WO2022257211A1 - Method and system for predicting remaining service life of lithium battery - Google Patents

Abstract

The present invention relates to the technical field of lithium batteries, and provides a method and system for predicting the remaining service life of a lithium battery. The method comprises: according to a historical battery capacity value of a lithium battery, obtaining a feature matrix of voltage, current and temperature data corresponding to each battery capacity value; considering the weight values of the voltage, current and temperature data and the effect of different feature types on the battery capacity values, combining feature matrices of the voltage, current and temperature data corresponding to the battery capacity values, and obtaining a measurement data feature matrix, a feature type matrix and a measurement data and feature type fusion matrix corresponding to different attention mechanisms; and splicing the measurement data feature matrix, the feature type matrix and the measurement data and feature type fusion matrix to obtain the predicted remaining service life of the lithium battery.

Description

Method and system for predicting remaining service life of lithium battery technical field

The invention belongs to the technical field of lithium batteries, in particular to a method and system for predicting the remaining service life of lithium batteries.

Background technique

The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

Lithium batteries are widely used in electric vehicles, aerospace, mobile devices and other fields due to their stability, safety and environmental friendliness. Although the repeated charging and discharging of lithium-ion batteries brings convenience to the operation of the equipment, as the charging and discharging cycles of lithium-ion batteries increase, their capacity will decrease and their safety will also deteriorate. If the lithium battery is not replaced before the capacity decays to a certain extent, it will have an unpredictable impact on the equipment and even cause a safety accident. Therefore, it is necessary to predict the remaining service life of lithium batteries.

Inspired by the excellent performance of machine learning in various fields, many models based on convolutional neural network (CNN) and recurrent neural network (RNN) have been applied to lithium battery remaining useful life (Remaining Useful Life, RUL) prediction. CNN treats each charging and discharging data as an independent feature vector, ignoring the key information in the feature. RNN can utilize time series information, but it often brings the problem of exploding or disappearing gradients. Long short-term memory unit (LSTM) is an optimization of RNN that can solve the above problems. The LSTM-based method usually uses the previous prediction value as the feature data of the next prediction. Due to the accumulation of prediction errors in this iterative method, the later battery capacity prediction will become more and more inaccurate. Although the above method can make reasonable predictions for the RUL of lithium-ion batteries, for complex battery capacity curves, especially in the case of long battery charge and discharge cycles, the prediction accuracy can be further improved.

Contents of the invention

In order to make full use of the key feature information of lithium batteries and eliminate the influence of the above LSTM prediction method, the present invention proposes a method and system for predicting the remaining service life of lithium batteries. Specifically, this method forms each measurement data and each feature into a feature matrix, weights the features through a mixed attention mechanism, and focuses on features that are beneficial to high prediction accuracy. At the same time, the present invention only uses the measurement data as the characteristics of the lithium battery, which eliminates the influence of the prediction accumulation error.

In order to achieve the above object, the present invention adopts the following technical solutions:

A first aspect of the present invention provides a method for predicting the remaining service life of a lithium battery.

A method for predicting the remaining service life of a lithium battery, comprising:

According to the historical battery capacity value of the lithium battery, the characteristic matrix of voltage, current and temperature data corresponding to each battery capacity value is obtained;

Considering the weight value of voltage, current and temperature data and the impact of different feature types on the battery capacity value, combined with the feature matrix of voltage, current and temperature data corresponding to the battery capacity value, the measurement data feature matrix and feature matrix corresponding to different attention mechanisms are obtained. type matrix and measurement data and feature type fusion matrix; the measurement data feature matrix, feature type matrix and measurement data and feature type fusion matrix are spliced to obtain the predicted remaining life of the lithium battery.

Further, after obtaining the historical battery capacity value of the lithium battery, it includes obtaining the voltage, current and temperature data corresponding to each battery capacity value; and preprocessing the voltage, current and temperature data corresponding to each battery capacity value, extracting A characteristic matrix of voltage, current, and temperature data corresponding to battery capacity values.

Further, the preprocessing includes: obtaining the row vectors of the voltage data matrix, current data matrix, and temperature data matrix corresponding to each battery capacity value, and then extracting the energy in the voltage row vector, current row vector, and temperature data row vector respectively features, volatility index features, skewness index features, and kurtosis index features.

Further, at least two of the linear regression features, energy features, volatility index features, skewness index features, and kurtosis index features in the voltage line vector, current line vector, and temperature data line vector are used to predict the remaining life of the lithium battery .

Further, the preprocessing includes: normalizing the voltage, current and temperature data corresponding to each battery capacity value.

Further, after acquiring the historical battery capacity value of the lithium battery includes, using a Gaussian filter to filter the historical battery capacity value of the lithium battery to filter out noise.

Further, the influence of different characteristic types on the battery capacity value includes: the influence of the weight of different characteristic types on the characteristic matrix of the voltage, current and temperature data corresponding to the capacity value.

A second aspect of the present invention provides a lithium battery remaining service life prediction system.

A lithium battery remaining service life prediction system, comprising:

The first processing module is configured to: obtain a characteristic matrix of voltage, current and temperature data corresponding to each battery capacity value according to the historical battery capacity value of the lithium battery;

The second processing module is configured to: consider the weight value of the voltage, current and temperature data and the impact of different feature types on the battery capacity value, and combine the feature matrix of the voltage, current and temperature data corresponding to the battery capacity value to obtain different attention The measurement data feature matrix, feature type matrix, and measurement data and feature type fusion matrix corresponding to the force mechanism; the measurement data feature matrix, feature type matrix, and measurement data and feature type fusion matrix are spliced to obtain the predicted remaining life of the lithium battery.

A third aspect of the present invention provides a computer readable storage medium.

A computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, it realizes the steps in the method for predicting the remaining service life of a lithium battery as described in the first aspect above.

A fourth aspect of the present invention provides a computer device.

A computer device, comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, when the processor executes the program, the remaining service life of the lithium battery as described in the first aspect above is realized The steps in the prediction method.

Compared with prior art, the beneficial effect of the present invention is:

In order to make full use of the key characteristics of lithium-ion batteries and eliminate the influence of the above-mentioned LSTM prediction method, the present invention forms a feature matrix for each measurement data and each feature, weights the features through a mixed attention mechanism, and focuses on high prediction accuracy The characteristics of favorable rate, while reducing and calculating, improve the prediction accuracy of the remaining service life of lithium batteries. At the same time, only the measured data are used as the characteristics of the lithium battery, which eliminates the influence of the cumulative error in prediction.

The method proposed by the invention is tested on the NASA lithium battery data set, and the experiment shows that the method can increase the prediction accuracy rate by 44.4%.

Advantages of additional aspects of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.

Description of drawings

The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention, and the schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations to the present invention.

Fig. 1 is the flow chart of the lithium battery remaining service life prediction method of the present invention;

Fig. 2 (a) is the capacity graph of general battery;

Figure 2(b) is the capacity curve of a general battery after a Gaussian filter;

Fig. 3 is when the present embodiment only adopts energy index as feature type, the contrast figure of prediction curve and target curve;

Fig. 4 is a comparison diagram between the forecast curve and the target curve when only the energy index and the volatility index are used as the feature types in this embodiment;

Fig. 5 is a comparison diagram between the forecast curve and the target curve when only energy index, volatility index, and skewness index are used as feature types in this embodiment;

Fig. 6 is a comparison diagram between the forecast curve and the target curve when only energy index, volatility index, skewness index, and kurtosis index are used as feature types in this embodiment;

FIG. 7 is a comparison diagram between the prediction curve and the target curve when the present embodiment only uses the attention mechanism for each physical parameter;

FIG. 8 is a comparison diagram between the prediction curve and the target curve when only the attention mechanism for each feature is used in this embodiment;

FIG. 9 is a comparison diagram between the prediction curve and the target curve when only the attention mechanism for each data point is used in this embodiment;

FIG. 10 is a comparison diagram between the prediction curve and the target curve when only the attention mechanism for each physical parameter and the attention mechanism for each feature are used in this embodiment;

FIG. 11 is a comparison diagram between the prediction curve and the target curve when the present embodiment only adopts the attention mechanism for each physical parameter and the attention mechanism for each data point;

FIG. 12 is a comparison diagram between the prediction curve and the target curve when only the attention mechanism for each feature and the attention mechanism for each data point are used in this embodiment;

FIG. 13 is a comparison diagram of the prediction curve and the target curve when the attention mechanism for each physical parameter, the attention mechanism for each feature, and the attention mechanism for each data point are used in this embodiment.

Fig. 14 is a schematic diagram of linear regression feature calculation in this embodiment.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

It should be noted that the following detailed description is exemplary and intended to provide further explanation of the present invention. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

It should be noted that the terminology used here is only for describing specific embodiments, and is not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly dictates otherwise, the singular is intended to include the plural, and it should also be understood that when the terms "comprising" and/or "comprising" are used in this specification, they mean There are features, steps, operations, means, components and/or combinations thereof.

It should be noted that the flowcharts and block diagrams in the figures show the architecture, functions and operations of possible implementations of the methods and systems according to various embodiments of the present disclosure. It should be noted that each block in a flowchart or a block diagram may represent a module, a program segment, or a part of a code, and the module, a program segment, or a part of a code may include one or more An executable instruction for a specified logical function. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or they may sometimes be executed in the reverse order, depending upon the functionality involved. It should also be noted that each block in the flowchart and/or block diagrams, and combinations of blocks in the flowchart and/or block diagrams, can be implemented using a dedicated hardware-based system that performs the specified functions or operations , or can be implemented using a combination of dedicated hardware and computer instructions.

Embodiment one

As shown in Figure 1, this embodiment provides a method for predicting the remaining service life of a lithium battery. This embodiment uses this method to be applied to a server as an example. It can be understood that this method can also be applied to a terminal, or It includes terminals, servers and systems, and is realized through the interaction between terminals and servers. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or it can provide cloud services, cloud database, cloud computing, cloud function, cloud storage, network server, cloud communication, intermediate Cloud servers for basic cloud computing services such as software services, domain name services, security service CDN, and big data and artificial intelligence platforms. The terminal may be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto. The terminal and the server may be connected directly or indirectly through wired or wireless communication, which is not limited in this application. In this embodiment, the method includes the following steps:

According to the historical battery capacity value of the lithium battery, the characteristic matrix of voltage, current and temperature data corresponding to each battery capacity value is obtained;

Considering the weight value of voltage, current and temperature data and the impact of different feature types on the battery capacity value, combined with the feature matrix of voltage, current and temperature data corresponding to the battery capacity value, the measurement data feature matrix and feature matrix corresponding to different attention mechanisms are obtained. type matrix and measurement data and feature type fusion matrix; the measurement data feature matrix, feature type matrix and measurement data and feature type fusion matrix are spliced to obtain the predicted remaining life of the lithium battery.

Specifically, in this embodiment, the battery capacity of the battery is used as the target output of the RUL prediction, and the data sequence of voltage, current and temperature corresponding to each capacity value is used as the input feature. Specifically, the method mainly includes three parts: 1. Data preprocessing and feature extraction; 2. Using the mixed attention mechanism to calculate the contribution value of different attribute features; 3. Using the trained network for battery RUL prediction. The implementation process of each part will be introduced in detail below.

1) Data preprocessing

Since the voltage, current, and temperature data lengths corresponding to different capacitance values are different, the data set needs to be preprocessed. Assuming that the capacitance value of the jth measurement is y _j , the corresponding voltage, current, and temperature sequence data are respectively

This method saves the data features as a two-dimensional matrix, and the row vector of the matrix is

where k ∈ {v,c,tp}. In order to get the first data of each row of data in the matrix

to the 10th data

First put

Divide into 10 equal parts, and take the mean value of each part as

arrive

value.

specific,

Represents the j-th voltage, current or temperature sequence data, because the duration of each observation is different,

may be as long as

Different lengths. In order to normalize the lengths of all input features to the same value, all the

All are divided into 10 equal parts, and the mean value of each part is calculated as

arrive

value.

and

different meaning,

is the original measurement data,

is transformed

This preprocessing method can ensure that the lengths of feature data corresponding to different capacitance values are consistent. feature

and

feature

The linear fitting factor of . feature

respectively

The energy (Eg), volatility index (FI), skewness index (SI), and kurtosis index (KI) characteristics of , the specific acquisition methods are as follows:

In formula (2),

Indicates the sampling rate; in formula (3)-(4),

and

Respectively

mean and standard deviation of .

Then normalize the acquired sample data:

in

and

data respectively

maximum and minimum values of .

Using the battery capacity at the jth cycle as the prediction target _yj , Figure 2(a) shows the capacity curve of a general battery. It can be seen from this that the raw capacity data has volatility, which may be caused by measurement errors, electromagnetic interference, complex chemicals, etc. Therefore, the raw volumetric data is processed using a Gaussian filter. The filtered results of the processed capacity curve are shown in Fig. 2(b). For statistical scaling, the target curve needs to be normalized to

2) The mixed attention mechanism calculates the contribution value of different attribute features

In order to focus on and weight important information in the feature matrix, a modified self-attention mechanism is used to calculate the contribution weights by considering different attributes of features.

Let the characteristic matrix of the jth cycle be

j denotes the jth measurement. matrix

The row data in are feature vectors about different measurements (voltage, current, and temperature), while the data in different columns represent different kinds of features. Since the eigenmatrix is calculated separately

Each row vector, each column vector and the contribution value of each feature need to use different attention mechanisms, so the method proposed in this paper is called lithium battery RUL prediction based on hybrid attention mechanism. Specifically, the following formula is used to calculate the weight of different measurements (or row data)

here

Represents the hidden layer parameters that need to be trained. Then, the output matrix

It can be expressed as

with corresponding attributes

The product of:

matrix

The columns represent different kinds of features. Similarly, for different types of features, we have similar weight calculation formulas and output matrix D _j = (D _j,1 ,D _j,2 ,…,D _j,16 ) calculation formulas:

Here β _j,i represents the weight of feature i (i=1, 2, . . . , 16) of the jth period. In addition to the separate outputs of row-based and column-based attention mechanisms, the weight matrix can also be a weight vector

and β _j = the product of (β _j,1 , β _j,2 , . . . , β _j,16 ). Assuming its output matrix is set to E _j , the elements of E _j

It can be obtained by the following formula:

3) Prediction of battery RUL

In this embodiment, the weighted features C _j , D _j and E _j will be concatenated and then input to the fully connected layer. Eventually, the network will output the predicted RUL of the lithium battery.

Among them, the fully connected layer is used to linearly transform the features. If the feature input of the fully connected layer is x, the output is y, and the parameter matrix is W, then the fully connected layer realizes the following functions:

y=Wx

In order to prove the technical solution of the embodiment, experiments are carried out on the NASA lithium battery data set, and the experiments show that the method can improve the prediction accuracy.

The lithium battery data set used in this embodiment comes from the Prognostics Center of Excellence of National Aeronautics and Space Administration (NASA), which is a widely used data set in lithium battery RUL prediction. This data set records the multi-physical parameter data of the content degradation of multiple lithium batteries during use. The invention will be verified on #5, #6, #7 and #18 batteries.

The specific use scenarios of the above four batteries are shown in Table 1, where 'V _up ' represents the constant charging voltage, 'V _low ' represents the voltage at the end of discharge, and 'I _char ' and 'I _dis ' represent the voltage during charge and discharge respectively. Current, 'Temperature' indicates the temperature of the battery in use, 'Original Capacity' indicates the capacity of the new battery, and finally 'Cycle' indicates the total charge and discharge cycles of the battery.

Table 1. #5, #6, #7 and #18 battery usage records

Battery	V up	V low	I char	I dis	temperature	raw capacity	cycle
#5	4.2	2.7	1.5	2	twenty four	1.86	168
#6	4.2	2.5	1.5	2	twenty four	2.04	168
#7	4.2	2.5	1.5	2	twenty four	1.89	168
#18	4.2	2.5	1.5	2	twenty four	1.85	132

Prediction result evaluation index:

In order to effectively evaluate our method from multiple perspectives, this example chooses two widely used metrics: Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). Suppose the target capacity value is y={y ₁ ,y ₂ ,…,y _n }, and the predicted capacity value is

Then RMSE and MAPE are defined as

The smaller the value of RMSE and MAPE, the closer the predicted result is to the real value.

In this embodiment, Liu et al. (Liu, Kailong, et al. "A Data-Driven Approach With Uncertainty Quantification for Predicting Future Capacities and Remaining Useful Life of Lithium-Ion Battery." IEEE Transactions on Industrial Electronics, vol.68, no .4, 2021, pp.3170–3180.) and Chen et al. (Chen, Liaogehao et al. "Remaining useful life prediction of lithium-ion battery with optimal input sequence selection and error compensation." Neurocomputing, vol.414, 2020 , pp.245-254.) The proposed method was compared. Table 2 shows the RMSE and MAPE of the method of this example and other methods on different batteries, which shows that the method of this example performs better than other methods on #5, #7 and #18 batteries.

Table 2 Performance comparison with other methods.

Embodiment two

This embodiment provides a method for predicting the remaining service life of a lithium battery.

A method for predicting the remaining service life of a lithium battery, comprising:

According to the historical battery capacity value of the lithium battery, the characteristic matrix of voltage, current and temperature data corresponding to each battery capacity value is obtained;

Considering the weight value of voltage, current and temperature data and the impact of different feature types on the battery capacity value, combined with the feature matrix of voltage, current and temperature data corresponding to the battery capacity value, the measurement data feature matrix and feature matrix corresponding to different attention mechanisms are obtained. type matrix and measurement data and feature type fusion matrix; the measurement data feature matrix, feature type matrix and measurement data and feature type fusion matrix are spliced to obtain the predicted remaining life of the lithium battery.

As one or more implementations, after obtaining the historical battery capacity value of the lithium battery, it includes obtaining the voltage, current and temperature data corresponding to each battery capacity value; and the voltage, current and temperature data corresponding to each battery capacity value The data is preprocessed to extract the characteristic matrix of the voltage, current and temperature data corresponding to the battery capacity value.

Preprocessing includes: obtaining the row vectors of the voltage data matrix, current data matrix, and temperature data matrix corresponding to each battery capacity value, and then extracting the energy features and fluctuation index features in the voltage row vectors, current row vectors, and temperature data row vectors respectively , skewness index feature and kurtosis index feature.

Using at least two of the linear regression features, energy features, volatility index features, skewness index features, and kurtosis index features in the voltage row vector, the current row vector, and the temperature data row vector to predict the remaining life of the lithium battery.

There are various technical solutions in this embodiment, specifically, different combinations of features in preprocessing; different attention mechanisms are combined, and so on. Here is an example to show and compare the realization effects of different combinations, and the example data set and evaluation criteria are the same as above. Experiments were performed on battery #5. Specifically, the data of the first 80 cycles of batteries #6, #7, #18, and battery #5 are used as the training data set, and the data of the remaining cycles of battery #5 are used for performance testing.

1) The influence of different feature combinations on the lithium battery RUL prediction results:

In order to predict the capacity of the lithium battery more accurately, some parameters in this embodiment need to be tested and adjusted. This dataset contains various monitoring data during charging and discharging, such as measured voltage, current and temperature data. Features involving linear regression (LR), energy index (Eg), volatility index (FI), skewness index (SI), and kurtosis index (KI) can be obtained from these three measures, and different combinations of these feature data will It has a greater impact on the prediction results, as shown in Figure 3-6. Table 3 shows the results based on different feature combinations, and Figure 3 shows the curves of target capacity and predicted capacity for different feature combinations. It can be seen that the common combination of all extracted features is better than the other combinations. Therefore, the following experiments are performed based on these feature combinations.

Table 3. Performance metrics for different feature combinations

Among them, linear regression refers to: if the abscissa represents the feature number i, and the ordinate represents the observed data (such as current, voltage or temperature data), then the relationship between i and the observed data can be fitted by a straight line, as shown in Figure 14 straight line in . The slope and intercept of the line are two linear regression features.

2) The influence of different attention mechanisms on the RUL prediction results of lithium batteries:

Three attention mechanisms are used in this example. The second experiment evaluates the attention mechanism used in the method, and Table 4 and Figure 4 show the results for different attention mechanisms. 'For each physical parameter' (A1) in Table 4 weights the three measurement series values separately based on equation (8). 'For each feature' (A2) means that sixteen weights will be obtained according to equation (10) to weight each feature, while 'for each data point' (A3) is based on equation (11) for the feature matrix Each point is weighted; as shown in Figure 7-13. From Table 4, it can be seen that the concatenation of A2 and A3 works best in all experiments, and the reason may be that more emphasis is placed on the type of features rather than the type of measurement data.

Table 4. Performance comparison of the impact of different attention mechanisms on lithium battery RUL prediction results

different methods	RMSE	MAPE(%)
For each physical parameter (A1)	0.0024	0.2530
For each feature (A2)	0.0043	0.4644
For each data point (A3)	0.0049	0.5089
A1 and A2	0.0057	0.5935
A1 and A3	0.0033	0.3589
A2 and A3	0.0020	0.2016
A1, A2 and A3	0.0024	0.2386

Embodiment three

This embodiment provides a lithium battery remaining service life prediction system.

A lithium battery remaining service life prediction system is characterized in that it comprises:

The first processing module is configured to: obtain a characteristic matrix of voltage, current and temperature data corresponding to each battery capacity value according to the historical battery capacity value of the lithium battery;

The second processing module is configured to: consider the weight value of the voltage, current and temperature data and the impact of different feature types on the battery capacity value, and combine the feature matrix of the voltage, current and temperature data corresponding to the battery capacity value to obtain different attention The measurement data feature matrix, feature type matrix, and measurement data and feature type fusion matrix corresponding to the force mechanism; the measurement data feature matrix, feature type matrix, and measurement data and feature type fusion matrix are spliced to obtain the predicted remaining life of the lithium battery.

Embodiment four

This embodiment provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps in the method for predicting the remaining service life of a lithium battery as described in Embodiment 1 or Embodiment 2 above are implemented .

Embodiment five

This embodiment provides a computer device, including a memory, a processor, and a computer program stored on the memory and operable on the processor. When the processor executes the program, the above-mentioned embodiment 1 or embodiment 2 is implemented. The steps in the method for predicting the remaining service life of the lithium battery.

Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, systems, or computer program products. Accordingly, the present invention can take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage and optical storage, etc.) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow diagram procedure or procedures and/or block diagram procedures or blocks.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented through computer programs to instruct related hardware, and the programs can be stored in a computer-readable storage medium. During execution, it may include the processes of the embodiments of the above-mentioned methods. Wherein, the storage medium may be a magnetic disk, an optical disk, a read-only memory (Read-Only Memory, ROM) or a random access memory (Random AccessMemory, RAM), etc.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (10)

1. A method for predicting the remaining service life of a lithium battery, comprising:

   According to the historical battery capacity value of the lithium battery, the characteristic matrix of voltage, current and temperature data corresponding to each battery capacity value is obtained;

   Considering the weight value of voltage, current and temperature data and the impact of different feature types on the battery capacity value, combined with the feature matrix of voltage, current and temperature data corresponding to the battery capacity value, the measurement data feature matrix and feature matrix corresponding to different attention mechanisms are obtained. type matrix and measurement data and feature type fusion matrix; the measurement data feature matrix, feature type matrix and measurement data and feature type fusion matrix are spliced to obtain the predicted remaining life of the lithium battery.
2. The method for predicting the remaining service life of a lithium battery according to claim 1, wherein after obtaining the historical battery capacity value of the lithium battery, it comprises, obtaining voltage, current and temperature data corresponding to each battery capacity value; and for each The voltage, current and temperature data corresponding to the battery capacity value are preprocessed to extract the characteristic matrix of the voltage, current and temperature data corresponding to the battery capacity value.
3. The method for predicting the remaining service life of a lithium battery according to claim 2, wherein the preprocessing comprises: obtaining the row vectors of the voltage data matrix, current data matrix, and temperature data matrix corresponding to each battery capacity value, and then respectively Extract the energy features, fluctuation index features, skewness index features and kurtosis index features in the voltage row vector, current row vector and temperature data row vector.
4. The method for predicting the remaining service life of a lithium battery according to claim 3, wherein the linear regression feature, energy feature, fluctuation index feature, skewness index feature and At least two of the kurtosis index features predict the remaining life of a lithium battery.
5. The method for predicting the remaining service life of a lithium battery according to claim 2, wherein the preprocessing includes: normalizing the voltage, current and temperature data corresponding to each battery capacity value.
6. The method for predicting the remaining service life of a lithium battery according to claim 2, wherein after obtaining the historical battery capacity value of the lithium battery, it includes, using a Gaussian filter to filter the historical battery capacity value of the lithium battery to filter out noise .
7. The method for predicting the remaining service life of a lithium battery according to claim 1, wherein the impact of the different feature types on the battery capacity value includes: the weight of different feature types on the characteristics of the voltage, current and temperature data corresponding to the capacity value matrix effect.
8. A lithium battery remaining service life prediction system is characterized in that it comprises:

   The first processing module is configured to: obtain a characteristic matrix of voltage, current and temperature data corresponding to each battery capacity value according to the historical battery capacity value of the lithium battery;

   The second processing module is configured to: consider the weight value of the voltage, current and temperature data and the impact of different feature types on the battery capacity value, and combine the feature matrix of the voltage, current and temperature data corresponding to the battery capacity value to obtain different attention The measurement data feature matrix, feature type matrix, and measurement data and feature type fusion matrix corresponding to the force mechanism; the measurement data feature matrix, feature type matrix, and measurement data and feature type fusion matrix are spliced to obtain the predicted remaining life of the lithium battery.
9. A computer-readable storage medium, on which a computer program is stored, characterized in that, when the program is executed by a processor, the steps in the method for predicting the remaining service life of a lithium battery according to any one of claims 1-7 are realized .
10. A computer device, comprising a memory, a processor, and a computer program stored on the memory and operable on the processor, wherein the processor implements any one of claims 1-7 when executing the program The steps in the method for predicting the remaining service life of the lithium battery.

Images