Lithium battery life prediction method, system, equipment and medium

By combining DPCNN and Transformer models, the problem of extracting local and global features in the prediction of the remaining lifespan of lithium-ion batteries is solved, achieving more accurate lifespan prediction and faster training speed.

CN118938061APending Publication Date: 2024-11-12XI'AN PETROLEUM UNIVERSITY
View PDF 0 Cites 7 Cited by

Patent Information

Application Number
CN202411343317.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-25
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Existing technologies struggle to simultaneously extract local and global features from lithium-ion battery life data and efficiently capture complex correlations between long-term sequences, resulting in inaccurate predictions of the remaining lifespan of lithium-ion batteries.

Method used

The DPCNN-Transformer model, which combines a deep pyramid convolutional neural network (DPCNN) and a Transformer module, captures various features in battery data through feature extraction and self-attention mechanisms to predict the remaining lifespan of lithium-ion batteries.

Benefits of technology

It improves the accuracy and robustness of predicting the remaining lifespan of lithium-ion batteries, reduces computational load, enhances the model's adaptability under different operating conditions, and achieves faster training speed and higher prediction accuracy.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118938061A_ABST
    Figure CN118938061A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of lithium ion batteries, in particular to a lithium battery life prediction method, system and device and a medium, and the method comprises the steps: obtaining the related characteristic parameter data of the remaining service life of a lithium ion battery; inputting the relevant characteristic parameter data of the remaining service life of the lithium ion battery into a pre-trained prediction model to obtain a prediction result of the remaining service life of the lithium ion battery; the prediction model adopts a DPCNN-Transform model, the DPCNN-Transform model comprises a DPCNN module and a Transform module, and the DPCNN module is used for carrying out feature extraction on input characteristic parameter data related to the remaining service life of the lithium ion battery to obtain a feature vector; and the Transform module predicts the remaining service life of the lithium ion battery based on the feature vector. According to the method, prediction of the remaining service life of the lithium ion battery can be accurately realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of lithium-ion batteries, and in particular to a method, system, device and medium for predicting the life of a lithium battery. Background Art

[0002] Compared with other electrochemical energy storage devices, lithium-ion batteries have higher energy density, faster charge and discharge speed, longer service life and lower self-discharge rate. Therefore, they have been widely used in many fields such as energy storage power stations, daily electronic products, electric vehicles, communication base stations, etc. However, as the use time of lithium-ion batteries continues to accumulate, these batteries will inevitably age gradually, resulting in a gradual decline in their performance. This performance degradation may cause safety hazards in some cases, especially when the battery is used improperly, and may even lead to catastrophic accidents. The remaining useful life (RUL) as a key indicator can not only provide decision makers with the necessary information to formulate corresponding management measures, but also significantly improve the safety and reliability of lithium-ion batteries in practical applications. Therefore, in-depth research and development of prediction technology for the RUL of lithium-ion batteries is of great significance in practical applications.

[0003] The prediction strategies for the remaining life of batteries can be summarized into two categories: one is model-driven method, and the other is data-driven method. The core of the model-driven method is to construct a partial differential equation system to simulate the intricate electrochemical dynamic process inside the battery. However, given that battery performance is affected by many factors, there are many variables that are difficult to accurately capture through pure physical models, such as the deep mechanism of battery aging, which often leads to limitations in model-based predictions and difficulty in achieving high accuracy. The data-driven method collects and analyzes the operating data of the battery under different working conditions, and uses machine learning technology to build a RUL prediction model, thereby revealing the changing laws of battery performance and evaluating its remaining service life. Traditional machine learning algorithms, such as support vector machines (SVM), decision trees, and random forests, have been applied to the prediction of battery life. However, these algorithms have not performed well in the prediction accuracy of the remaining service life. In particular, in the face of increasingly large data sets, these traditional models have obvious limitations in deeply mining data features, efficiently utilizing information, and accurately capturing subtle changes in battery performance degradation.

[0004] In recent years, various deep learning models, especially long short-term memory networks (LSTM) and convolutional neural networks (CNN), have gradually attracted attention and favor in the field of battery life prediction. However, the inherent locality and spatial invariance of the CNN model and its various variants limit its comprehensive capture of global features, thereby weakening the effective transmission of key information between battery aging cycles. Although the LSTM model and its various variants are good at processing sequence data, their ability to capture long-term contextual dependencies is still insufficient when faced with extremely lengthy raw battery data. In addition, although many other deep learning methods have been tried to be applied to the prediction of remaining battery life, they also face many challenges and limitations when dealing with the complexity and uncertainty of battery performance degradation, making it difficult to achieve satisfactory prediction results.

[0005] Therefore, how to simultaneously extract local and global features from battery life data, and how to efficiently capture and analyze the complex correlations between long-term series to achieve accurate prediction of the remaining service life of lithium-ion batteries, are issues that need to be addressed urgently. Summary of the invention

[0006] The present invention provides a lithium battery life prediction method, system, device and medium, aiming to solve the current problem of inaccurate prediction of the remaining service life of lithium-ion batteries.

[0007] The objective of the present invention is achieved by the following technical solutions:

[0008] In a first aspect, the present invention provides a lithium battery life prediction method, comprising:

[0009] Obtain characteristic parameter data related to the remaining service life of lithium-ion batteries;

[0010] Inputting characteristic parameter data related to the remaining service life of the lithium-ion battery into a pre-trained prediction model to obtain a prediction result of the remaining service life of the lithium-ion battery;

[0011] The prediction model adopts a DPCNN-Transformer model, which includes a DPCNN module and a Transformer module; the DPCNN module is used to extract features from input characteristic parameter data related to the remaining service life of the lithium-ion battery to obtain a feature vector; and the Transformer module predicts the remaining service life of the lithium-ion battery based on the feature vector.

[0012] As a further improvement of the present invention, the characteristic parameter data related to the remaining service life of the lithium-ion battery includes at least one of a battery capacity time series sorted based on the number of charge and discharge cycles, a discharge voltage time series sorted based on the number of charge and discharge cycles, and a charging current time series sorted based on the number of charge and discharge cycles.

[0013] As a further improvement of the present invention, the characteristic parameter data related to the remaining service life of the lithium-ion battery is preprocessed data, and the preprocessing step includes normalizing the characteristic parameter data related to the remaining service life of the lithium-ion battery, and smoothing the normalized characteristic parameter data using an exponentially weighted moving average;

[0014] The formula for the smoothing process is:

[0015] EWMA(t)=α×x(t)+(1-α)×EWMA(t-1)

[0016] Where EWMA(t) represents the exponentially weighted moving average at the tth time point; X(t) represents the observed value at the tth time point; α is the attenuation factor, whose value is between 0 and 1; EWMA(t-1) represents the exponentially weighted moving average at the t-1th time point.

[0017] As a further improvement of the present invention, the DPCNN module is used to extract features from the input characteristic parameter data related to the remaining service life of the lithium-ion battery to obtain a feature vector, which specifically includes:

[0018] The characteristic parameter data related to the remaining service life of the lithium-ion battery is segmented by a sliding window to obtain a discharge voltage subsequence and a charge voltage subsequence, and the discharge voltage subsequence and the charge voltage subsequence are used as input data of the DPCNN module;

[0019] The input data is input to the DPCNN module, multi-scale feature extraction is performed through the feature pyramid convolution layer, and a battery capacity sequence with a feature vector is output.

[0020] As a further improvement of the present invention, the Transformer module is used to determine the sequence position coding information according to the battery capacity sequence output by the DPCNN module, and obtain the output matrix of the attention mechanism layer based on the sequence position coding information. After the output matrix passes through the feedforward network, the downward trend of the lithium-ion battery usage time series is learned, and the remaining service life of the lithium-ion battery is predicted based on the downward trend of the lithium-ion battery usage time series.

[0021] As a further improvement of the present invention, the remaining service life of the lithium-ion battery is predicted based on the downward trend of the lithium-ion battery usage time series, and the prediction method is:

[0022]

[0023] In the formula, represents the prediction result at time t; f is the activation function; W p and b p The weights and biases used for prediction; FHN is the output of the hidden layer of the Transformer at time t.

[0024] As a further improvement of the present invention, the loss function used by the Transformer module during training is:

[0025]

[0026] Where L(θ) is the loss function; M is the number of prediction cycles; x t It is the real value of lithium-ion battery capacity; is the predicted value of the lithium-ion battery capacity; T is the first time point, which is used to define the start and end time of the loss function calculation.

[0027] In a second aspect, the present invention provides a lithium battery life prediction system for implementing the lithium battery life prediction method as described above, the system comprising:

[0028] A data acquisition module is used to obtain characteristic parameter data related to the remaining service life of lithium-ion batteries;

[0029] A battery remaining service life prediction module is used to input characteristic parameter data related to the remaining service life of the lithium-ion battery into a pre-trained prediction model to obtain a prediction result of the remaining service life of the lithium-ion battery;

[0030] The prediction model adopts a DPCNN-Transformer model, which includes a DPCNN module and a Transformer module. The DPCNN module is used to extract features from input characteristic parameter data related to the remaining service life of the lithium-ion battery to obtain a feature vector; the Transformer module predicts the remaining service life of the lithium-ion battery based on the feature vector.

[0031] In a third aspect, the present invention also provides a computer-readable storage medium storing one or more programs, wherein the one or more programs include instructions, which, when executed by a computing device, enable the computing device to execute the above-mentioned lithium battery life prediction method.

[0032] In a fourth aspect, the present invention further provides a computing device, comprising:

[0033] One or more processors, a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include steps for executing a lithium battery life prediction method.

[0034] The beneficial effects of the present invention are as follows: the lithium battery life prediction method of the present invention accurately estimates the remaining service life of lithium-ion batteries by constructing a prediction model based on the DPCNN-Transformer network. The prediction model makes full use of the DPCNN module to effectively capture the local and global features of the input data, process the characteristic parameter data related to the remaining service life of the lithium-ion battery, and realize multi-level feature extraction of the input data, so that the prediction model has stronger robustness to slight changes or noise in the data. The Transformer module is good at processing the dependencies between sequences, which can further enhance the model's ability to capture long-term dependent information. Through the self-attention mechanism, it can automatically learn the important parts of the characteristic parameters related to the remaining service life of the battery, further enhancing the model's adaptability under different working conditions. The combination of the DPCNN module and the Transformer module significantly improves the accuracy of the RUL prediction of lithium-ion batteries.

[0035] Furthermore, the DPCNN module effectively reduces the length of the sequence processed by the subsequent Transformer module by reducing the size of the feature map, thereby reducing the amount of calculation and speeding up the model training while maintaining high prediction accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention 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 some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 A flow chart of a lithium battery life prediction method is provided for an embodiment of the present invention;

[0038] Figure 2 A schematic diagram of a lithium battery life prediction method is provided for an embodiment of the present invention;

[0039] Figure 3 The capacity degradation curve of the lithium-ion battery provided by the embodiment of the present invention;

[0040] Figure 4 A battery terminal voltage curve under discharge state of a lithium-ion battery provided by an embodiment of the present invention;

[0041] Figure 5 A current curve measured when the lithium-ion battery is in a charging state provided by an embodiment of the present invention;

[0042] Figure 6 This is a diagram of the RUL estimation results of a lithium-ion battery provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] In order to make the purpose and technical solution of the present invention clearer and easier to understand, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0044] For the understanding of the scheme of the present invention, the relevant terms are explained:

[0045] 1. DPCNN: Deep Pyramid Convolutional Neural Network;

[0046] 2. Transformer: A neural network model based on the attention mechanism, which is used to process sequence-to-sequence tasks. It captures the dependencies in the input sequence through the self-attention mechanism.

[0047] 3. EWMA: Exponentially weighted moving average is a weighted moving average technique used for smoothing time series data;

[0048] 4. Multi-head Attention: multi-head attention mechanism;

[0049] 5. Position-wise Feed-Forward Networks: Feedforward network.

[0050] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments, wherein the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0051] Example 1

[0052] like Figure 1 , Figure 2As shown, this embodiment provides a lithium battery life prediction method, which establishes a prediction model for battery life analysis through a DPCNN-Transformer network, and analyzes the battery's past data to predict the remaining battery life. The pyramid pooling layer of the DPCNN module can extract features at both the local and global levels. The Transformer module performs well in processing long sequence data, thanks to its self-attention mechanism, which enables the model to better capture long-term dependencies. Combining DPCNN and Transformer to establish a lithium-ion battery life prediction model can give full play to the unique advantages of the two modules, more comprehensively capture various features in the battery data, and improve prediction performance and accuracy. The following specific implementation methods are included.

[0053] First, obtain characteristic parameter data related to the remaining service life of the lithium-ion battery.

[0054] Among them, the characteristic parameter data related to the remaining service life of the lithium-ion battery includes at least one of a battery capacity time series sorted by the number of charge and discharge cycles, a discharge voltage time series sorted by the number of charge and discharge cycles, and a charging current time series sorted by the number of charge and discharge cycles. This embodiment adopts the optimal solution of simultaneously acquiring three types of time series data. Of course, the battery capacity time series and the discharge voltage time series sorted by the number of charge and discharge cycles, or the discharge voltage time series and the charging current time series can be used to predict the remaining service life of the lithium-ion battery, but the prediction accuracy decreases thereafter.

[0055] Specifically, Figure 3 As shown in the figure, the battery capacity time series based on the number of charge and discharge cycles is expressed as:

[0056] Cp=(c1,c2,...,c k , ..., c n )

[0057] Where k∈[1,n], n is the number of charge and discharge cycles, c k is the battery capacity at the kth charge and discharge cycle.

[0058] like Figure 4 As shown in the figure, the discharge voltage time series sorted by the number of charge and discharge cycles is expressed as:

[0059] Vt=(v1,v2,...,v p , ..., v n )

[0060] Where p∈[1,n], n is the number of charge and discharge cycles, v p It is the dynamic sequence of battery terminal voltage during the discharge process in a charge and discharge cycle.

[0061] like Figure 5 As shown in the figure, the charging current time series based on the number of charge and discharge cycles is expressed as:

[0062] Ic=(i1,i2,...,i a , ..., i n )

[0063] Among them, q∈[1,n], n is the number of charge and discharge cycles, i q It is the current dynamic sequence measured in real time during the charging process in a charge and discharge cycle.

[0064] As one example in this embodiment, the number of charge and discharge cycles is set to 100.

[0065] The initial lithium-ion battery remaining service life related characteristic parameter data obtained is pre-processed data to eliminate noise in the data. The pre-processing step includes normalizing the obtained lithium-ion battery remaining service life related characteristic parameter data and smoothing the normalized characteristic parameter data using an exponentially weighted moving average method.

[0066] Specifically, the normalization process includes: inputting a data sequence X with n points = (x1, x2, ..., x n ), normalize it to (0, 1). The normalized expression is:

[0067]

[0068] Wherein, X is the characteristic parameter related to the degradation of the remaining service life of the lithium-ion battery (i.e., the battery capacity time series Cp, the discharge voltage time series Vt or the charging current time series Ic); X0 represents the maximum value of the data sequence; and x′ is the normalized value corresponding to x.

[0069] The smoothing expression is:

[0070] EWMA(t)=α×x(t)+(1-α)×EWMA(t-1)

[0071] In the formula, EWMA(t) represents the exponentially weighted moving average at the tth time point; X(t) represents the observed value at the tth time point; α is the attenuation factor (that is, the smoothing factor), whose value is between 0 and 1; EWMA(t-1) represents the exponentially weighted moving average at the t-1th time point.

[0072] Next, the characteristic parameter data related to the remaining service life of the lithium-ion battery is input into the pre-trained prediction model to obtain the prediction result of the remaining service life of the lithium-ion battery. The prediction model adopts the DPCNN-Transformer model, which includes the DPCNN module and the Transformer r module.

[0073] The DPCNN module is used to extract features from the input lithium-ion battery remaining service life related characteristic parameter data to obtain a feature vector.

[0074] Specifically, the characteristic parameter data related to the remaining service life of the lithium-ion battery is segmented through a sliding window to obtain a discharge voltage subsequence and a charge voltage subsequence, and the discharge voltage subsequence and the charge voltage subsequence are used as input data of the DPCNN module;

[0075] The discharge voltage subsequence and the charge voltage subsequence are input into the DPCNN module, multi-scale feature extraction is performed through the feature pyramid convolution layer, and a battery capacity sequence with a feature vector is output.

[0076] The DPCNN module uses a feature pyramid convolution layer to extract multi-scale features. The parameter quantity and computational complexity of the pyramid convolution layer are estimated as follows:

[0077]

[0078] FLOPs = Para*(H×W)

[0079] Where para is the pyramid convolution parameter; FLOPs is the computational complexity; H and W represent the height and width of the convolution kernel respectively; C i is the number of input channels of the feature pyramid; is the convolution kernel size of each layer; C o1 , C o2 ..., C on For Depth The corresponding output feature dimension.

[0080] As an example in this embodiment, the data is divided into 20 subsequences, and features are extracted through the DPCNN model. The data is divided into 20 subsequences through a sliding window, and then the DPCNN module is used to extract features for each subsequence, with the discharge voltage sequence Vt and the charging current sequence Ic as module inputs, and the battery capacity sequence Cp as module outputs.

[0081] The Transformer module predicts the remaining service life of lithium-ion batteries based on the feature vector.

[0082] Specifically, the Transformer module is used to determine the sequence position encoding information according to the battery capacity sequence output by the DPCNN module, and obtain the output matrix of the attention mechanism layer based on the sequence position encoding information. After the output matrix passes through the feedforward network, the downward trend of the lithium-ion battery usage time series is learned, and the remaining service life of the lithium-ion battery is predicted based on the downward trend of the lithium-ion battery usage time series.

[0083] Specifically, first determine the sequence position encoding information, which can be expressed as:

[0084]

[0085]

[0086] Among them, PE(pos, 2i) and PE(pos, 2i+1) represent the position encoding values ​​of position pos in dimensions 2i and 2i+1 respectively; pos is the position index in the input sequence; i represents the feature dimension; d model Indicates the length of the input sequence.

[0087] In the Transformer module of this embodiment, the multi-head attention mechanism is used to improve the self-attention layer and calculate the relationship between features. Based on the input sequence position information PE(·), the query vector Q(Query), the key vector K(Key) and the value vector V(Value) are calculated, and then the output matrix of the attention mechanism layer is obtained as follows:

[0088]

[0089] head i =Attention(QW I Q , KW I K , VW I V ), i = 1, 2, ..., n

[0090] MH(Q,K,V)=Concat(head1,head2,...,head n )W o

[0091] Among them, MH(·) is the output matrix of the multi-head attention mechanism layer; d h is the dimension of the attention head; W I Q , W I K , W I VThey represent the mapping weights of the query vector, key vector, and value vector respectively; Wo is the weight of multi-head attention, and Concat is the concatenation vector.

[0092] After the multi-head attention mechanism layer, the position feedforward network is used for calculation. The position feedforward network is a fully connected feedforward network, and its calculation expression is as follows:

[0093] FFN(A)=ReLU(0,MH*W1+b1)W2+b2

[0094] Among them, MH is the position feedforward network input, that is, the output matrix of the attention mechanism layer; W1 and b1 are the weight and bias of the first hidden layer; W2 and b2 are the weight and bias of the output layer; ReLU is the activation function, which returns zero when the input is negative, otherwise it returns the input value itself.

[0095] According to the downward trend of the lithium-ion battery usage time series, the remaining service life of the lithium-ion battery is predicted. The specific prediction method used is:

[0096]

[0097] In the formula, represents the prediction result at time t; f is the activation function; W p and b p The weights and biases used for prediction; FHN is the output of the hidden layer of the Transformer at time t.

[0098] The data used for the prediction model training in this embodiment are historical characteristic parameters related to the degradation of the remaining service life of lithium-ion batteries, that is, the battery capacity time series, discharge voltage time series and charging current time series sorted by the number of charge and discharge cycles. Initialize the parameters of the prediction model, where the parameters include word embedding matrix, position encoding, weight matrix of each layer, etc. The data input to the prediction model is forward propagated through each layer of the model, and the loss function is calculated based on the output of the prediction model and the true label. The model parameters are back-propagated using the loss function, and the gradient is calculated. The optimizer is used to update the prediction model parameters according to the calculated gradient to optimize the prediction model.

[0099] The loss function used by the Transformer module during training is:

[0100]

[0101] Where L(θ) is the loss function; M is the number of prediction cycles; x t It is the real value of lithium-ion battery capacity; is the predicted value of the lithium-ion battery capacity; T is the first time point, wherein the first time point is a specific time point used to define the start and end time of the loss function calculation. The number of predicted cycles in the example in this embodiment is 500.

[0102] In order to verify the accuracy of the prediction model in this embodiment, the prediction model is evaluated using indicators such as the relative error of the remaining charge and discharge times, the root mean square error, and the mean absolute error.

[0103] Relative error of remaining charge and discharge times (RE):

[0104]

[0105] Root Mean Square Error (RMSE):

[0106]

[0107] Mean Absolute Error (MAE):

[0108]

[0109] Where K is the number of predicted cycles; x t is the actual value of capacity; is the capacity forecast value.

[0110] Based on the above evaluation indicators, this embodiment uses the prediction model constructed by the existing RNN, GRU and LSTM and the prediction model constructed by the DPCNN-Transformer of this embodiment to make predictions based on the degradation-related characteristic parameters of the remaining service life of the same lithium-ion battery, respectively, and obtains corresponding prediction results, and compares them based on RE value, MAE and RMSE, as shown in Table 1. Table 1 is the test results based on the NASA data set. It can be seen from Table 1 that the RE value of the present invention is lower than that of the comparison method by more than 0.1, and MAE and RMSE are also better than the comparison method. Among the four different prediction models, this method achieves the best effect.

[0111] Table 1 Comparison of evaluation indicators of prediction results based on NASA dataset

[0112]

[0113] like Figure 6 As shown, a comparison chart between the prediction results and the true values ​​of the method proposed in this embodiment based on the NASA data set shows that the method in this embodiment can make a good prediction of the RUL of lithium-ion batteries.

[0114] Example 2

[0115] This embodiment also provides a lithium battery life prediction system, which is used to implement the lithium battery life prediction method in embodiment 1. The system includes:

[0116] A data acquisition module is used to obtain characteristic parameter data related to the remaining service life of lithium-ion batteries;

[0117] A battery remaining service life prediction module is used to input characteristic parameter data related to the remaining service life of the lithium-ion battery into a pre-trained prediction model to obtain a prediction result of the remaining service life of the lithium-ion battery;

[0118] The prediction model adopts the DPCNN-Transformer model, which includes the DPCNN module and the Transformer module. The DPCNN module is used to extract the features of the input characteristic parameter data related to the remaining service life of the lithium-ion battery to obtain the feature vector; the Transformer module predicts the remaining service life of the lithium-ion battery based on the feature vector.

[0119] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0120] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0121] Example 3

[0122] In another embodiment of the present invention, a terminal device is provided, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, which are suitable for implementing one or more instructions, and are specifically suitable for loading and executing one or more instructions to implement corresponding method processes or corresponding functions; the processor described in the embodiment of the present invention can be used for the operation of the lithium battery life prediction method, including:

[0123] Obtain characteristic parameter data related to the remaining service life of lithium-ion batteries;

[0124] Inputting characteristic parameter data related to the remaining service life of the lithium-ion battery into a pre-trained prediction model to obtain a prediction result of the remaining service life of the lithium-ion battery;

[0125] The prediction model adopts the DPCNN-Transformer model, which includes a DPCNN module and a Transformer module; the DPCNN module is used to extract features from the input characteristic parameter data related to the remaining service life of the lithium-ion battery to obtain a feature vector; the Transformer module predicts the remaining service life of the lithium-ion battery based on the feature vector.

[0126] Example 4

[0127] In another embodiment of the present invention, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It is understandable that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and the extended storage medium supported by the terminal device, and can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that more specific examples (non-exhaustive list) of the computer-readable storage medium here include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0128] Computer readable storage media also include data signals propagated in baseband or as part of a carrier wave, which carry readable program codes. Such propagated data signals can take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable storage medium can also be any readable medium other than a readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or device. The program code contained on the readable storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above.

[0129] Program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0130] The processor may load and execute one or more instructions stored in a computer-readable storage medium to implement the corresponding steps of the lithium battery life prediction method in the above embodiment; the processor may load and execute the following steps:

[0131] Obtain characteristic parameter data related to the remaining service life of lithium-ion batteries;

[0132] Inputting characteristic parameter data related to the remaining service life of the lithium-ion battery into a pre-trained prediction model to obtain a prediction result of the remaining service life of the lithium-ion battery;

[0133] The prediction model adopts the DPCNN-Transformer model, which includes a DPCNN module and a Transformer module; the DPCNN module is used to extract features from the input characteristic parameter data related to the remaining service life of the lithium-ion battery to obtain a feature vector; the Transformer module predicts the remaining service life of the lithium-ion battery based on the feature vector.

[0134] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software 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 the present invention.

[0135] In the embodiments provided by the present invention, it should be understood that the disclosed systems / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are only schematic. For example, the division of the modules or units 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, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0136] 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.

[0137] In addition, each functional unit in each embodiment of the present invention 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 unit may be implemented in the form of hardware or in the form of software functional units.

Claims

1. A lithium battery life prediction method, characterized in that: include: Obtain characteristic parameter data related to the remaining service life of lithium-ion batteries; Inputting characteristic parameter data related to the remaining service life of the lithium-ion battery into a pre-trained prediction model to obtain a prediction result of the remaining service life of the lithium-ion battery; The prediction model adopts a DPCNN-Transformer model, which includes a DPCNN module and a Transformer module; the DPCNN module is used to extract features from input characteristic parameter data related to the remaining service life of the lithium-ion battery to obtain a feature vector; and the Transformer module predicts the remaining service life of the lithium-ion battery based on the feature vector.

2. The lithium battery life prediction method according to claim 1, characterized in that: The characteristic parameter data related to the remaining service life of the lithium-ion battery includes at least one of a battery capacity time series sorted based on the number of charge and discharge cycles, a discharge voltage time series sorted based on the number of charge and discharge cycles, and a charging current time series sorted based on the number of charge and discharge cycles.

3. The lithium battery life prediction method according to claim 2, characterized in that: The characteristic parameter data related to the remaining service life of the lithium-ion battery is preprocessed data, and the preprocessing step includes normalizing the characteristic parameter data related to the remaining service life of the lithium-ion battery, and smoothing the normalized characteristic parameter data using an exponentially weighted moving average; The formula for the smoothing process is: EWMA(t)=α×x(t)+(1-α)×EWMA(t-1) Where EWMA(t) represents the exponentially weighted moving average at the tth time point; X(t) represents the observed value at the tth time point; α is the attenuation factor, whose value is between 0 and 1; EWMA(t-1) represents the exponentially weighted moving average at the t-1th time point.

4. The lithium battery life prediction method according to claim 1, characterized in that: The DPCNN module is used to extract features from the inputted characteristic parameter data related to the remaining service life of the lithium-ion battery to obtain a feature vector, which specifically includes: The characteristic parameter data related to the remaining service life of the lithium-ion battery is segmented by a sliding window to obtain a discharge voltage subsequence and a charge voltage subsequence, and the discharge voltage subsequence and the charge voltage subsequence are used as input data of the DPCNN module; The input data is input to the DPCNN module, multi-scale feature extraction is performed through the feature pyramid convolution layer, and a battery capacity sequence with a feature vector is output.

5. The lithium battery life prediction method according to claim 1, characterized in that: The Transformer module is used to determine the sequence position encoding information according to the battery capacity sequence output by the DPCNN module, and obtain the output matrix of the attention mechanism layer based on the sequence position encoding information. After the output matrix passes through the feedforward network, the downward trend of the lithium-ion battery usage time series is learned, and the remaining service life of the lithium-ion battery is predicted according to the downward trend of the lithium-ion battery usage time series.

6. The lithium battery life prediction method according to claim 5, characterized in that: The remaining service life of the lithium-ion battery is predicted based on the downward trend of the lithium-ion battery usage time series, and the prediction method is: In the formula, represents the prediction result at time t; f is the activation function; W p and b p The weights and biases used for prediction; FHN is the output of the hidden layer of the Transformer at time t.

7. The lithium battery life prediction method according to claim 5, characterized in that: The loss function used by the Transformer module during training is: Where L(θ) is the loss function; M is the number of prediction cycles; x t It is the real value of lithium-ion battery capacity; is the predicted value of the lithium-ion battery capacity; T is the first time point, which is used to define the start and end time of the loss function calculation.

8. A lithium battery life prediction system, used to implement the lithium battery life prediction method according to any one of claims 1 to 7, characterized in that: The system includes: Data acquisition module, to obtain characteristic parameter data related to the remaining service life of lithium-ion batteries; A battery remaining service life prediction module is used to input characteristic parameter data related to the remaining service life of the lithium-ion battery into a pre-trained prediction model to obtain a prediction result of the remaining service life of the lithium-ion battery; The prediction model adopts a DPCNN-Transformer model, which includes a DPCNN module and a Transformer module. The DPCNN module is used to extract features from input characteristic parameter data related to the remaining service life of the lithium-ion battery to obtain a feature vector; the Transformer module predicts the remaining service life of the lithium-ion battery based on the feature vector.

9. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions, which, when executed by a computing device, enable the computing device to execute the lithium battery life prediction method described in any one of claims 1 to 7.

10. A computing device, characterized in that: include: One or more processors, a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include steps for executing the lithium battery life prediction method described in any one of claims 1 to 7.

Citation Information

Cited By

  • Oxygen lance service life prediction method, system and device and readable storage medium

    CN119493964A

  • Method, device and equipment for predicting remaining service life of battery, medium and product

    CN119575222A

  • Remaining service life prediction method and device and electronic equipment

    CN119959768A

  • Remaining service life prediction method, device and electronic equipment

    CN119959768B

  • Life prediction method based on lithium battery characteristic relation and convolutional Transform neural network

    CN119989940A