Battery health degree evaluation method, electronic equipment and computer storage medium

Abstract

The invention discloses a battery health degree assessment method, electronic equipment and a computer storage medium, and the method comprises the steps: obtaining the battery static data of a battery, and inputting the battery static data into a preset health degree estimation model to obtain a health degree initial calculation value; and obtaining battery dynamic data of the battery, and inputting the battery dynamic data and the health degree initial calculation value into a preset health degree evaluation model to obtain the health degree of the battery. According to the method and the device, the health degree of the battery is measured and calculated in a multi-dimensional manner, the initial calculated value of the health degree is obtained based on the static data of the battery, and the health degree of the battery is obtained based on the dynamic data of the battery and the initial calculated value of the health degree, so that the health state of the battery can be evaluated more comprehensively and accurately. Besides, in the application, special equipment technical personnel do not need to be arranged to assist in evaluating the health degree of the battery on site, so that the cost of evaluating the health degree of the battery is greatly reduced, and the circulation safety of the battery in the market is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy batteries, and in particular to a battery health assessment method, electronic equipment and computer storage medium. Background Art

[0002] As the world pays more and more attention to environmental protection, new energy batteries are increasingly used, such as electric vehicles, energy storage systems, etc. However, the performance of new energy batteries is affected by many factors, such as temperature, number of charge and discharge cycles, battery aging, etc. Therefore, accurately assessing the health status of batteries is crucial for safe use, maintenance, and even calculation of battery asset profit and loss.

[0003] Existing methods for evaluating the health of new energy batteries usually measure basic parameters such as battery voltage and current, and then construct charge and discharge curves to evaluate the health status of the battery.

[0004] However, there are some problems with the above existing new energy battery health assessment method. The method only considers data of a single dimension, which may lead to inaccurate assessment results and fail to effectively ensure the safe use of the battery. Secondly, the existing new energy battery health assessment method often requires the assistance of specialized equipment technicians, which undoubtedly increases the cost of battery health assessment. Summary of the invention

[0005] The purpose of this application is to provide a battery health assessment method, electronic device and computer storage medium, aiming to solve the problem that the existing technology only considers data of a single dimension when performing battery health detection and assessment, and requires the deployment of specialized equipment technicians, resulting in inaccurate assessment results and high assessment costs.

[0006] To achieve the above objectives:

[0007] In a first aspect, an embodiment of the present application provides a battery health assessment method, the method comprising:

[0008] Acquire battery static data of the battery, and input the battery static data into a preset health estimation model to obtain an initial health calculation value;

[0009] The battery dynamic data of the battery is obtained, and the battery dynamic data and the initial calculated value of the health are input into a preset health evaluation model to obtain the health of the battery.

[0010] In one embodiment, inputting the battery static data into a preset health estimation model to obtain an initial health calculation value includes:

[0011] Acquire the charging depth from the battery static data, and acquire the charging capacity according to the charging depth;

[0012] Acquire the current battery capacity according to the charging depth and the charging capacity;

[0013] The current battery capacity and the initial battery capacity are input into the health estimation model to obtain the health initial calculation value.

[0014] In one embodiment, before inputting the battery dynamic data and the initial calculated value of the health into a preset health evaluation model to obtain the health of the battery, the method includes:

[0015] Obtain the historical battery dynamic data and historical health initial calculation value of the battery;

[0016] Preprocessing the historical battery dynamic data;

[0017] Performing feature extraction on the preprocessed historical battery dynamic data to obtain historical battery dynamic features;

[0018] The health evaluation model is obtained according to the historical battery dynamic characteristics and the historical initial health calculation value.

[0019] In one embodiment, the obtaining the health evaluation model according to the historical battery dynamic characteristics and the historical health initial calculation value includes:

[0020] Aggregating the historical battery dynamic characteristics and the historical health initial calculation values ​​according to a preset period to construct a training data set and a test data set;

[0021] Inputting the training data set into the model to be trained to obtain a model to be tested;

[0022] The test data set is input into the model to be tested to obtain the health assessment model.

[0023] In one embodiment, after inputting the battery dynamic data and the initial calculated value of the health into a preset health assessment model to obtain the health of the battery, the method further includes:

[0024] The health status is input into a preset life prediction model to obtain the remaining life of the battery.

[0025] In one embodiment, before inputting the health status into a preset life prediction model to obtain the remaining life of the battery, the method further includes:

[0026] Acquire historical performance data of the battery and a dynamic health corresponding to the historical performance data;

[0027] The life prediction model is obtained according to the historical usage performance data and the dynamic health degree.

[0028] In one embodiment, the acquiring the life prediction model according to the historical usage performance data and the dynamic health degree includes:

[0029] Obtaining at least two of the cumulative mileage, cumulative years, cumulative charging capacity, cumulative discharging capacity, and cumulative current and voltage from the historical performance data;

[0030] Constructing linear regression models of the dynamic health and the accumulated mileage, accumulated years, accumulated charging capacity, accumulated discharge capacity, and accumulated current and voltage respectively;

[0031] At least two linear regression models corresponding to the historical performance data are fitted to obtain the life prediction model.

[0032] In one embodiment, after inputting the battery dynamic data and the initial calculated value of the health into a preset health assessment model to obtain the health of the battery, the method further includes:

[0033] Corresponding alarm feedback is performed according to the score interval in which the health level is located.

[0034] In a second aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the battery health assessment method as described above when executing the computer program.

[0035] In a third aspect, an embodiment of the present application provides a computer storage medium, wherein the computer storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the battery health assessment method described above are implemented.

[0036] In this application, the battery health is measured in a multi-dimensional way, the initial calculated value of health is obtained based on the static data of the battery, and then the health of the battery is obtained based on the dynamic data of the battery and the initial calculated value of health, which can more comprehensively and accurately evaluate the health status of the battery. In addition, in this application, there is no need to equip special equipment technicians to assist in the battery health assessment, which greatly reduces the cost of battery health assessment and improves the safety of battery circulation in the market. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A flowchart of a battery health assessment method provided by an embodiment of the present invention.

[0038] Figure 2A schematic diagram of the structure of a monitoring system provided by an embodiment of the present invention.

[0039] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention.

[0040] Description of Figure Numbers:

[0041] 310. Processor; 311. Memory; 312. Network interface; 313. Bus system. DETAILED DESCRIPTION

[0042] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0043] It should be noted that, in this article, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element. In addition, components, features, and elements with the same name in different embodiments of the present application may have the same meaning or different meanings, and their specific meanings need to be determined by their explanation in the specific embodiment or further combined with the context of the specific embodiment.

[0044] It should be understood that, although the terms first, second, third, etc. may be used to describe various information in this article, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this article, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein can be interpreted as "at the time of..." or "when..." or "in response to determination". Furthermore, as used in this article, the singular forms "one", "one" and "the" are intended to also include plural forms, unless there is an opposite indication in the context. It should be further understood that the terms "comprising", "including" indicate that there are described features, steps, operations, elements, components, projects, kinds, and / or groups, but do not exclude the existence, occurrence or addition of one or more other features, steps, operations, elements, components, projects, kinds, and / or groups. The terms "or" and "and / or" used herein are interpreted as inclusive, or mean any one or any combination. Thus, “A, B, or C” or “A, B and / or C” means “any of the following: A; B; C; A and B; A and C; B and C; A, B, and C.” An exception to this definition will occur only when a combination of elements, functions, steps, or operations are inherently mutually exclusive in some manner.

[0045] It should be understood that, although the various steps in the flowchart in the embodiment of the present application are displayed in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and it can be performed in other orders. Moreover, at least a portion of the steps in the figure may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and their execution order is not necessarily performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0046] It should be noted that in this article, step codes such as S1 and S2 are used for the purpose of expressing the corresponding content more clearly and concisely, and do not constitute a substantial limitation on the order. When implementing the step, those skilled in the art may execute S2 first and then S1, etc., but these should all be within the scope of protection of this application.

[0047] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0048] In the subsequent description, the suffixes such as "module", "component" or "unit" used to represent elements are only used to facilitate the description of the present application, and have no specific meanings. Therefore, "module", "component" or "unit" can be used in a mixed manner.

[0049] The following is an explanation of the terms that may be involved in this application, as follows:

[0050] SOC (State of Charge), SOC indicates the ratio of the current remaining power of the battery to the total capacity of the battery, usually expressed as a percentage. It reflects the current degree of charge and discharge of the battery and is an important parameter used by the battery management system (BMS) to control the charge and discharge process. When the battery is fully charged, the SOC is 100%, and when the battery is fully discharged, the SOC is close to 0%.

[0051] SOH (State of Health), SOH indicates the state of the battery relative to a new battery, which reflects the health or performance level of the battery. This parameter covers many aspects of the battery, including but not limited to battery capacity, internal resistance, discharge efficiency, etc. The SOH of a new battery is defined as 100%, and as the battery ages, the SOH will gradually decrease. When the SOH drops to a certain level, the battery may no longer be suitable for use.

[0052] The OCV-SOC curve refers to the relationship curve between the open-circuit voltage (OCV) of the battery and the state of charge (SOC) of the battery. OCV refers to the voltage of the battery when no current flows, that is, the voltage of the battery in a completely static state. SOC is the percentage of the battery's current remaining power, usually expressed as a capacity relative to the battery's fully charged state. The OCV-SOC curve can reflect the battery's state of charge. For the battery management system (BMS), an accurate OCV-SOC relationship is crucial for estimating the battery's SOC.

[0053] XGBoost (eXtreme Gradient Boosting) is an efficient and scalable tree boosting system that is widely used in the field of machine learning, especially in various data science competitions and practical applications. It has achieved leading results.

[0054] Please refer to Figure 1 , Figure 1 A flow chart of a battery health assessment method is shown, the method comprising:

[0055] S1. Obtain the battery static data of the battery, and input the battery static data into the preset health estimation model to obtain the initial calculated value of health. Specifically, in this embodiment, the battery static data can be data generated when the vehicle is stationary, such as parking time, SOC or SOH when parked, parking time, battery voltage, current, temperature and other state parameters when parked, etc. On the other hand, it can also be battery attribute data, such as battery material type, battery rated capacity, battery rated kilowatt-hour, number of modules, number of parallel connections, number of series connections, number of monomers, years of use, number of charges, number of discharges, etc. It can also be physical and chemical properties of the battery, such as tap density, specific surface area, specific energy, voltage platform, cycle performance, transition metal content, raw material cost, environmental performance and safety performance.

[0056] Furthermore, by inputting the static data of the battery into the preset health estimation model, the initial calculated value of the health of the battery can be obtained, and the health of the battery can be preliminarily evaluated according to the initial calculated value of the health, and whether to execute step S2 is determined according to the evaluation result. For example, when the initial calculated value of the health does not reach the preset preliminary health qualification threshold, step S2 is not executed, otherwise, step S2 is executed.

[0057] S2. Obtain the battery dynamic data of the battery, and input the battery dynamic data and the initial calculated value of the health into a preset health assessment model to obtain the health of the battery. Specifically, in this embodiment, the health can be expressed as a percentage, which is used to evaluate the performance status or remaining life of the battery. On the one hand, the battery dynamic data can be data generated or recorded by the battery during the driving of the vehicle, such as driving time, mileage, average discharge current per 100 kilometers, minimum discharge temperature, total discharge time, kinetic energy recovery time, etc. On the other hand, it can also be data generated or recorded when the battery is charging, such as maximum charging current, minimum charging voltage, maximum charging temperature, SOC before and after charging, total charging time, electric energy obtained through kinetic energy recovery, etc.

[0058] Furthermore, the battery health is obtained by inputting the battery dynamic data and the initial calculated health value obtained from step S1 into the preset health assessment model. Exemplarily, the health can be divided into "healthy", "sub-healthy", "degraded", "severely degraded", etc. according to the level of quality. The monitoring system carried by the vehicle itself or the monitoring system of the battery swap station can upload the health generated each time to the cloud server. The maintenance personnel of the car brand or the operation and maintenance personnel of the battery swap station can follow up with the driver in real time according to the health, and quickly carry out targeted processing actions, effectively avoiding safety problems caused by battery failure, improving the safety of battery use, and helping enterprises accurately evaluate the profit and loss of battery assets.

[0059] By executing steps S1 to S2, first, the initial calculated value of health is obtained based on the static data of the battery, and then the health of the battery is obtained based on the dynamic data of the battery and the initial calculated value of health. Compared with the prior art that only cuts in from a single dimension, the present application scheme measures the health of the battery in a multi-dimensional manner, which can more comprehensively and accurately evaluate the health status of the battery. Secondly, the present application scheme does not require the on-site assistance of specialized equipment technicians, and all the equipment used is based on the equipment already installed in the vehicle or battery swap station, and there is no need to configure other monitoring equipment, which greatly reduces the cost of battery health assessment.

[0060] Optionally, in the present application, the "battery" may specifically be a power battery used to provide power and endurance for an electric vehicle, or a storage battery used to provide energy for low-power electrical equipment in the vehicle, and is not limited thereto.

[0061] Optionally, the subject that executes steps S1 to S2 may be, for example, a vehicle-mounted device or a monitoring system carried by the vehicle itself, which automatically executes steps S1 to S2 according to a preset monitoring cycle. For example, it may also be a monitoring system of a battery swap station, which automatically executes steps S1 to S2 when the vehicle enters the battery swap station and performs the battery swap process, which is not limited here.

[0062] In one embodiment, the step of inputting the battery static data into a preset health estimation model to obtain an initial health calculation value in step S1 includes:

[0063] S11. Obtain the depth of charge from the battery static data, and obtain the charge capacity according to the depth of charge. Specifically, the battery static data mainly includes the static data before and after the charging segment, and the static data before and after the charging segment mainly includes the parking time, parking SOC, parking time and temperature distribution characteristics, voltage, current, OCV-SOC curve at different temperatures and battery rated capacity, etc. In practical applications, it is necessary to measure the open circuit voltage of the battery at a certain moment. This is usually done after the battery has been stationary for a period of time to ensure that the voltage is stable and reflects the true OCV value. Therefore, in this embodiment, the voltage at the end of the static state is used as the estimated OCV.

[0064] In one embodiment, the depth of charge can be obtained by looking up a table. Specifically, according to the extracted charging start and end equivalent OCV voltage OCV i And the corresponding temperature T i , find the OCV-SOC curve CurveT at the temperature closest to the sampling temperature, and perform linear interpolation to obtain (OCV 1 and OCV 2 ) corresponding to the interpolation value (SOC 1 and SOC 2 ). Among them, the charging depth ΔSOC is calculated by the following formula:

[0065] ΔSOC=SOC 2 -SOC 1 ;

[0066] Furthermore, based on the above results, the charging segments in which the charging depth ΔSOC is greater than the charging depth threshold and there is no data interruption or loss in the charging process are screened, and the charging capacity ΔQ is calculated using the ampere-hour integral. The charging capacity ΔQ is calculated using the following formula:

[0067]

[0068] Where I(t) is the current at time t.

[0069] S12, obtaining the current battery capacity according to the charging depth and the charging capacity. Specifically, in this embodiment, based on the charging depth and the charging capacity obtained in step S11, the current battery capacity is calculated by the following formula:

[0070]

[0071] S13, input the current battery capacity and the initial battery capacity into the health estimation model to obtain the initial health calculation value. Specifically, in this embodiment, the current battery capacity and the initial battery capacity of the battery obtained in step S12 are input into the health estimation model to obtain the initial health calculation value. The health estimation model is specifically the following calculation formula:

[0072]

[0073] Among them, SOH is the initial calculated value of health, Q is the current battery capacity, Q ini is the initial battery capacity.

[0074] Furthermore, in the process of executing steps S11 to S13, since the static integration method is open-loop, it assumes that the voltage-capacity curve of the battery is a linear relationship, but the voltage-capacity curve of the battery is usually nonlinear, and errors may accumulate over time. Therefore, after step S13, after the battery is discharged to a certain depth, it is necessary to perform regular calibration, such as using the open circuit voltage method or the Kalman filter method, to improve the accuracy of the estimation.

[0075] In one embodiment, before inputting the battery dynamic data and the initial calculated value of the health into a preset health assessment model to obtain the health of the battery in step S2, the following steps are included:

[0076] S21, obtaining historical battery dynamic data and historical health initial calculation value of the battery. Specifically, in this embodiment, the historical battery dynamic data mainly includes two data collected by the vehicle-mounted equipment or the battery monitoring system or the battery swap station, and the two data are the basic data of battery discharge during driving and the basic data of battery parking and charging.

[0077] Among them, the basic data of discharge during driving mainly include maximum discharge current, average discharge current per 100 kilometers, minimum discharge SOC, maximum discharge temperature, minimum discharge temperature, discharge time, etc. The basic data of battery parking charging mainly include charging power, maximum charging current, minimum charging current, maximum charging voltage, minimum charging voltage, maximum charging SOC, minimum charging SOC, maximum charging temperature, minimum charging temperature, SOC before charging, SOC after charging, charging time, etc.

[0078] In another embodiment, the historical battery dynamic data may also include battery static attribute data. The battery static attribute data mainly includes battery material type, battery rated capacity, battery rated kilowatt-hour, number of modules, number of parallel connections, number of series connections, number of cells, years of use, number of charges, number of discharges, etc.

[0079] S22, preprocessing the historical battery dynamic data. Specifically, in this embodiment, the historical battery dynamic data obtained in step S21 is cleaned, missing values ​​and abnormal values ​​are processed, feature selection, feature conversion and other preprocessing methods are performed to ensure the integrity and high quality of the data.

[0080] Among them, the preprocessing specifically includes: step 1, mean smoothing the outliers in the historical battery dynamic data to reduce the impact of outliers on the overall data; step 2, zero-filling the missing values ​​in the historical battery dynamic data to ensure the integrity of the overall data; step 3, normalizing the time series data information in the historical battery dynamic data to make the data features have similar scales, which helps the model better understand the weights of each data feature; step 4, aligning the historical battery dynamic data according to time to reduce the noise caused by inconsistent data collection time and improve data quality.

[0081] S23, extracting features from the preprocessed historical battery dynamic data to obtain historical battery dynamic features. Specifically, in this embodiment, by extracting features from the preprocessed historical battery dynamic data to obtain historical battery dynamic features, this process can reduce the impact of noise and outliers on the model, improve the robustness of the model and the operational efficiency of the algorithm. Among them, the principal component analysis method, the independent component analysis method and other methods can be used for feature extraction.

[0082] S24. Obtain a health assessment model based on historical battery dynamic characteristics and historical initial health calculation values. Specifically, in this embodiment, the historical battery dynamic characteristics and historical initial health calculation values ​​are used as the material for model training. After multiple rounds of cyclic training, when ① the model converges, and / or the number of training times reaches a preset number threshold; ② the accuracy of the model generation result reaches a preset accuracy threshold, the trained model is determined as the health assessment model. Among them, based on the extracted features, an algorithm model that can accurately measure the health of the battery is constructed. Exemplarily, the algorithm model can be constructed using machine learning methods such as logistic regression, support vector machine, and neural network.

[0083] In one embodiment, the step of obtaining a health evaluation model according to historical battery dynamic characteristics and historical initial health calculation values ​​in step S24 includes:

[0084] S241. Aggregate historical battery dynamic characteristics and historical initial calculated values ​​of health according to a preset cycle to construct a training data set and a test data set. Specifically, in this embodiment, the historical battery dynamic characteristics include battery driving process characteristics and battery charging characteristics. Among them, the battery driving process characteristics include maximum discharge current, average discharge current per 100 kilometers, minimum discharge SOC, maximum discharge temperature, minimum discharge temperature, discharge time, etc. Battery charging characteristics include charging power, maximum charging current, minimum charging current, maximum charging voltage, minimum charging voltage, maximum charging SOC, minimum charging SOC, maximum charging temperature, minimum charging temperature, SOC before charging, SOC after charging, charging time, etc.

[0085] Furthermore, the preset period can be, for example, every minute, every hour, every day, every week, every month, etc. In this embodiment, the preset period is preferably every day. Considering the problem of data time alignment, the daily statistical historical battery dynamic features and historical health initial calculation values ​​are aggregated by day, and the training data set and the test data set are constructed according to the preset ratio. The cumulative features are used to calculate the latest daily features, and the health SOH is used to calculate the average SOH value within each day. The features and label data are aligned by date association.

[0086] Exemplarily, constructing historical battery dynamic features may include:

[0087] Charging start SOC range: [[0,20), [20,40), [40,60), [60,80), [80,100];

[0088] Average charging temperature range: [(,-10),[-10,10),[10,30),[30,)];

[0089] Average charging rate range: [[0,0.4),[0.4,0.8),[0.8,1.2),[1.2,)].

[0090] S242, input the training data set into the model to be trained to obtain the model to be tested. Specifically, in this embodiment, the XGBoost model is used as the model to be trained, and the training data set is input into the model to be trained as training material for repeated training, thereby obtaining the model to be tested.

[0091] S243, input the test data set into the model to be tested to obtain a health assessment model. Specifically, in this embodiment, the test data set is input into the model to be tested, and when the accuracy of the output result is greater than a preset accuracy threshold, the model to be tested is determined as a health assessment model.

[0092] In actual implementation, by inputting the current initial calculated value of health into the trained health assessment model, the health status (SOH) of the battery of the current vehicle will be automatically output and recorded and saved. The health status of the battery is obtained by the following formula:

[0093] SOH predict =f soh_model (X1,X2,…,Xn);

[0094] Furthermore, in actual project applications, considering the increasing number of connected batteries, the continuous enrichment of training data sets and the more significant differences in the dynamic use of each battery, in order to improve the prediction accuracy of the model, the model can be trained and updated according to the system operation time cycle. At the same time, taking into account the impact of data volume and training time, the training frequency can be increased in the early stage of model launch, and the training frequency can be reduced after the number of batteries increases and the model accuracy tends to stabilize.

[0095] In one embodiment, after inputting the battery dynamic data and the initial calculated value of the health into a preset health evaluation model to obtain the health of the battery in step S2, the method further includes:

[0096] S3. Input the health status into a preset life prediction model to obtain the remaining life of the battery. Specifically, in this embodiment, after obtaining the health status of the battery in step S2, the remaining life of the battery can be obtained by inputting it into the life prediction model. In this way, on the one hand, it is possible to predict when the battery needs to be replaced, thereby avoiding system interruptions caused by sudden battery failure. On the other hand, it can reduce unnecessary battery replacements, help operators better plan power resources, improve energy efficiency, and reduce maintenance costs.

[0097] In one embodiment, before inputting the health status into a preset life prediction model to obtain the remaining life of the battery in step S3, the method further includes:

[0098] S31. Obtain the historical performance data of the battery and the dynamic health corresponding to the historical performance data. Specifically, in this embodiment, since the above health assessment model is a model trained for multi-dimensional characteristics of a group, it is not suitable for the life prediction of a single battery. Therefore, it is considered to simplify the model and construct a battery remaining life regression prediction model based on the historical aging trajectory data of a single battery, that is, the historical performance data, in which the battery SOH changes with the service life assessment index. Among them, the historical performance may include years of use, cumulative mileage, cumulative number of charge / discharge cycles, charging current / voltage, etc.

[0099] Furthermore, the dynamic life EOL curve of the battery health can be obtained through the health assessment model. In this embodiment, a linear regression model is constructed based on the dynamic life EOL curve of the battery health, and then the linear change relationship between the health and the life assessment indicators (service life, cumulative mileage, cumulative number of charge / discharge cycles, charging current / voltage) is obtained, and then the dynamic health corresponding to the historical usage performance data is obtained.

[0100] In one implementation, taking SOH=80% as the EOL value of the single cell remaining life, the corresponding EOL life can be predicted in real time. As the battery ages, its historical trajectory data gradually accumulates, and the accuracy of life prediction also increases.

[0101] S32, obtaining a life prediction model based on historical performance data and dynamic health. Specifically, in this embodiment, a linear regression model is used as a training model for the life prediction model, and the historical performance data and the corresponding dynamic health are input into the training model as training materials to obtain the life prediction model.

[0102] In one embodiment, the step of obtaining a life prediction model based on historical performance data and dynamic health in step S32 includes:

[0103] S321, obtaining at least two of the cumulative mileage, cumulative years, cumulative charging capacity, cumulative discharge capacity, and cumulative current and voltage from the historical performance data. Specifically, in this embodiment, the at least two may include any two, any three, any four, or all of the five groups of data.

[0104] S322, respectively construct linear regression models of dynamic health and cumulative mileage, cumulative years, cumulative charging capacity, cumulative discharge capacity, and cumulative current and voltage. Specifically, in this embodiment, based on the dynamic record of battery historical SOH battery health, that is, dynamic health and cumulative mileage, cumulative years, cumulative charging capacity, cumulative discharge capacity, and cumulative current and voltage are respectively constructed to fit the changing rules of various life assessment indicators and health.

[0105] S323, fitting the linear regression models corresponding to at least two historical performance data to obtain a life prediction model. Specifically, in this embodiment, the linear regression models corresponding to the at least two historical performance data selected in step S321 are fitted to obtain a life prediction model.

[0106] Exemplarily, steps S321 to S323 may be implemented in the following manner:

[0107] 1. Build a linear regression model based on historical performance data

[0108] There is a set of historical performance data that records the battery life evaluation indicators (such as mileage, age, number of charge and discharge cycles) and the corresponding battery health (SOH). A linear model is established to link the life indicator (x) with SOH (y).

[0109] 1. Collect the mileage, age, number of charge and discharge cycles and corresponding SOH value of a group of single batteries.

[0110] 2. Select the linear regression model y=kx+b, where k is the slope and b is the intercept.

[0111] 3. Train the model using historical performance data to find the best-fitting k and b values.

[0112] 4. Set upper and lower limits for the slope k to ensure that the model's predictions are within a reasonable range.

[0113] 2. Obtain the dynamic life history curve of battery health

[0114] 1. From the time the battery leaves the factory, record the SOH value and life evaluation indicators (accumulated mileage, accumulated years, accumulated charging capacity, accumulated discharge capacity, accumulated current and voltage) daily.

[0115] 2. Plot the daily SOH value and life assessment index into a curve to observe the changes in battery health over time.

[0116] 3. Linear regression model for training life evaluation indicators

[0117] 1. Select the life assessment indicators (accumulated mileage, accumulated years, accumulated charging capacity, accumulated discharge capacity, accumulated current and voltage).

[0118] 2. For each indicator, use the dynamic health (historical SOH data) to train a linear regression model to fit the changing pattern of life assessment indicators and health.

[0119] 4. Predicting the remaining battery life

[0120] 1. Input the current health of the battery into the trained life prediction model.

[0121] 2. The life prediction model outputs the life indicator corresponding to the current health status.

[0122] 3. Set the battery EOL as the life indicator when the SOH drops to 80%.

[0123] 4. Subtract the EOL indicator from the current life indicator to get the remaining RUL life of the battery.

[0124] V. Evaluation of the detection model effect

[0125] When evaluating the effectiveness of a prediction model, we usually focus on aspects such as the model’s accuracy, stability, and practicality. Common evaluation indicators include:

[0126] 1. Accuracy:

[0127] Definition: The ratio of the number of samples predicted correctly to the total number of samples.

[0128] Formula: Accuracy = (TP + TN) / (TP + TN + FP + FN).

[0129] Among them, TP (true positive example), TN (true negative example), FP (false positive example), and FN (false negative example).

[0130] 2. Precision:

[0131] Definition: The proportion of samples predicted to be positive that are actually positive.

[0132] Formula: Precision==TP / (TP+FP).

[0133] 3. Recall:

[0134] Definition: The proportion of samples that are correctly predicted to be positive among all samples that are actually positive.

[0135] Formula: Recall = TP / (TP+FN).

[0136] In one embodiment, it is determined whether any one of the precision, recall and F1 score (accuracy) meets the preset accuracy threshold. If not, it indicates that the model has not yet reached the expected performance level and needs further improvement. The model can be retrained by the method performed in step S97 to improve the training effect of the model, and finally a battery swap station health assessment model that can reach the accuracy threshold is obtained.

[0137] In another embodiment, the accuracy of the model can also be determined by:

[0138] (1)mse=mean_squared_error(y_test,y_pred);

[0139] The mean squared error (MSE) is used to calculate the mean squared error between the true value (`y_test`) and the predicted value (`y_pred`). It represents the average of the squares of the difference between the predicted value and the true value, and can measure the error degree of the predicted value. The smaller the mse, the more accurate the model's prediction.

[0140] (2)r 2 =r 2 _score(y_test,y_pred);

[0141] The R-squared value is also called the coefficient of determination. It is used to measure the accuracy of model predictions. It indicates how much of the variation in the dependent variable can be explained by the independent variable. The R-squared value ranges from 0 to 1. The closer the value is to 1, the stronger the explanatory power of the model and the more accurate the prediction.

[0142] (3)print(f'Mean Squared Error:{mse:.2f}'); print(f'R-squared:{r2:.2f}');

[0143] The above two lines of code correspond to the output results of mean square error (mse) and R square value respectively. `:.2f` is a format string, which means to retain two decimal places.

[0144] In summary, the above two lines of code are used to calculate and print out the mean square error (mse) and R-squared value of the model. These two indicators can intuitively obtain the accuracy and explanatory power of the model prediction.

[0145] Based on the establishment and specific application of the evaluation model described in the above embodiments, the current prediction accuracy of battery swap stations can reach 90%, the recall rate can reach 90%, and the prediction value accuracy can reach 85%. It has been widely used in the production of new energy battery health and remaining life detection.

[0146] In one embodiment, after inputting the battery dynamic data and the initial calculated value of the health into a preset health evaluation model to obtain the health of the battery in step S1, the method further includes:

[0147] The corresponding alarm feedback is executed according to the score interval in which the health is located. Specifically, in this embodiment, a set of quantifiable health level tables is first established. After the health of the battery is obtained in this round of detection cycle, the on-board equipment or the monitoring system carried by the vehicle itself or the monitoring system of the battery swap station will automatically match the score corresponding to the health with the health level table, and after confirming the score interval in which the health is located, the alarm feedback corresponding to the score interval is executed.

[0148] In one embodiment, in the health level table, the health level can be divided into four levels according to the level of quality, namely "healthy", "sub-healthy", "degraded", and "severely degraded", and the score ranges corresponding to the four levels are [90,95), [80,90), [60,80), and [0,60), and the alarm feedback corresponding to the four score ranges are "APP prompts alarm text", "Initiate work order and manual intervention", "SMS reminder + APP prompts alarm text + initiates work order", and "Telephone alarm on-duty personnel for emergency processing, such as immediate visit to the site for inspection and processing".

[0149] Please refer to Figure 2 , Figure 2 The schematic diagram of the structure of a monitoring system is shown. The monitoring system can be installed in a car or a battery swap station. The basic data layer is used to store all data information of the battery, mainly including basic data of the parked battery, basic data of the battery parking charging, basic data of the battery recharge, basic data of the battery driving discharge, and static attribute data of the battery. The data processing layer is used to pre-process the data in the basic data layer to ensure the integrity and high quality of the data. The basic feature engineering layer is used to extract features from the pre-processed data information. The extracted features include but are not limited to the following: Figure 2 The basic characteristics of battery materials, battery driving process characteristics, charging characteristic data and parking characteristics are displayed.

[0150] The single cell battery health estimation model includes a health SOH calculation model based on the static integration method and a single cell battery health dynamic estimation model SOH prediction model. This model is mainly used to estimate the SOH state in the cloud, that is, to obtain the initial calculated value of the battery health to evaluate the current battery consumption and aging.

[0151] The single cell battery health dynamic estimation model uses historical data to train the relationship between the health status SOH and the operating condition characteristics related to consumption and aging, so as to achieve the purpose of outputting the current health status SOH under operating conditions that cannot be calculated by the physical model.

[0152] The single cell remaining life EOL model, the life prediction model is a data-driven regression prediction model, which is used to regress and predict the EOL (the state where the battery reaches the end of its service life) end life according to the trajectory of the health status SOH changing with the life index, and output the current battery remaining life score.

[0153] In this way, the entire monitoring system can collect and analyze the battery data in real time when it is working or not working. Through the collaboration between the above model layers, it can promptly discover and deal with potential battery problems, effectively improving the stability and reliability of the battery.

[0154] Based on the same inventive concept as the above embodiments, an embodiment of the present invention provides an electronic device, such as Figure 3 As shown, the device includes: a processor 310 and a memory 311 storing a computer program; wherein, Figure 3 The processor 310 shown in the figure is not used to indicate that the number of the processor 310 is one, but is only used to indicate the positional relationship of the processor 310 relative to other devices. In actual applications, the number of the processor 310 may be one or more; similarly, Figure 3 The memory 311 shown in the figure has the same meaning, that is, it is only used to refer to the position relationship of the memory 311 relative to other devices. In practical applications, the number of memories 311 can be one or more. When the processor 310 runs the computer program, the method applied to the above device is implemented.

[0155] The device may also include: at least one network interface 312. The various components in the device are coupled together via a bus system 313. It is understood that the bus system 313 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 313 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, Figure 3 Various buses are labeled as bus system 313.

[0156] The memory 311 may be a volatile memory or a nonvolatile memory, or may include both volatile and nonvolatile memories. The nonvolatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disk, or a compact disc read-only memory (CD-ROM); the magnetic surface memory may be a disk memory or a tape memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), synchronous static random access memory (SSRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM, SyncLink Dynamic Random Access Memory), and direct RAM bus random access memory (DRRAM, Direct Rambus Random Access Memory).The memory 311 described in the embodiments of the present invention is intended to include but is not limited to these and any other suitable types of memories.

[0157] The memory 311 in the embodiment of the present invention is used to store various types of data to support the operation of the device. Examples of these data include: any computer program used to operate on the device, such as an operating system and an application; contact data; phone book data; messages; pictures; videos, etc. Among them, the operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., which are used to implement various basic services and process hardware-based tasks. The application program may include various applications, such as a media player (Media Player), a browser (Browser), etc., which are used to implement various application services. Here, the program that implements the method of the embodiment of the present invention may be included in the application program.

[0158] Based on the same inventive concept as the above-mentioned embodiment, this embodiment further provides a computer-readable storage medium, in which a computer program is stored. The computer-readable storage medium may be a ferromagnetic random access memory (FRAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); it may also be various devices including one or any combination of the above memories, such as a mobile phone, a computer, a tablet device, a personal digital assistant, etc. When the computer program stored in the computer-readable storage medium is executed by the processor, the above method is implemented. For the specific steps implemented when the computer program is executed by the processor, please refer to Figure 1 The description of the illustrated embodiment will not be repeated here.

[0159] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0160] In this document, the terms "comprises," "comprising," or any other variations thereof, are intended to cover a non-exclusive inclusion of elements other than those listed and may also include additional elements not expressly listed.

[0161] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A battery health assessment method, characterized in that: The method comprises: Acquire battery static data of the battery, and input the battery static data into a preset health estimation model to obtain an initial health calculation value; The battery dynamic data of the battery is obtained, and the battery dynamic data and the initial calculated value of the health are input into a preset health evaluation model to obtain the health of the battery.

2. The method according to claim 1, characterized in that The step of inputting the battery static data into a preset health estimation model to obtain an initial health calculation value includes: Acquire the charging depth from the battery static data, and acquire the charging capacity according to the charging depth; Acquire the current battery capacity according to the charging depth and the charging capacity; The current battery capacity and the initial battery capacity are input into the health estimation model to obtain the health initial calculation value.

3. The method according to claim 1, characterized in that Before inputting the battery dynamic data and the initial calculated value of the health into a preset health evaluation model to obtain the health of the battery, the method includes: Obtain the historical battery dynamic data and historical health initial calculation value of the battery; Preprocessing the historical battery dynamic data; Performing feature extraction on the preprocessed historical battery dynamic data to obtain historical battery dynamic features; The health evaluation model is obtained according to the historical battery dynamic characteristics and the historical initial health calculation value.

4. The method according to claim 3, characterized in that The obtaining the health evaluation model according to the historical battery dynamic characteristics and the historical health initial calculation value includes: Aggregating the historical battery dynamic characteristics and the historical health initial calculation values ​​according to a preset period to construct a training data set and a test data set; Inputting the training data set into the model to be trained to obtain a model to be tested; The test data set is input into the model to be tested to obtain the health assessment model.

5. The method according to claim 1, characterized in that After inputting the battery dynamic data and the initial calculated value of the health into a preset health evaluation model to obtain the health of the battery, the method further includes: The health status is input into a preset life prediction model to obtain the remaining life of the battery.

6. The method according to claim 5, characterized in that Before inputting the health status into a preset life prediction model to obtain the remaining life of the battery, the method further includes: Acquire historical performance data of the battery and a dynamic health corresponding to the historical performance data; The life prediction model is obtained according to the historical usage performance data and the dynamic health degree.

7. The method according to claim 6, characterized in that The obtaining the life prediction model according to the historical usage performance data and the dynamic health degree includes: Obtaining at least two of the cumulative mileage, cumulative years, cumulative charging capacity, cumulative discharging capacity, and cumulative current and voltage from the historical performance data; Constructing linear regression models of the dynamic health and the accumulated mileage, accumulated years, accumulated charging capacity, accumulated discharge capacity, and accumulated current and voltage respectively; At least two linear regression models corresponding to the historical performance data are fitted to obtain the life prediction model.

8. The method according to claim 1, characterized in that After inputting the battery dynamic data and the initial calculated value of the health into a preset health evaluation model to obtain the health of the battery, the method further includes: Corresponding alarm feedback is performed according to the score interval in which the health level is located.

9. An electronic device, characterized in that: The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the battery health assessment method as claimed in any one of claims 1 to 8 when executing the computer program.

10. A computer storage medium, characterized in that: The computer storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the battery health assessment method according to any one of claims 1 to 8 are implemented.

Images