Battery health degree estimation method based on GRU neural network and related device

Abstract

The embodiment of the invention relates to the technical field of new energy vehicle batteries, and provides a GRU neural network-based battery health degree estimation method and a related device, and the method comprises the steps: obtaining the historical charging monitoring data of a battery; fitting a battery capacity curve according to the historical charging monitoring data of the battery; extracting first battery charging characteristic data from the battery historical charging monitoring data; screening the first battery charging characteristic data by adopting a spearman correlation analysis method to obtain second battery charging characteristic data; adjusting the GRU neural network by adopting the second battery charging characteristic data and the battery capacity curve to obtain a battery health degree estimation model; the battery health degree estimation model is used for estimating the battery health degree, the calculated amount of the battery health estimation model in the training process can be reduced, and the accuracy of predicting the battery health degree by the battery health model is improved.

Description

Technical Field

[0001] The present application relates to the technical field of new energy vehicle batteries, and specifically to a battery health prediction method and related devices based on a GRU neural network. Background Art

[0002] With the popularity of electric vehicles and various portable electronic devices, battery health monitoring and management has become increasingly important. Battery health is a key indicator to measure the degree of battery performance degradation, which is directly related to the endurance and service life of the device.

[0003] Most of the existing battery health assessment methods use a large amount of battery life data and related battery monitoring data to train a battery health prediction model to estimate the battery health. However, since the battery monitoring data is very complex and some battery monitoring data has little influence on the battery life data, the training process of the battery health prediction model is slow and the accuracy is not high. Summary of the invention

[0004] The embodiment of the present application provides a battery health prediction method and related devices based on a GRU neural network. The battery health prediction model is obtained by fitting the battery capacity curve according to the historical battery charging monitoring data, and extracting the battery monitoring data with the highest battery health through the correlation analysis method to train the GRU neural network. This can reduce the amount of calculation of the battery health prediction model during the training process and improve the accuracy of the battery health model in predicting the battery health.

[0005] A first aspect of an embodiment of the present application provides a battery health prediction method based on a GRU neural network, the method comprising:

[0006] Obtain historical battery charging monitoring data;

[0007] Fitting a battery capacity curve according to the battery historical charging monitoring data;

[0008] Extracting first battery charging characteristic data from the battery historical charging monitoring data;

[0009] Using the spearman correlation analysis method to screen the first battery charging characteristic data to obtain second battery charging characteristic data;

[0010] The GRU neural network is adjusted using the second battery charging characteristic data and the battery capacity curve to obtain a battery health estimation model;

[0011] The battery health prediction model is used to predict the battery health.

[0012] Furthermore, fitting the battery capacity curve according to the battery historical charging monitoring data includes:

[0013] Screening the battery historical charging monitoring data to extract stable growth historical charging data;

[0014] According to the stable growth historical charging data, a local weighted regression method is used to eliminate abnormal data to obtain the stable growth historical charging data after eliminating the abnormal data;

[0015] According to the stable growth historical charging data after removing the abnormal data, the battery capacity curve is fitted by using the lowess smoothing method.

[0016] Furthermore, the screening of the battery historical charging monitoring data to extract the stable growth historical charging data includes:

[0017] The data that does not meet the first screening condition, the second screening condition, the third screening condition and the fourth screening condition in the battery historical charging monitoring data are used to obtain stable growth historical charging data;

[0018] The first screening condition includes a first preset threshold value of the SOC span of the historical charging monitoring data;

[0019] The second screening condition includes that the SOC of the historical charging monitoring data is continuously increasing;

[0020] The third screening condition includes that the absolute value of the current of the historical charging monitoring data is greater than a second preset threshold;

[0021] The fourth screening condition is that the absolute value of the total current of the same SOC field in the historical charging monitoring data is greater than 2000.

[0022] Furthermore, the GRU neural network is adjusted by using the second battery charging characteristic data and the battery capacity curve to obtain a battery health prediction model, including:

[0023] Normalizing the second battery charging characteristic data and the battery capacity curve to obtain normalized second battery charging characteristic data and the battery capacity curve;

[0024] Using a PCA method to reduce the dimension of the normalized second battery charging characteristic data and battery capacity curve to obtain the second battery charging characteristic data and battery capacity curve after dimension reduction;

[0025] The second battery charging feature data after dimension reduction is used as a training feature, and the battery capacity curve is used as a label corresponding to the training feature to adjust the GRU neural network to obtain a battery health prediction model.

[0026] Furthermore, the second battery charging characteristic data includes:

[0027] Maximum total voltage, minimum total voltage, average total voltage, average voltage of the highest voltage cell, average voltage of the lowest voltage cell, average voltage difference, maximum voltage difference, average temperature of temperature extremes, average temperature of the highest temperature cell, average temperature of the lowest temperature cell, average temperature, charging time per unit SOC and total charging capacity.

[0028] In this example, the historical battery charging monitoring data is first obtained, and the battery capacity curve is fitted based on the historical battery charging monitoring data as the label for GRU model training. The second battery charging feature data with the highest correlation is extracted from the historical battery charging monitoring data through the spearman correlation analysis method as the training data for the GRU model. Finally, the battery capacity curve and the second battery charging feature data are used to adjust and train the GRU neural network to obtain a battery health prediction model, which can reduce the amount of calculation of the battery health prediction model during the training process, improve the training efficiency of the battery health prediction model, and improve the accuracy of the battery health model in predicting battery health.

[0029] A second aspect of an embodiment of the present application provides a battery health estimation device, the device comprising:

[0030] A first acquisition unit, used to acquire historical battery charging monitoring data;

[0031] A first processing unit, configured to fit a battery capacity curve according to the battery historical charging monitoring data;

[0032] A second acquisition unit, configured to extract first battery charging characteristic data from the battery historical charging monitoring data;

[0033] a second processing unit, configured to screen the first battery charging characteristic data by using a spearman correlation analysis method to obtain second battery charging characteristic data;

[0034] A third processing unit is used to adjust the GRU neural network by using the second battery charging characteristic data and the battery capacity curve to obtain a battery health estimation model;

[0035] The estimating unit is used to estimate the battery health by using the battery health estimation model.

[0036] Further, in the aspect of fitting the battery capacity curve according to the historical battery charging monitoring data, the first processing unit is used to:

[0037] Screening the battery historical charging monitoring data to extract stable growth historical charging data;

[0038] According to the stable growth historical charging data, a local weighted regression method is used to eliminate abnormal data to obtain the stable growth historical charging data after eliminating the abnormal data;

[0039] According to the stable growth historical charging data after removing the abnormal data, the battery capacity curve is fitted by using the lowess smoothing method.

[0040] Further, in the aspect of screening the battery historical charging monitoring data to extract the stable growth historical charging data, the first processing unit is used to:

[0041] The data that does not meet the first screening condition, the second screening condition, the third screening condition and the fourth screening condition in the battery historical charging monitoring data are used to obtain stable growth historical charging data;

[0042] The first screening condition includes a first preset threshold value of the SOC span of the historical charging monitoring data;

[0043] The second screening condition includes that the SOC of the historical charging monitoring data is continuously increasing;

[0044] The third screening condition includes that the absolute value of the current of the historical charging monitoring data is greater than a second preset threshold;

[0045] The fourth screening condition is that the absolute value of the total current of the same SOC field in the historical charging monitoring data is greater than 2000.

[0046] Further, in the aspect of adjusting the GRU neural network by using the second battery charging characteristic data and the battery capacity curve to obtain the battery health prediction model, the third processing unit is used to:

[0047] Normalizing the second battery charging characteristic data and the battery capacity curve to obtain normalized second battery charging characteristic data and the battery capacity curve;

[0048] Using a PCA method to reduce the dimension of the normalized second battery charging characteristic data and battery capacity curve to obtain the second battery charging characteristic data and battery capacity curve after dimension reduction;

[0049] The second battery charging feature data after dimension reduction is used as a training feature, and the battery capacity curve is used as a label corresponding to the training feature to adjust the GRU neural network to obtain a battery health prediction model.

[0050] 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 the battery health prediction method based on the GRU neural network in the first aspect of the embodiment of the present application.

[0051] The fourth aspect of an embodiment of the present application provides a computer-readable storage medium, wherein the above-mentioned computer-readable storage medium stores a computer program for electronic data exchange, wherein the above-mentioned computer program enables the computer to execute part or all of the steps described in the battery health prediction method based on the GRU neural network in the first aspect of the embodiment of the present application.

[0052] 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

[0053] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0054] Figure 1 A schematic diagram of the overall process of a battery health estimation method based on a GRU neural network is provided for an embodiment of the present application;

[0055] Figure 2 A schematic diagram of specific steps for fitting a battery capacity curve in a battery health estimation method based on a GRU neural network is provided for an embodiment of the present application;

[0056] Figure 3 A schematic diagram of specific steps for training a battery health prediction model of a battery health prediction method based on a GRU neural network is provided for an embodiment of the present application;

[0057] Figure 4 A schematic diagram of the structure of a battery health estimation device is provided for an embodiment of the present application;

[0058] Figure 5 A schematic diagram of the structure of a terminal provided in an embodiment of the present application;

[0059] Reference numerals:

[0060] First acquisition unit-1, first processing unit-2, second acquisition unit-3, second processing unit-4, third processing unit-5, estimation unit-6. DETAILED DESCRIPTION

[0061] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0062] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0063] Reference to "embodiments" in this application means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described in this application may be combined with other embodiments.

[0064] In order to better understand the battery health prediction method based on the GRU neural network provided in the embodiment of the present application, the following first briefly introduces the scenario of applying the battery health prediction method based on the GRU neural network. The existing battery health evaluation method estimates the battery health by training the battery health prediction model, but because the battery monitoring data is very complex, and some battery monitoring data has little impact on the battery life data, the training process of the battery health prediction model is slow and the accuracy is not high.

[0065] The battery health prediction method based on GRU neural network is applied to the battery health prediction device. Figure 1 FIG. 1 shows a schematic diagram of the overall process of a battery health prediction method based on a GRU neural network. Figure 1 As shown, including:

[0066] S1. Obtain the historical battery charging monitoring data. First, deploy a data acquisition terminal on the vehicle whose battery data is to be collected to monitor the vehicle battery monitoring data in real time, including time, longitude, latitude, altitude, remaining battery power, charge and discharge status, current, total voltage, all cell temperatures, all single probe temperatures and other parameter data. After obtaining these parameter data, extract the historical battery charging monitoring data from them as the data basis for subsequent processing.

[0067] S2. Fitting a battery capacity curve according to the battery historical charging monitoring data. Specifically, Figure 2 The figure shows a schematic diagram of the specific steps of fitting the battery capacity curve in the battery health estimation method based on the GRU neural network. Figure 2 As shown, including:

[0068] S201, screening the battery historical charging monitoring data, extracting stable growth historical charging data. Since the SOC of the battery platform is calculated by the battery management system (BMS), directly extracting the charging data will lead to errors, so it is necessary to screen the battery historical charging monitoring data and remove the data affected by the errors, specifically including:

[0069] The data in the battery historical charging monitoring data that does not meet the first screening condition, the second screening condition, the third screening condition and the fourth screening condition are used to obtain stable growth historical charging data,

[0070] Specifically, the first screening condition includes that the SOC span of the historical charging monitoring data is greater than a first preset threshold, the first preset threshold is 30%, and the calculation formula of the first screening condition is: ΔSOC>30%.

[0071] The second screening condition includes that the SOC of the historical charging monitoring data is continuously increasing, and the calculation formula of the second screening condition is: SOC2-SOC1≤1.

[0072] The third screening condition includes that the absolute value of the current of the historical charging monitoring data is greater than the second preset threshold value, and the calculation formula of the third preset threshold value is: |I|>80A, where I is the charging current value.

[0073] The fourth screening condition is that the absolute value of the sum of the currents of the same SOC field in the historical charging monitoring data is greater than 2000, and the calculation formula of the fourth preset threshold is:

[0074] S202: According to the stable growth historical charging data, a local weighted regression method is used to eliminate abnormal data to obtain the stable growth historical charging data after the abnormal data is eliminated.

[0075] S203, fitting the battery capacity curve using the Lowess smoothing method according to the stable growth historical charging data after the abnormal data is removed.

[0076] S3. Extracting first battery charging characteristic data from the battery historical charging monitoring data.

[0077] S4. Using the spearman correlation analysis method to screen the first battery charging characteristic data to obtain the second battery charging characteristic data. Specifically, the calculation formula of the spearman correlation analysis method is as follows:

[0078]

[0079] Among them, ρ is the correlation analysis coefficient, d i is the difference in the rank value corresponding to the i-th data, and n is the number of observed samples.

[0080] After the first battery charging characteristic data is subjected to correlation analysis and calculation, a correlation analysis calculation result is obtained, the calculation results are arranged in order according to size, and the top 13 characteristic data are extracted to obtain the second battery charging characteristic data, including: maximum total voltage, minimum total voltage, average total voltage, average voltage of the highest voltage cell, average voltage of the lowest voltage cell, average pressure difference, maximum pressure difference, average value of temperature extremes, average temperature of the highest temperature cell, average temperature of the lowest temperature cell, average temperature, unit SOC charging time and total charging capacity.

[0081] S5. Use the second battery charging characteristic data and the battery capacity curve to adjust the GRU neural network to obtain a battery health estimation model. Figure 3 The figure shows a schematic diagram of the specific steps of training a battery health prediction model of a battery health prediction method based on a GRU neural network, as shown in FIG. Figure 3 As shown, including:

[0082] S501, normalizing the second battery charging characteristic data and the battery capacity curve to obtain normalized second battery charging characteristic data and the battery capacity curve. In this example, the normalization process uses a MAXMIN normalization formula, and the specific form is as follows:

[0083]

[0084] Among them, X′ is the normalized data, X is the original data, Xmin is the minimum value in the data, and Xmax is the maximum value in the data.

[0085] S502, using a PCA method to reduce the dimension of the normalized second battery charging characteristic data and the battery capacity curve to obtain the reduced-dimensional second battery charging characteristic data and the battery capacity curve. PCA dimensionality reduction can map high-dimensional data to a low-dimensional space, retain important features of the high-dimensional data, and remove noise influence. Specifically, for the normalized data sample X n×m =[x1,x2,...,xn] T Centralize and calculate the covariance matrix. The specific formula is as follows:

[0086]

[0087] Among them, X0 is the decentralized sample matrix, x i is the original feature sample, and m is the number of samples.

[0088]

[0089] Among them, X0 is the decentralized sample matrix, and C is the covariance matrix.

[0090] Then, the covariance matrix is ​​decomposed by eigenvalues, and the characteristic matrix is ​​obtained by sorting. The first k are taken to form the matrix p. Finally, the original data is mapped to the k-dimensional coordinate system after dimensionality reduction through linear transformation, using the following formula:

[0091]

[0092] Among them, X n×m Finally, the feature matrix Y after dimensionality reduction is obtained. The present invention reduces the extracted 13 features to 8 dimensions to obtain the second battery charging feature data after dimensionality reduction, which is convenient for subsequent model training.

[0093] S503. Use the reduced-dimensional second battery charging feature data as training features, and use the battery capacity curve as the label corresponding to the training features to adjust the GRU neural network to obtain a battery health prediction model. This solution adopts the GRU-Attention model. GRU can capture short-term dependencies through short-term memory and gating mechanisms. The Attention mechanism can help the model focus on long-span information. The combination of the two can effectively capture the key factors of electric vehicle life aging.

[0094] S6. Use the battery health prediction model to estimate the battery health. Specifically, when in use, the acquired real-time vehicle battery monitoring data is input into the battery health prediction model to obtain the estimated battery capacity, and then the SOH is calculated using the SOH calculation formula. The specific SOH calculation formula is as follows:

[0095]

[0096] Among them, CC is the current maximum capacity of the battery pack, and CR is the rated capacity of the battery pack.

[0097] In this example, the historical battery charging monitoring data is first obtained, and the battery capacity curve is fitted based on the historical battery charging monitoring data as the label for GRU model training. The second battery charging feature data with the highest correlation is extracted from the historical battery charging monitoring data through the spearman correlation analysis method as the training data for the GRU model. Finally, the battery capacity curve and the second battery charging feature data are used to adjust and train the GRU neural network to obtain a battery health prediction model, which can reduce the amount of calculation of the battery health prediction model during the training process, improve the training efficiency of the battery health prediction model, and improve the accuracy of the battery health model in predicting battery health.

[0098] In line with the above, see Figure 4 , Figure 4 The present invention provides a schematic diagram of a battery health prediction device. Figure 4 As shown, the device comprises:

[0099] A first acquisition unit 1 is used to acquire battery historical charging monitoring data;

[0100] A first processing unit 2, used for fitting a battery capacity curve according to the battery historical charging monitoring data;

[0101] A second acquisition unit 3 is used to extract first battery charging characteristic data from the battery historical charging monitoring data;

[0102] A second processing unit 4 is used for screening the first battery charging characteristic data by using a spearman correlation analysis method to obtain second battery charging characteristic data;

[0103] A third processing unit 5 is used to adjust the GRU neural network using the second battery charging characteristic data and the battery capacity curve to obtain a battery health estimation model;

[0104] The estimating unit 6 is used to estimate the battery health by using the battery health estimation model.

[0105] In a possible implementation, in the aspect of fitting a battery capacity curve according to the historical battery charging monitoring data, the first processing unit is configured to:

[0106] Screening the battery historical charging monitoring data to extract stable growth historical charging data;

[0107] According to the stable growth historical charging data, a local weighted regression method is used to eliminate abnormal data to obtain the stable growth historical charging data after eliminating the abnormal data;

[0108] According to the stable growth historical charging data after removing the abnormal data, the battery capacity curve is fitted by using the lowess smoothing method.

[0109] In a possible implementation, in the aspect of screening the battery historical charging monitoring data to extract the stable growth historical charging data, the first processing unit is used to:

[0110] The data that does not meet the first screening condition, the second screening condition, the third screening condition and the fourth screening condition in the battery historical charging monitoring data are used to obtain stable growth historical charging data;

[0111] The first screening condition includes a first preset threshold value of the SOC span of the historical charging monitoring data;

[0112] The second screening condition includes that the SOC of the historical charging monitoring data is continuously increasing;

[0113] The third screening condition includes that the absolute value of the current of the historical charging monitoring data is greater than a second preset threshold;

[0114] The fourth screening condition is that the absolute value of the total current of the same SOC field in the historical charging monitoring data is greater than 2000.

[0115] Further, in the aspect of adjusting the GRU neural network by using the second battery charging characteristic data and the battery capacity curve to obtain the battery health prediction model, the third processing unit is used to:

[0116] Normalizing the second battery charging characteristic data and the battery capacity curve to obtain normalized second battery charging characteristic data and the battery capacity curve;

[0117] Using a PCA method to reduce the dimension of the normalized second battery charging characteristic data and battery capacity curve to obtain the second battery charging characteristic data and battery capacity curve after dimension reduction;

[0118] The second battery charging feature data after dimension reduction is used as a training feature, and the battery capacity curve is used as a label corresponding to the training feature to adjust the GRU neural network to obtain a battery health prediction model.

[0119] For the above embodiments, please refer to Figure 5 , Figure 5 A schematic diagram of the structure of a terminal provided in an embodiment of the present application, such as Figure 5As shown, it includes a processor, an input device, an output device and a memory, which are connected to each other, wherein the memory is used to store a computer program, the computer program includes program instructions, the processor is configured to call the program instructions, and the program includes instructions for executing the following steps;

[0120] Obtain historical battery charging monitoring data;

[0121] Fitting a battery capacity curve according to the battery historical charging monitoring data;

[0122] Extracting first battery charging characteristic data from the battery historical charging monitoring data;

[0123] Using the spearman correlation analysis method to screen the first battery charging characteristic data to obtain second battery charging characteristic data;

[0124] The GRU neural network is adjusted using the second battery charging characteristic data and the battery capacity curve to obtain a battery health estimation model;

[0125] The battery health prediction model is used to predict the battery health.

[0126] In this example, the historical battery charging monitoring data is obtained, and the battery capacity curve is fitted based on the historical battery charging monitoring data as the label for GRU model training. The second battery charging feature data with the highest correlation is extracted from the historical battery charging monitoring data through the spearman correlation analysis method as the training data for the GRU model. Finally, the battery capacity curve and the second battery charging feature data are used to adjust and train the GRU neural network to obtain a battery health prediction model. This can reduce the amount of calculation of the battery health prediction model during the training process, improve the training efficiency of the battery health prediction model, and improve the accuracy of the battery health model in predicting battery health.

[0127] The above mainly introduces the scheme of the embodiment of the present application from the perspective of the execution process on the method side. It is understandable that in order to realize the above functions, the terminal includes a hardware structure and / or software module corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of each example described in the embodiments provided herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0128] 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.

[0129] 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 battery health prediction methods based on the GRU neural network recorded in the above method embodiments.

[0130] 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 battery health prediction methods based on the GRU neural network recorded in the above method embodiments.

[0131] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described order of actions, because according to the present application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0132] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0133] In the several embodiments provided in the present application, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are only schematic, such as the division of the units, which is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the device or unit can be electrical or other forms.

[0134] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0135] In addition, the functional units in the various embodiments of the application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated units may be implemented in the form of hardware or in the form of software program modules.

[0136] If the integrated unit is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a memory, including a number of instructions to enable a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned memory includes: U disk, read-only memory (ROM), random access memory (RAM), mobile hard disk, disk or optical disk and other media that can store program codes.

[0137] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable memory, which can include: a flash drive, a read-only memory, a random access memory, a magnetic disk or an optical disk, etc.

[0138] The embodiments of the present application are introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of ​​the present application. At the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A battery health prediction method based on GRU neural network, characterized in that: include: Obtain historical battery charging monitoring data; Fitting a battery capacity curve according to the historical charging monitoring data of the battery; Extracting first battery charging characteristic data from the battery historical charging monitoring data; Using the spearman correlation analysis method to screen the first battery charging characteristic data to obtain second battery charging characteristic data; The GRU neural network is adjusted using the second battery charging characteristic data and the battery capacity curve to obtain a battery health estimation model; The battery health prediction model is used to predict the battery health.

2. The method for estimating battery health based on a GRU neural network according to claim 1, characterized in that: The step of fitting a battery capacity curve according to the battery historical charging monitoring data comprises: Screening the battery historical charging monitoring data to extract stable growth historical charging data; According to the stable growth historical charging data, a local weighted regression method is used to eliminate abnormal data to obtain the stable growth historical charging data after eliminating the abnormal data; According to the stable growth historical charging data after removing the abnormal data, the battery capacity curve is fitted by using the lowess smoothing method.

3. The method for estimating battery health based on a GRU neural network according to claim 2, characterized in that: The screening of the battery historical charging monitoring data to extract the stable growth historical charging data includes: The data that does not meet the first screening condition, the second screening condition, the third screening condition and the fourth screening condition in the battery historical charging monitoring data are used to obtain stable growth historical charging data; The first screening condition includes that the SOC span of the historical charging monitoring data is greater than a first preset threshold; The second screening condition includes that the SOC of the historical charging monitoring data is continuously increasing; The third screening condition includes that the absolute value of the current of the historical charging monitoring data is greater than a second preset threshold; The fourth screening condition is that the absolute value of the total current of the same SOC field in the historical charging monitoring data is greater than 2000.

4. The method for estimating battery health based on a GRU neural network according to claim 3, characterized in that: The method of adjusting the GRU neural network by using the second battery charging characteristic data and the battery capacity curve to obtain a battery health prediction model includes: Normalizing the second battery charging characteristic data and the battery capacity curve to obtain normalized second battery charging characteristic data and the battery capacity curve; Using a PCA method to reduce the dimension of the normalized second battery charging characteristic data and battery capacity curve to obtain the second battery charging characteristic data and battery capacity curve after dimension reduction; The second battery charging feature data after dimension reduction is used as a training feature, and the battery capacity curve is used as a label corresponding to the training feature to adjust the GRU neural network to obtain a battery health prediction model.

5. The method for estimating battery health based on a GRU neural network according to claim 1, characterized in that: The second battery charging characteristic data includes: Maximum total voltage, minimum total voltage, average total voltage, average voltage of the highest voltage cell, average voltage of the lowest voltage cell, average voltage difference, maximum voltage difference, average temperature of temperature extremes, average temperature of the highest temperature cell, average temperature of the lowest temperature cell, average temperature, charging time per unit SOC and total charging capacity.

6. A battery health prediction device, characterized in that: include: A first acquisition unit, used to acquire historical battery charging monitoring data; A first processing unit, configured to fit a battery capacity curve according to the battery historical charging monitoring data; A second acquisition unit, configured to extract first battery charging characteristic data from the battery historical charging monitoring data; a second processing unit, configured to screen the first battery charging characteristic data by using a spearman correlation analysis method to obtain second battery charging characteristic data; A third processing unit is used to adjust the GRU neural network by using the second battery charging characteristic data and the battery capacity curve to obtain a battery health estimation model; The estimating unit is used to estimate the battery health by using the battery health estimation model.

7. The battery health prediction device according to claim 6, characterized in that: In the aspect of fitting a battery capacity curve according to the historical battery charging monitoring data, the first processing unit is used to: Screening the battery historical charging monitoring data to extract stable growth historical charging data; According to the stable growth historical charging data, a local weighted regression method is used to eliminate abnormal data to obtain the stable growth historical charging data after eliminating the abnormal data; According to the stable growth historical charging data after removing the abnormal data, the battery capacity curve is fitted by using the lowess smoothing method.

8. The battery health prediction device according to claim 7, characterized in that: In the aspect of screening the battery historical charging monitoring data to extract the stable growth historical charging data, the first processing unit is used to: The data that does not meet the first screening condition, the second screening condition, the third screening condition and the fourth screening condition in the battery historical charging monitoring data are used to obtain stable growth historical charging data; The first screening condition includes a first preset threshold value of the SOC span of the historical charging monitoring data; The second screening condition includes that the SOC of the historical charging monitoring data is continuously increasing; The third screening condition includes that the absolute value of the current of the historical charging monitoring data is greater than a second preset threshold; The fourth screening condition is that the absolute value of the total current of the same SOC field in the historical charging monitoring data is greater than 2000.

9. A terminal, characterized in that: It includes a processor, an input device, an output device and a memory, which 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 the battery health prediction method based on the GRU neural network as described in any one of claims 1-5.

10. 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 the battery health prediction method based on the GRU neural network as described in any one of claims 1 to 5.

Images