A smart battery health status prediction method based on big data and related devices

Abstract

The present invention belongs to the field of battery technology and provides a smart battery health status prediction method based on big data and related devices. The present invention can accurately predict the health status of the battery by extracting the health characteristics and capacity of qualified charging segments. The present invention is suitable for estimating the health status of the battery in real vehicles, realizing the gradient utilization of the battery, and helping users adjust their usage habits, alleviating users' mileage anxiety, extending the battery's service life and improving the battery's safety, and further providing a reference for the safe operation of the battery.

Description

Technical Field

[0001] The present invention relates to the field of battery technology, and in particular to a smart battery health status prediction method based on big data. Background Art

[0002] Accurately estimating and predicting the health status of lithium-ion batteries is crucial for the management of the battery's entire life cycle and the realization of battery gradient utilization. However, most current technologies are based on well-controlled laboratory environments and do not take into account the complexity of actual operating modes and the inconsistency of battery cells. There is an urgent need for a technical solution that can accurately estimate the health status of batteries based on the actual operating data of the battery, thereby achieving gradient utilization of batteries, helping users adjust their usage habits, extending battery life, and improving battery safety. Summary of the invention

[0003] In view of the defects in the prior art, the present invention provides a smart battery health status prediction method and related devices based on big data to achieve gradient utilization of batteries, help users adjust their usage habits, extend battery life and improve battery safety.

[0004] A first aspect of an embodiment of the present application provides a method for estimating the health status of a smart battery based on big data, characterized by comprising:

[0005] According to the collected charging data of the m original charging segments, an original charging data set is obtained;

[0006] According to a preset SOC screening condition, the original charging data in the original charging data set is screened to obtain a first charging data set, where the first charging data set includes k first charging segments;

[0007] Screening the first charging data in the first charging data set to obtain o health data;

[0008] The GAM fitting method is used to fit o health data to obtain the battery health decay curve;

[0009] Extracting health features from the original charging data in the original charging data set according to the battery health decay curve to obtain health feature data;

[0010] Based on the Pearson correlation analysis method, the health characteristic data were screened to obtain the correlation characteristics;

[0011] Normalizing the correlation features to obtain normalized feature data;

[0012] According to the normalized feature data and the battery health decay curve, the LSTM model is optimized by the Bayesian optimization method to obtain the battery health prediction model;

[0013] The battery health prediction model is used to predict the health of the smart battery, and the health prediction result of the smart battery is obtained.

[0014] Furthermore, the preset SOC screening conditions include preset SOC span screening conditions, preset SOC growth screening conditions and preset SOC continuity screening conditions.

[0015] Further, the first charging data in the first charging data set is screened to obtain o health data, including:

[0016] Calculate k battery maximum capacity data according to first charging data of k first charging segments in the first charging data set;

[0017] Use the box method to filter out the abnormal capacity data in the k battery maximum capacity data to obtain o battery maximum capacity data;

[0018] According to the o maximum capacity data of the batteries and the calibrated capacity of the smart battery, the health of the smart battery is evaluated to obtain o health data.

[0019] Furthermore, k pieces of battery maximum capacity data are calculated based on the first charging data of k first charging segments in the first charging data set, including:

[0020] The method shown in the following formula is used to calculate k battery maximum capacity data according to the first charging data of k first charging segments in the first charging data set:

[0021]

[0022] Among them, C i represents the maximum battery capacity corresponding to the i-th first charging segment, Q c (i) represents all the electricity passed through the i-th first charging segment, ΔSOC(i) represents the SOC change value of the i-th first charging segment, t2 and t1 represent the charging cut-off time and the charging start time respectively, I(t) represents the charging current at the current moment, SOC(t2) and SOC(t1) represent the SOC values ​​corresponding to the charging cut-off time t2 and the charging start time t1 respectively.

[0023] Furthermore, the health characteristic data includes battery accumulation characteristics, battery operation characteristics, and battery consistency characteristics.

[0024] A second aspect of an embodiment of the present application provides a smart battery health status prediction device based on big data, characterized in that it includes a collection unit, a first screening unit, a second screening unit, a fitting unit, a feature extraction unit, a third screening unit, a normalization processing unit, an optimization unit and a prediction unit, wherein:

[0025] A collection unit, used to obtain an original charging data set according to the collected charging data of the m original charging segments;

[0026] A first screening unit, configured to screen the original charging data in the original charging data set according to a preset SOC screening condition to obtain a first charging data set, wherein the first charging data set includes k first charging segments;

[0027] A second screening unit, configured to screen the first charging data in the first charging data set to obtain o health data;

[0028] A fitting unit, used to fit o health data using a GAM fitting method to obtain a battery health decay curve;

[0029] A feature extraction unit, used for extracting health features from the original charging data in the original charging data set according to the battery health decay curve to obtain health feature data;

[0030] The third screening unit is used to screen the health characteristic data based on the Pearson correlation analysis method to obtain correlation characteristics;

[0031] A normalization processing unit, used for normalizing the correlation features to obtain normalized feature data;

[0032] An optimization unit is used to optimize the LSTM model by using the Bayesian optimization method according to the normalized feature data and the battery health decay curve to obtain a battery health prediction model;

[0033] The estimation unit is used to estimate the health of the smart battery by using a battery health estimation model to obtain a health estimation result of the smart battery.

[0034] A third aspect of an embodiment of the present application provides a terminal, comprising a processor, an input device, an output device and a memory, wherein the processor, input device, output device and memory are interconnected, wherein the memory is used to store a computer program, the computer program comprises program instructions, and the processor is configured to call the program instructions to execute the step instructions of a method for predicting the health status of a smart battery based on big data in the first aspect of an embodiment of the present application.

[0035] The fourth aspect of the embodiments of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program for electronic data exchange, wherein the computer program enables the computer to execute part or all of the steps described in the method for predicting the health status of a smart battery based on big data in the first aspect of the embodiments of the present application.

[0036] A fifth aspect of the embodiments of the present application provides a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to execute some or all of the steps described in the first aspect of the embodiments of the present application. The computer program product may be a software installation package. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the specific embodiments of the present invention, the following will briefly introduce the drawings required for the specific embodiments or the prior art description. In all the drawings, each element or part is not necessarily drawn according to the actual scale.

[0038] Figure 1 A flowchart of a method for estimating the health status of a smart battery based on big data provided by an embodiment of the present invention;

[0039] Figure 2 A schematic diagram of raw charging data provided by an embodiment of the present invention;

[0040] Figure 3 A schematic diagram of a battery health decay curve provided by an embodiment of the present invention;

[0041] Figure 4 A framework diagram of an LSTM model provided for an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The following embodiments of the technical solution of the present invention are described in detail in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and are therefore only used as examples, and cannot be used to limit the protection scope of the present invention.

[0043] It should be noted that, unless otherwise specified, the technical terms or scientific terms used in this application should have the common meanings understood by those skilled in the art to which the present invention belongs.

[0044] like Figure 1 As shown, in one embodiment, a smart battery health status prediction method based on big data is provided, including:

[0045] 1. Obtain an original charging data set based on the collected original charging data of the m original charging segments;

[0046] Specifically, by deploying data collection terminals, the platform can monitor various charging data of smart batteries in real time. Figure 2 As shown, each original charging segment in the original charging data set contains multiple charging data, each charging data includes time, longitude, latitude, altitude, SOC (remaining battery capacity), charge and discharge status, current, total voltage, all monomer temperatures, all single probe temperatures, etc. These charging data are collected in real time and saved in the database for subsequent analysis and processing.

[0047] The present invention obtains an original charging data set based on the charging data of m original charging segments collected by the data collection terminal. Each original charging segment in the original charging data set includes the SOC value corresponding to the start of charging of the smart battery and the SOC value corresponding to the end of charging. The unit of SOC (remaining battery power) is percentage (%), which represents the current power state or charging progress of the battery.

[0048] 2. Filtering the original charging data of the original charging data set according to a preset SOC filtering condition to obtain a first charging data set, where the first charging data set includes k first charging segments;

[0049] Due to uncertainty factors such as measurement errors in the sensors of the data acquisition terminal, internal faults in the battery, and errors in the SOC measurement, the battery capacity of the collected smart battery may have noise and deviations. Therefore, it is necessary to filter the original charging data in the original charging data set according to the preset SOC filtering conditions to obtain the first charging data set.

[0050] According to the preset SOC screening condition, the steps for screening the raw charging data in the raw charging data set are as follows:

[0051] The preset SOC screening conditions include a preset SOC span screening condition, a preset SOC growth screening condition, and a preset SOC continuity screening condition.

[0052] Preferably, the preset SOC span screening condition means that only the original charging segments with sufficient span of SOC(t1) below 40% and SOC(t2) above 70% are retained, and the remaining original charging segments that do not meet the preset SOC span screening condition are eliminated.

[0053] Among them, SOC(t1) represents the SOC value corresponding to when the smart battery starts charging, and SOC(t2) represents the SOC value corresponding to when the smart battery ends charging.

[0054] Preferably, the preset SOC growth screening condition is to remove the original charging data of the same original charging segment whose unit SOC charging power is greater than the charging power threshold, and only leave the original charging data whose unit SOC charging power is less than or equal to the charging power threshold. Figure 2 As shown, assuming that the SOC value includes 18 pieces of original charging data (each row is a piece of charging data) when the SOC value is 35, the unit SOC charging power is the sum of the currents passing through the 18 pieces of original charging data.

[0055] Specifically, the unit SOC charge is calculated using the method shown in the following formula:

[0056]

[0057] △C is the amount of power charged in the n pieces of original charging data with the same SOC value in the current original charging segment, I i is the current passing through the i-th raw charging data, where n is the number of raw charging data contained in the same SOC value, and t is the data collection time interval.

[0058] Preferably, the charging power threshold is 3000.

[0059] Preferably, the preset SOC continuous screening condition means eliminating the charging data whose charging time interval between two adjacent original charging data in the same original charging segment is greater than 2 minutes, and leaving the charging data whose charging time interval between two adjacent charging data in the original charging segment is less than or equal to 2 minutes.

[0060] Specifically, the charging time interval T of two adjacent charging data in the same original charging segment is expressed by the method shown in the following formula: X -T X-1 <120s, where T X is the current time corresponding to the charging data in row x, T X-1 The current time corresponding to the charging data in row x-1.

[0061] 3. Screen the first charging data in the first charging data set to obtain o health data;

[0062] According to the first charging data of k first charging segments in the first charging data set, o health data are obtained, including:

[0063] (1) calculating k battery maximum capacity data according to first charging data of k first charging segments in the first charging data set;

[0064] The first charging data set includes k first charging segments, and it is known that k≤m. Each first charging segment includes a plurality of first charging data.

[0065] Use the method shown in the following formula to calculate k battery maximum capacity data based on the first charging data of k first charging segments in the first charging dataset:

[0066]

[0067] where C i represents the maximum capacity of the battery corresponding to the i-th first charging segment, Q c (i) represents all the electricity passed through the i-th first charging segment, ΔSOC(i) represents the SOC change value of the i-th first charging segment, t2 and t1 respectively represent the charging cut-off time and the charging start time, and I(t) represents the charging current at the current moment. SOC(t2) and SOC(t1) respectively represent the SOC values corresponding to the charging cut-off time t2 and the charging start time t1.

[0068] (2) Use the box plot method to screen out the abnormal capacity data in the k battery maximum capacity data to obtain o battery maximum capacity data;

[0069] Specifically, use the box plot method to quickly screen out the abnormal capacity data in the k battery maximum capacity data to obtain o battery maximum capacity data: when C i < Q1 - 1.5IQR or C i > Q3 + 1.5IQR, this data is abnormal capacity data.

[0070] Q1 is the first quartile, equal to the 25% number after arranging all the data values in the k battery maximum capacity data from small to large. Q3 is the third quartile, equal to the 75% number after arranging all the data values in the k battery maximum capacity data from small to large. IQR is the interquartile range, equal to the difference between the third quartile and the first quartile, that is, IQR = Q3 - Q1.

[0071] (3) Evaluate the health of the intelligent battery based on the o battery maximum capacity data and the calibrated capacity of the intelligent battery to obtain o health data.

[0072] Specifically, use the method shown in the following formula to evaluate the health of the intelligent battery based on the o battery maximum capacity data and the calibrated capacity of the intelligent battery to obtain o health data:

[0073]

[0074] where D is the health (i.e., SOH), and C o is the calibrated capacity of the intelligent battery.

[0075] 4. Use the GAM fitting method to fit o health data to obtain the battery health decay curve;

[0076] The GAM fitting method is used to fit o health data, and the following is obtained: Figure 3 The battery health degradation curve is shown.

[0077] 5. According to the battery health decay curve, extract the health features of the original charging data in the original charging data set to obtain health feature data;

[0078] According to the battery health decay curve, the raw data in the original data set is manually extracted to obtain health features. Useful health features are manually extracted for subsequent analysis and processing.

[0079] According to the battery health decline curve, we can understand the trend of capacity decline. Understanding the trend of capacity decline will help identify the health features most related to capacity decline.

[0080] Preferably, the health characteristics include battery accumulation characteristics, battery operation characteristics and battery consistency characteristics.

[0081] Specifically, the battery cumulative characteristics include: charging time and total charging capacity.

[0082] The battery operation characteristics include: maximum total voltage, minimum voltage of the highest single cell, maximum temperature of the highest single cell, average temperature of the lowest single cell, minimum total voltage, average voltage of the highest single cell, minimum temperature of the highest single cell, average value of temperature extremes, total voltage variation range, maximum voltage of the lowest single cell, average temperature of the highest single cell, average temperature, average total voltage, minimum voltage of the lowest single cell, maximum temperature of the lowest single cell, average total current, maximum voltage of the highest single cell, average voltage of the lowest single cell, minimum temperature of the lowest single cell, and time difference between the first 1V time and the last 1V time.

[0083] Battery consistency characteristics include: maximum voltage difference and average voltage difference.

[0084] 6. Screen the health characteristic data based on the Pearson correlation analysis method to obtain the correlation characteristic data;

[0085] Preferably, the method shown in the following formula is used to screen the health characteristics based on the Pearson correlation analysis method to obtain the correlation characteristics:

[0086]

[0087] Where P is the Pearson correlation coefficient, which indicates the linear correlation between health feature A and health degree D, N is the number of samples (here the number of samples refers to the number of first charging data), that is, the number of data points in the two sets of data of health feature A and health degree D, R(A i ) The observed value of the ith health feature A, is the average value of the observed value of health characteristic A, R(D i ) the observed value of the i-th D health, is the average value of D health observations.

[0088] After the correlation analysis is performed on each health feature, the correlation analysis calculation results (i.e., the Pearson correlation coefficient mentioned above) are obtained, and the correlation analysis calculation results are arranged in order of size, and the top 11 health features are selected and screened out as correlation features.

[0089] Preferably, the correlation characteristics include: maximum total voltage, average voltage of the lowest voltage monomer, average temperature of the lowest voltage monomer, charging time, minimum total voltage, average pressure difference, average value of temperature extremes, total charging capacity, average voltage of the highest voltage monomer, average temperature of the highest temperature monomer, and average temperature.

[0090] 7. Normalize the correlation feature data to obtain normalized feature data;

[0091] Preferably, the correlation feature data is normalized using the method shown in the following formula to obtain normalized feature data:

[0092]

[0093] Among them, A1 is the correlation feature data, A 1min ,A 1max are the minimum and maximum values ​​in the correlation feature data respectively, and A1′ is the normalized feature data.

[0094] 8. According to the normalized feature data and the battery health decay curve, the LSTM model is optimized by the Bayesian optimization method to obtain the battery health prediction model;

[0095] The normalized feature data is used as the training feature, and the battery health decay curve is used as the label corresponding to the training feature to optimize the LSTM model to obtain the battery health prediction model. Bayesian optimization uses a probability model to estimate the posterior distribution of the objective function, which can effectively explore the entire search space and avoid falling into the local optimal solution. LSTM can learn long-term dependent information and effectively avoid the gradient vanishing problem. Bayesian optimization will optimize the bias and weight of LSTM, which can effectively improve the performance indicators of the model, reduce the variance of the prediction, and collect uncertainty on specific data points.

[0096] Preferably, the weights and biases of the LSTM model are optimized by the Bayesian optimization method, and the formulas for optimizing the weights and biases are as follows:

[0097]

[0098] Where ρ is the standard deviation parameter of the sample linear transformation parameter; μ is the mean parameter of the sample linear change parameter; The weight used for the linear change of the i-th layer on the n-th sample; The bias used for the linear change of the i-th layer on the n-th sample. The weights and biases are optimized during the iteration process.

[0099] like Figure 4 As shown in the figure, the framework diagram of the LSTM model is provided. In the input layer, 1DCNN is used as a feature extractor to provide input features for the LSTM layer model. In the LSTM layer, the Bayesian method is used to optimize the bias and hyperparameters of the LSTM and optimize the performance indicators of the model. In the output layer, the model output is mapped to the final value through a fully connected layer.

[0100] In one embodiment, the trained LSTM model is evaluated using MAPE and RMSE. MAPE can intuitively reflect the error ratio between the predicted value and the actual value, but it is very sensitive to outliers. RMSE can quantify the size of the prediction error and is less sensitive to outliers. It is usually used to measure the prediction error of continuous variables.

[0101]

[0102] Where D i is the current health, The current health predicted by the model.

[0103] 9. Use the battery health prediction model to predict the health of the smart battery and obtain the health prediction result of the smart battery.

[0104] The present invention can accurately predict the health status of the battery by extracting the health characteristics and capacity of qualified charging segments. The present invention is applicable to the estimation of the health status of the battery in real vehicles, realizes the gradient utilization of the battery, helps users adjust their usage habits, alleviates users' mileage anxiety, prolongs the battery's service life and improves the battery's safety, and further provides a reference for the safe operation of the battery.

[0105] In one of the embodiments, a smart battery health status prediction device based on big data is also provided, which is characterized by comprising a collection unit, a first screening unit, a second screening unit, a fitting unit, a feature extraction unit, a third screening unit, a normalization processing unit, an optimization unit and a prediction unit, wherein:

[0106] A collection unit, used to obtain an original charging data set according to the collected charging data of the m original charging segments;

[0107] A first screening unit, configured to screen the original charging data in the original charging data set according to a preset SOC screening condition to obtain a first charging data set, wherein the first charging data set includes k first charging segments;

[0108] A second screening unit, configured to screen the first charging data in the first charging data set to obtain o health data;

[0109] A fitting unit, used to fit o health data using a GAM fitting method to obtain a battery health decay curve;

[0110] A feature extraction unit, used for extracting health features from the original charging data in the original charging data set according to the battery health decay curve to obtain health feature data;

[0111] The third screening unit is used to screen the health characteristic data based on the Pearson correlation analysis method to obtain correlation characteristics;

[0112] A normalization processing unit, used for normalizing the correlation features to obtain normalized feature data;

[0113] An optimization unit is used to optimize the LSTM model by using the Bayesian optimization method according to the normalized feature data and the battery health decay curve to obtain a battery health prediction model;

[0114] The estimation unit is used to estimate the health of the smart battery by using a battery health estimation model to obtain a health estimation result of the smart battery.

[0115] The embodiment of the present application can divide the terminal into functional units according to the above method example. For example, each functional unit can be divided according to each function, or two or more functions can be integrated into one processing unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional units. It should be noted that the division of units in the embodiment of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0116] An embodiment of the present application also provides a terminal including a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute any one of the smart battery health status prediction methods based on big data as recorded in the above method embodiments.

[0117] An embodiment of the present application also provides a computer storage medium, wherein the computer storage medium stores a computer program for electronic data exchange, and the computer program enables a computer to execute part or all of the steps of any one of the big data-based smart battery health status prediction methods recorded in the above method embodiments.

[0118] An embodiment of the present application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program enables a computer to execute part or all of the steps of any one of the big data-based smart battery health status prediction methods recorded in the above method embodiments.

[0119] In the description of the present invention, a large number of specific details are described. However, it is understood that embodiments of the present invention can be practiced without these specific details. In some instances, well-known methods, structures and techniques are not shown in detail so as not to obscure the understanding of this description.

[0120] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, without contradiction.

[0121] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A smart battery health status prediction method based on big data, characterized in that: include: According to the collected charging data of the m original charging segments, an original charging data set is obtained; According to a preset SOC screening condition, the original charging data in the original charging data set is screened to obtain a first charging data set, wherein the first charging data set includes k first charging segments; Screening the first charging data in the first charging data set to obtain o pieces of health data; The GAM fitting method is used to fit o health data to obtain the battery health decay curve; Extracting health features from the original charging data in the original charging data set according to the battery health decay curve to obtain health feature data; The health characteristic data are screened based on the Pearson correlation analysis method to obtain correlation characteristics; Normalizing the correlation features to obtain normalized feature data; According to the normalized feature data and the battery health decay curve, the LSTM model is optimized by the Bayesian optimization method to obtain the battery health prediction model; The battery health prediction model is used to predict the health of the smart battery to obtain a health prediction result of the smart battery.

2. The method for estimating the health status of a smart battery based on big data as claimed in claim 1, characterized in that: The preset SOC screening conditions include a preset SOC span screening condition, a preset SOC growth screening condition and a preset SOC continuity screening condition.

3. A method for estimating the health status of a smart battery based on big data as claimed in claim 1 or 2, characterized in that: The first charging data in the first charging data set is screened to obtain o health data, including: Calculate k battery maximum capacity data according to the first charging data of k first charging segments in the first charging data set; Using a box method to filter out abnormal capacity data from the k battery maximum capacity data to obtain o battery maximum capacity data; According to the maximum capacity data of the battery and the rated capacity of the smart battery, the health of the smart battery is evaluated to obtain health data.

4. The method for estimating the health status of a smart battery based on big data as claimed in claim 3, characterized in that: According to the first charging data of k first charging segments in the first charging data set, k battery maximum capacity data are calculated, including: The method shown in the following formula is used to calculate k battery maximum capacity data according to the first charging data of k first charging segments in the first charging data set: Among them, C i represents the maximum battery capacity corresponding to the i-th first charging segment, Q c (i) represents all the electricity passed through the i-th first charging segment, ΔSOC(i) represents the SOC change value of the i-th first charging segment, t2 and t1 represent the charging cut-off time and the charging start time respectively, I(t) represents the charging current at the current moment, SOC(t2) and SOC(t1) represent the SOC values ​​corresponding to the charging cut-off time t2 and the charging start time t1 respectively.

5. The method for estimating the health status of a smart battery based on big data as claimed in claim 1, characterized in that: The health characteristic data includes battery accumulation characteristics, battery operation characteristics and battery consistency characteristics.

6. A smart battery health status prediction device based on big data, characterized in that: It includes a collection unit, a first screening unit, a second screening unit, a fitting unit, a feature extraction unit, a third screening unit, a normalization processing unit, an optimization unit and an estimation unit, wherein: The collecting unit is used to obtain an original charging data set according to the collected charging data of the m original charging segments; The first screening unit is used to screen the original charging data in the original charging data set according to a preset SOC screening condition to obtain a first charging data set, wherein the first charging data set includes k first charging segments; The second screening unit is used to screen the first charging data in the first charging data set to obtain o health data; The fitting unit is used to fit o health data using a GAM fitting method to obtain a battery health decay curve; The feature extraction unit is used to extract health features from the original charging data in the original charging data set according to the battery health decay curve to obtain health feature data; The third screening unit is used to screen the health feature data based on the Pearson correlation analysis method to obtain correlation features; The normalization processing unit is used to perform normalization processing on the correlation feature to obtain normalized feature data; The optimization unit is used to optimize the LSTM model by a Bayesian optimization method according to the normalized feature data and the battery health decay curve to obtain a battery health prediction model; The estimating unit is used to estimate the health of the smart battery using the battery health estimation model to obtain a health estimation result of the smart battery.

7. A terminal, characterized in that: It includes a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute a smart battery health status prediction method based on big data as described in any one of claims 1 to 5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which includes program instructions. When the program instructions are executed by a processor, the processor executes a smart battery health status prediction method based on big data as described in any one of claims 1 to 5.

Images