data eigenvalues of the preset small-rate charge and discharge and substitute them into the second corresponding relationship to obtain the corresponding predicted battery capacity; A second prediction module, configured to, if the dQ / dV-V curve data of the preset small-rate charge and discharge in the standby state cannot be obtained, obtain the cycle number data at each temperature and each charge and discharge current in the battery historical charge and discharge data, predict the dQ / dV-V curve data eigenvalues based on the first corresponding relationship, and combine the second corresponding relationship to obtain the predicted battery capacity; A state evaluation module, configured to evaluate the battery pack state parameters based on the dQ / dV-V curve data eigenvalues and the predicted battery capacity.

8. A computer device, characterized in that, It includes: A memory and a processor, the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to execute the battery pack state evaluation method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, Computer instructions are stored on the computer-readable storage medium, and the computer instructions are used to cause a computer to execute the battery pack state evaluation method according to any one of claims 1-6.

10. A computer program product, characterized in that, It includes computer instructions, and the computer instructions are used to cause a computer to execute the battery pack state evaluation method according to any one of claims 1-6.

Images