Lithium ion battery remaining service life prediction method and system based on multi-dimensional data hybrid deep neural network

Abstract

The invention discloses a lithium ion battery remaining service life prediction method and system based on a multi-dimensional data hybrid deep neural network, and relates to the technical field of lithium ion batteries. The problem that the existing method for predicting the remaining service life of the lithium ion battery based on deep learning basically needs data of a plurality of battery cycles, and data storage of a battery management system is challenged is solved. The method comprises the following steps: carrying out a charge-discharge aging experiment on a battery to obtain a battery performance index and observation data, establishing a lithium battery original aging data set and carrying out preprocessing; sorting and extracting the data set through feature engineering; building a hybrid deep neural network model; training the hybrid deep neural network model by adopting k-fold cross validation; and after the k-fold cross validation training is completed, loading the model and calculating the output of each fold of model on the data of the test set, and averaging to predict the remaining service life of the lithium ion battery. The method is applied to the new-energy automobile field.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium-ion batteries, and in particular to a method for predicting the remaining service life of lithium-ion batteries based on a multi-dimensional data hybrid deep neural network. Background Art

[0002] Lithium-ion batteries have become a widely used power source in new energy vehicles due to their high energy density, wide operating temperature range, long cycle life and environmental friendliness. Therefore, the performance of lithium-ion batteries directly determines the core indicators of new energy vehicles such as cruising range, charging time and safety performance, and is one of the important factors affecting the development of new energy vehicles. However, during the long-term operation of lithium-ion batteries, as the number of charge and discharge times increases and the operating temperature changes, problems such as battery aging and performance degradation will gradually occur. When the performance drops to a certain level, the internal structure of the battery collapses, and various safety problems such as local short circuits can lead to serious safety accidents. Therefore, during the normal operation of the battery, it is necessary to accurately estimate the remaining service life of the battery to ensure the safe and stable operation of the battery. At present, although researchers have done a lot of work in battery status assessment, prediction and fault diagnosis, due to the complex working mechanism of lithium-ion batteries themselves and their strong nonlinear characteristics, the main electrochemical reactions inside them, such as solid-phase diffusion, liquid-phase diffusion, solid-liquid interface charge transfer, growth and decomposition of solid-phase electrolyte interface films, cannot be accurately described by ordinary differential equations. The operating conditions of new energy vehicles are complex and changeable. For example, the mutual coupling of various external factors such as electricity, heat, and mechanics will have a significant impact on the battery, resulting in serious impact on the accurate estimation of the remaining service life of the battery. Therefore, it is often difficult to achieve high-precision, wide-adaptability, and high-robust life prediction of lithium-ion batteries. In general, the prediction of the remaining service life (RUL) of lithium-ion batteries can be divided into model-based methods and data-driven methods according to technical methods.

[0003] Model-based methods usually require highly accurate models, are easily disturbed by external conditions, and respond slowly to load changes. Therefore, these methods are limited by the complexity of the model, limited robustness, dynamic adaptability, and prediction accuracy. In addition, the constructed battery aging models are usually only targeted at specific application environments or battery types, and have limited application scope. More importantly, these methods usually perform poorly in predicting the early life of lithium-ion batteries.

[0004] The data-driven model estimates the health status of the battery by using advanced algorithms such as Gaussian process regression, support vector machine and machine learning to establish a nonlinear relationship model between data and battery status based on signals such as voltage and current collected from the battery management system. The advantage of this method is that it does not rely on the complex physical process of the battery aging mechanism. As an extended field of machine learning, deep learning can automatically extract effective features from large-scale data sets by constructing a multi-layer neural network architecture, showing excellent fitting ability for complex systems. In recent years, deep learning technology has received increasing attention in the prediction of lithium-ion battery life due to its superiority. In particular, battery capacity degradation data presents nonlinear time series characteristics. Recurrent neural network (RNN) is an effective tool for processing such data and predicting future series, and its applicability and efficiency have been verified by many researchers. Convolutional neural network (CNN) is an efficient feature extraction technology that simplifies the complexity of feature extraction compared with RNN and its derivative algorithms. CNN can effectively alleviate the overfitting problem through its unique structure, especially when the amount of data is limited or when facing capacity regeneration. Therefore, CNN has become a common method in the field of battery life prediction and has been widely used in related research in recent years.

[0005] However, deep learning-based methods generally require data from multiple battery cycles, which poses a great challenge to the data storage of battery management systems, especially when continuously monitoring and processing large amounts of battery data in real time. This requires more efficient data compression and processing strategies to optimize storage and computing resources. Summary of the invention

[0006] The present invention aims at the problem that the existing methods for predicting the remaining service life of lithium-ion batteries based on deep learning basically require data of multiple battery cycles, which poses a great challenge to the data storage of the battery management system. A method for predicting the remaining service life of lithium-ion batteries based on a multi-dimensional data hybrid deep neural network is proposed. The method comprises:

[0007] Step S1: Perform charge and discharge aging experiments on the battery to obtain performance indicators of various cycles of the battery and battery cycle observation data, establish an original aging data set for the lithium battery, and pre-process the aging data set;

[0008] Step S2: Divide the preprocessed data set into a training set for model training and a test set for model evaluation, and organize and extract the data set through feature engineering;

[0009] Step S3: Building a hybrid deep neural network model based on one-dimensional convolutional neural network, two-dimensional convolutional neural network and Transformer;

[0010] Step S4: Use k-fold cross validation to train the hybrid deep neural network model. When the loss function of the training model no longer continues to decrease, save the model and record it as the optimal model of the fold;

[0011] Step S5: After the k-fold cross-validation training is completed, the model is loaded and the output of each fold of the model on the test set data is calculated, and the average is taken to predict the remaining service life of the lithium-ion battery.

[0012] Furthermore, a preferred method is proposed, wherein the performance indicators of each battery cycle in step S1 include discharge capacity, charging time, and internal resistance; and the battery cycle observation data includes voltage, current, temperature, and capacity during the battery charge and discharge cycle.

[0013] Furthermore, a preferred method is proposed, in which the charge and discharge aging experiment in step S1 is a multi-stage charging and constant current discharge process at a constant temperature, including: first using a constant current to charge the battery to a specific state of charge, then using another constant current to charge the battery in the charged state to a cut-off voltage, and then using a constant voltage to charge to the cut-off current of the battery, and discharging with the same current to the cut-off voltage.

[0014] Furthermore, a preferred method is proposed, in which the preprocessing in step S1 includes: using a linear interpolation method to remove outliers, and normalizing the data without outliers.

[0015] Furthermore, a preferred method is proposed, in which the feature engineering in step S2 includes:

[0016] The four curves of current, voltage, temperature and capacity of each cycle are linearly interpolated into four curves with 100 data points according to the time series. The four curves form a set of 4×100 two-dimensional data matrices; the three values ​​of discharge capacity, charging time and battery internal resistance of each cycle are combined into a 1×3 one-dimensional data matrix.

[0017] Furthermore, a preferred method is proposed, wherein step S3 comprises:

[0018] The one-dimensional data matrix is ​​input into a three-layer one-dimensional convolutional neural network to extract the battery aging information from the one-dimensional scalar features, and then the output of the one-dimensional convolutional neural network is processed through a fully connected layer;

[0019] The two-dimensional data matrix is ​​input into a three-layer two-dimensional convolutional neural network connected front and back, and the output of the two-dimensional convolutional neural network is further extracted through the Transformer network;

[0020] The output of the Transformer network and the fully connected layer are integrated and then output through two fully connected layers. The output obtained is the estimated value of the remaining service life of the lithium-ion battery.

[0021] The processing flow of the Transformer network is:

[0022] The data first passes through the position encoding module to preserve the time step information of the data;

[0023] The time series data is processed through a multi-head self-attention mechanism to calculate the correlation of each time step in the sequence. The self-attention mechanism performs weighted aggregation of the input features of each time step through a weight matrix.

[0024] Furthermore, a preferred method is proposed, in which the k-fold cross validation in step S4 refers to using 10-fold cross validation.

[0025] Based on the same inventive concept, the present invention also proposes a lithium-ion battery remaining service life prediction system based on a multi-dimensional data hybrid deep neural network, the system comprising:

[0026] The charge-discharge aging experiment unit is used to perform charge-discharge aging experiments on the battery to obtain the performance indicators of each battery cycle and battery cycle observation data, establish the original aging data set of lithium batteries, and pre-process the aging data set;

[0027] A data extraction unit is used to divide the preprocessed data set into a training set for model training and a test set for model evaluation, and to organize and extract the data set through feature engineering;

[0028] Model building unit, used to build a hybrid deep neural network model based on one-dimensional convolutional neural network, two-dimensional convolutional neural network and Transformer;

[0029] The training unit is used to train the hybrid deep neural network model using k-fold cross validation. When the loss function of the training model no longer continues to decrease, the saved model is recorded as the optimal model of the fold;

[0030] The prediction unit is used to load the model and calculate the output of each fold model on the test set data after the k-fold cross-validation training is completed, and then take the average to predict the remaining service life of the lithium-ion battery.

[0031] Based on the same inventive concept, the present invention also proposes a computer device, including a memory and a processor, wherein the memory stores a computer program. When the processor runs the computer program stored in the memory, the processor executes a method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network as described in any one of the above items.

[0032] Based on the same inventive concept, the present invention also proposes a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of a method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network as described above are executed.

[0033] The present invention is beneficial in that:

[0034] The present invention solves the problem that existing methods for predicting the remaining useful life of lithium-ion batteries based on deep learning basically require data from multiple battery cycles, which poses a great challenge to the data storage of the battery management system.

[0035] The present invention proposes a method for predicting the remaining service life of lithium-ion batteries based on a multi-dimensional data hybrid deep neural network, and designs a hybrid deep neural network model based on a one-dimensional convolutional neural network, a two-dimensional convolutional neural network and a Transformer, which can effectively utilize one-dimensional scalar features and two-dimensional vector features. Through forward propagation, the two-dimensional convolutional neural network effectively extracts the time series data features of the battery, and the Transformer uses the self-attention mechanism to calculate the correlation between each time step in the input sequence, effectively captures the global dependency, and further enhances the feature expression ability. The one-dimensional convolutional neural network can directly capture the key information related to battery aging from the key performance parameters of the battery. This multi-branch design ensures the synergy between different data dimensions and fully explores the potential aging characteristics. The multi-branch, multi-dimensional data fusion method can more accurately extract features related to the remaining life of the battery, thereby accurately predicting the remaining life of the battery, thereby assisting the battery management system to effectively manage and control the battery.

[0036] Compared with the traditional method that relies on multiple cycle data, the method proposed in the present invention can more efficiently identify and extract the aging characteristics of lithium-ion batteries because it uses a multi-branch multi-dimensional hybrid neural network. It only needs one cycle of charge and discharge data to accurately predict the remaining service life of the battery, thereby significantly reducing the demand for data storage in the battery management system. This design not only improves data processing efficiency, but also enhances the adaptability of the model, enabling it to maintain high performance when the amount of data is limited.

[0037] The present invention is applied to the field of new energy vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 A flow chart of a method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network according to the first embodiment;

[0039] Figure 2This is a schematic diagram of the prediction of the remaining service life of a short-life lithium-ion battery according to the eleventh embodiment, wherein the horizontal axis represents the actual service life and the vertical axis represents the predicted service life;

[0040] Figure 3 A schematic diagram of the prediction of the remaining service life of a long-life lithium-ion battery according to the eleventh embodiment, wherein the horizontal axis represents the actual service life and the vertical axis represents the predicted service life;

[0041] Figure 4 Schematic diagram of error distribution of the remaining service life prediction of all lithium-ion batteries according to the eleventh embodiment, wherein the horizontal axis represents the test set number and the vertical axis represents the prediction error;

[0042] Figure 5 A schematic diagram of the prediction of the remaining service life of a comparative example short-life lithium-ion battery according to the eleventh embodiment, wherein the horizontal axis represents the actual service life and the vertical axis represents the predicted service life;

[0043] Figure 6 A schematic diagram of the prediction of the remaining service life of a comparative example long-life lithium-ion battery according to the eleventh embodiment, wherein the horizontal axis represents the actual service life and the vertical axis represents the predicted service life;

[0044] Figure 7 This is a schematic diagram of the error distribution of the remaining service life prediction of all lithium-ion batteries in the comparative example described in Implementation Example 11, wherein the horizontal axis represents the test set number and the vertical axis represents the prediction error. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments.

[0046] Implementation method 1, see Figure 1 The present embodiment describes a method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network, the method comprising:

[0047] Step S1: Perform charge and discharge aging experiments on the battery to obtain performance indicators of various cycles of the battery and battery cycle observation data, establish an original aging data set for the lithium battery, and pre-process the aging data set;

[0048] Step S2: Divide the preprocessed data set into a training set for model training and a test set for model evaluation. Batteries can be randomly allocated into training sets and test sets in a ratio of 3:1 according to battery life, and the data set is sorted and extracted through feature engineering;

[0049] Step S3: Build a hybrid deep neural network model based on one-dimensional convolutional neural network, two-dimensional convolutional neural network and Transformer, and summarize and output information through a fully connected layer;

[0050] Step S4: Use k-fold cross validation to train the hybrid deep neural network model. When the loss function of the training model no longer continues to decrease, save the model and record it as the optimal model of the fold;

[0051] Step S5: After the k-fold cross-validation training is completed, the model is loaded and the output of each fold of the model on the test set data is calculated, and the average is taken to predict the remaining service life of the lithium-ion battery.

[0052] Compared with the traditional method that relies on multiple cycle data, the method for predicting the remaining service life of lithium-ion batteries based on a multi-dimensional data hybrid deep neural network proposed in this embodiment can more efficiently identify and extract the aging characteristics of lithium-ion batteries because it uses a multi-branch multi-dimensional hybrid neural network. It only needs one cycle of charge and discharge data to accurately predict the remaining service life of the battery, thereby significantly reducing the demand for data storage in the battery management system. This design not only improves data processing efficiency, but also enhances the adaptability of the model, enabling it to maintain high performance when the amount of data is limited.

[0053] At the same time, the multi-branch, multi-dimensional data fusion hybrid deep neural network designed in this embodiment can more efficiently identify and extract the aging characteristics of lithium-ion batteries. Through one-dimensional convolutional neural networks and two-dimensional convolutional neural networks, the key performance parameters and time series data of the battery can be processed respectively, effectively improving the feature expression ability of the hybrid deep neural network model.

[0054] Furthermore, the method for predicting the remaining service life of lithium-ion batteries proposed in this embodiment combines the advantages of convolutional neural networks (CNN) and Transformer. Transformer captures the correlation between each time step in the input sequence through the self-attention mechanism, enhancing the understanding of global dependencies. This enables the model to maintain high performance even when the amount of data is limited. The stability and accuracy of the hybrid deep neural network model can be ensured through k-fold cross-validation training and multi-branch design. The optimal model of each fold can integrate the characteristics of different data and improve the reliability of the prediction. The multi-dimensional data fusion design ensures the synergy between different data dimensions, and can fully explore the potential features related to battery aging, thereby improving the accuracy of the prediction of the remaining service life.

[0055] Implementation method 2. This implementation method is a further limitation of the method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network described in implementation method 1. The performance indicators of each battery cycle in step S1 include discharge capacity, charging time, and internal resistance; the battery cycle observation data includes voltage, current, temperature, and capacity during the battery charge and discharge cycle.

[0056] In this embodiment, performance indicators such as discharge capacity, charging time and internal resistance are introduced so that the model can more comprehensively reflect the health status and performance degradation of the battery. These indicators help to capture the aging characteristics of the battery more accurately. The battery cycle observation data includes multiple observation data such as voltage, current, temperature and capacity, which can provide detailed information about the battery under different working conditions and enhance the model's understanding of battery behavior. By comprehensively considering multiple performance indicators and observation data, the model can identify and capture more complex feature relationships. For example, changes in voltage and temperature may be directly related to changes in internal resistance, and comprehensive analysis helps to reveal the potential mechanism of battery aging. By considering different charge and discharge cycle conditions (such as different voltages and temperatures), the model has stronger adaptability and can maintain good prediction performance in different working environments and conditions. This is particularly important for battery management systems in practical applications.

[0057] Implementation method three. This implementation method is a further limitation of the method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network described in implementation method one. The charge and discharge aging experiment in step S1 is a multi-stage charging and constant current discharge process under constant temperature, including: first using a constant current to charge the battery to a specific state of charge, then using another constant current to charge the battery in the charged state to a cut-off voltage, and then using a constant voltage to charge to the cut-off current of the battery, and discharging with the same current to the cut-off voltage.

[0058] In this embodiment, the battery state of charge (SOC) represents the ratio of the remaining capacity of the battery after it has been used for a period of time or has been left unused for a long time to the capacity of the battery in a fully charged state. The specific state of charge described in this embodiment is 60%, 70%, and 80%. These three specific states of charge.

[0059] In this embodiment, through the process of multi-stage charging and constant current discharge, the charging state of the battery can be accurately controlled to provide richer battery behavior data. This helps to more accurately model the performance degradation characteristics of the battery and improve the accuracy of the prediction. The multi-stage charging and discharging process can better capture the dynamic characteristics of the battery under different working conditions. By analyzing the performance of the battery under different states of charge, the key factors affecting the battery life can be identified. Using different charging currents and states of charge, the experiment can evaluate the impact of different charging strategies on battery performance, thereby providing a basis for optimizing battery use and charging schemes. At the same time, the hybrid deep neural network model has the ability of adaptive learning and can continuously optimize model parameters as new data is input. This flexibility enables the prediction model to adapt to different types and brands of lithium-ion batteries, increasing its wide application. Through accurate prediction methods, the reliance on a large amount of experimental data can be reduced, and the cost of research and development and testing can be reduced. Battery life prediction can be carried out in the design stage, saving time and resources.

[0060] Implementation method 4. This implementation method is a further limitation of the method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network described in implementation method 1. The preprocessing in step S1 includes: using a linear interpolation method to remove outliers, and normalizing the data without abnormalities, that is, standardizing the data without abnormalities to the "0-1" range.

[0061] In this embodiment, linear interpolation can be used to effectively fill in missing values ​​or outliers in the data to ensure the integrity of the data set. This helps to avoid deviations introduced by abnormal data, thereby improving the prediction accuracy of the model. Removing outliers can improve the overall quality of the data set, making the data input to the model more consistent and reliable, and helping to improve the learning effect and generalization ability of the model. Further, the data without abnormalities is standardized to the "0-1" interval so that data with different features are on the same scale, avoiding model training problems caused by differences in the range of characteristic values. This is particularly important for deep neural networks because it can speed up convergence and improve training efficiency. Through the preprocessing step, the model's sensitivity to fluctuations and noise in the input data is reduced, thereby enhancing the robustness of the model in practical applications and improving the stability of the prediction results.

[0062] Embodiment 5: This embodiment further limits the method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network described in embodiment 1. The feature engineering in step S2 includes:

[0063] The four curves of current, voltage, temperature, and capacity of each cycle are linearly interpolated into four curves with 100 data points according to the time series, and the four curves form a set of 4×100 two-dimensional data matrices. The three values ​​of discharge capacity, charging time, and battery internal resistance of each cycle are combined into a 1×3 one-dimensional data matrix.

[0064] In this embodiment, the current, voltage, temperature and capacity curves of each cycle are interpolated to the same number of data points (100), which helps to eliminate data inconsistencies between different cycles, making the input data more standardized and convenient for subsequent processing and analysis. The linear interpolation method used in this implementation equation can effectively retain the trend and change of the original data, so that the hybrid deep neural network model can capture the dynamic characteristics of the battery during the charging and discharging process, and avoid information loss due to sparse data points. Furthermore, by forming a 4×100 two-dimensional data matrix of 4 curves, the hybrid deep neural network model can simultaneously consider the mutual influence of multiple features, thereby improving the understanding of battery performance. The introduction of this multidimensional data can enhance the learning ability of the hybrid deep neural network model for complex relationships. Combining the discharge capacity, charging time and internal resistance into a 1×3 one-dimensional data matrix makes the subsequent feature extraction and modeling process more concise and efficient, which helps to reduce the computational complexity and improve the training speed of the hybrid deep neural network model.

[0065] At the same time, combining time series data with static features (such as discharge capacity, charging time, and internal resistance) allows the hybrid deep neural network model to consider both dynamic changes and static characteristics, thereby enhancing its learning and generalization capabilities. Deep neural networks usually require high-dimensional feature inputs. Organizing data in this matrix form can better meet the input requirements of the hybrid deep neural network model and improve the performance of the model.

[0066] Furthermore, data organized in matrix form can be more easily visualized, which helps analyze the performance of the battery under different charge and discharge states.

[0067] Embodiment 6: This embodiment further limits the method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network described in embodiment 5, and step S3 includes:

[0068] The one-dimensional data matrix is ​​input into a three-layer one-dimensional convolutional neural network to extract the battery aging information from the one-dimensional scalar features, and then the output of the one-dimensional convolutional neural network is processed through a fully connected layer;

[0069] The two-dimensional data matrix is ​​input into a three-layer two-dimensional convolutional neural network connected front and back, and the output of the two-dimensional convolutional neural network is further extracted through the Transformer network;

[0070] The output of the Transformer network and the fully connected layer are integrated and then output through two fully connected layers. The output obtained is the estimated value of the remaining service life of the lithium-ion battery.

[0071] The processing flow of the Transformer network is:

[0072] The data first passes through the position encoding module to preserve the time step information of the data;

[0073] The time series data is processed through a multi-head self-attention mechanism to calculate the correlation of each time step in the sequence. The self-attention mechanism uses a weight matrix to perform weighted aggregation of the input features of each time step. This mechanism ensures that the model can focus on the long-range dependencies between any two points in the time series. Subsequently, through a feedforward neural network and normalization operation, deep features are extracted and redundant noise is reduced. After multiple layers of this structure, the model finally outputs a rich feature representation for subsequent life prediction.

[0074] The hybrid deep neural network model described in this embodiment is a multi-branch feature extraction structure. This structure effectively captures the multi-dimensional features of the lithium-ion battery aging process by splitting the one-dimensional data and the two-dimensional data. Specifically: The hybrid deep neural network model includes two branches:

[0075] Branch 1: The key performance indicators of the battery (internal resistance, discharge capacity and charging time) are processed through a one-dimensional convolutional neural network (CNN1d) to extract aging characteristic parameters at different cycle scales. The design of this branch enables the network to quickly analyze the performance changes of the battery at different cycle stages.

[0076] Branch 2: Processing cyclic observation data (current, voltage, temperature and capacity curves) through a two-dimensional convolutional neural network (CNN2d). This part of the data reflects the more complex spatiotemporal characteristics that can be reflected within the cycle during the battery aging process. CNN2d can very effectively extract the local features of this data. The Transformer layer after CNN2d can further enhance the ability to capture the global features of the data through its self-attention mechanism, focusing on the important features of different cycle stages during the battery aging process, thereby effectively improving the accuracy and robustness of battery life prediction.

[0077] Therefore, the hybrid deep neural network model proposed in this embodiment can realize the life prediction of lithium-ion batteries with only one cycle of data input through the powerful feature extraction capability of multi-dimensional data. Compared with the traditional method, the method proposed in this embodiment reduces the dependence on a large amount of historical data, greatly improving the calculation efficiency and flexibility in practical applications.

[0078] Therefore, the core innovation of this implementation is not the change in the basic network level, but the design of a multi-branch feature extraction architecture, thereby utilizing the powerful feature extraction capabilities of CNN and Transformer to achieve more efficient and accurate life prediction.

[0079] The advantages of this embodiment are explained in conjunction with the first embodiment. Traditional methods usually require data from multiple battery cycles for training, which poses a great challenge to the data storage of the battery management system. The hybrid deep neural network model designed in this embodiment only requires one cycle of charge and discharge data to accurately predict the remaining service life of the battery, significantly reducing the data storage requirements.

[0080] Implementation method seven: This implementation method is a further limitation of the method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network described in implementation method one, and the k-fold cross validation in step S4 refers to using 10-fold cross validation.

[0081] In this embodiment, by dividing the data into 10 subsets, using 9 training sets and 1 validation set each time, the dependence of the hybrid deep neural network model on specific data partitions can be effectively reduced, and the stability and reliability of the prediction results can be improved. Through 10-fold cross-validation, the training data can be more fully utilized to ensure that each data point can play a role in model training and validation, thereby improving the efficiency of data use. Through multiple training and validation, the performance of the hybrid deep neural network model on unseen data can be better evaluated, thereby reducing the risk of overfitting of the model and ensuring its generalization ability. At the same time, the 10-fold cross-validation used in this embodiment is particularly effective when the sample size is small, and can maximize the use of limited data resources to help build a more accurate prediction model.

[0082] Embodiment 8: A lithium-ion battery remaining service life prediction system based on a multi-dimensional data hybrid deep neural network described in this embodiment, the system includes:

[0083] The charge-discharge aging experiment unit is used to perform charge-discharge aging experiments on the battery to obtain the performance indicators of each battery cycle and battery cycle observation data, establish the original aging data set of lithium batteries, and pre-process the aging data set;

[0084] A data extraction unit is used to divide the preprocessed data set into a training set for model training and a test set for model evaluation, and to organize and extract the data set through feature engineering;

[0085] The model building unit is used to build a hybrid deep neural network model based on one-dimensional convolutional neural network, two-dimensional convolutional neural network and Transformer, and to aggregate and output information through a fully connected layer;

[0086] The training unit is used to train the hybrid deep neural network model using k-fold cross validation. When the loss function of the training model no longer continues to decrease, the saved model is recorded as the optimal model of the fold;

[0087] The prediction unit is used to load the model and calculate the output of each fold model on the test set data after the k-fold cross-validation training is completed, and then take the average to predict the remaining service life of the lithium-ion battery.

[0088] Embodiment 9. A computer device described in this embodiment includes a memory and a processor, wherein the memory stores a computer program. When the processor runs the computer program stored in the memory, the processor executes a method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network as described in any one of embodiments 1 to 7.

[0089] Embodiment 10. A computer-readable storage medium described in this embodiment stores a computer program, and when the computer program is executed by a processor, the steps of a method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network as described in any one of embodiments 1 to 7 are executed.

[0090] Implementation Method 11: See Figures 2 to 7 This embodiment provides a specific example of the method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network described in the first embodiment, and is also used to explain the second to seventh embodiments, specifically:

[0091] Step 1: At (30±1)℃, the 18650 lithium-ion battery of lithium iron phosphate was subjected to charge and discharge aging experiments. The charging protocol is: first charge at 3C, 4C, 5C, and 6C rates to 60%, 70%, and 80% of the rated capacity, where 10 batteries are charged to three different states of charge at each charging rate, totaling 120 batteries. Then charge at 1C constant current to 3.65V, and then charge at constant voltage to a cut-off current of 1 / 20C; the discharge protocol is: all batteries are discharged to 2.0V at a 4C rate. The battery termination condition is that the rated capacity drops to 80% of the initial capacity. At the end of each cycle, record the discharge capacity, charging time, internal resistance and other battery performance indicators of the battery in this cycle, and collect the change curves of battery current, voltage, temperature, capacity and other parameters over time during the cycle process, obtain the observation data of each battery in each cycle, and establish the original aging data set of lithium batteries. And replace the abnormal values ​​collected during the test with the mean before and after, and perform normalization preprocessing on the data without abnormalities.

[0092] Step 2: Divide the data set obtained by preprocessing in step 1 into a training set for model training and a test set for model evaluation in a ratio of 3:1 based on the life value of each battery. Then, the four curves of current, voltage, temperature, and capacity of each cycle are linearly interpolated into four curves with 100 data points according to the time series. The four curves form a set of 4×100 two-dimensional data matrices. The three values ​​of discharge capacity, charging time, and battery internal resistance of each cycle are combined into a 1×3 one-dimensional data matrix. The remaining service life of each cycle of the battery is used as a label.

[0093] Step 3: Build a hybrid deep neural network model based on one-dimensional convolutional neural network, two-dimensional convolutional neural network and Transformer, and summarize and output information through the fully connected layer. The specific method is as follows:

[0094] The two-dimensional data matrix obtained in step 2 is input into the three-layer two-dimensional convolutional neural network connected front and back, and the output of the two-dimensional convolutional neural network is further extracted through the Transformer network. The one-dimensional data matrix obtained in step 2 is input into the three-layer one-dimensional convolutional neural network, and the output of the one-dimensional convolutional neural network is processed through the fully connected layer. After the output of the above-mentioned Transformer and fully connected layers is integrated and processed, and then passed through two fully connected layers, the output of the hybrid deep neural network model is the estimated value of the remaining service life of the lithium-ion battery.

[0095] Step 4: Use 10-fold cross validation to train the constructed hybrid deep neural network model. When the loss function of the training model no longer continues to decrease, save the model as the optimal model of that fold. After the 10-fold cross validation training is completed, load the model and calculate the output of each fold model on the test set data. The average is used to predict the remaining service life of the lithium-ion battery. The life prediction value is as shown in the attached figure. Figure 2 , Attachment Figure 3 The error distribution diagram is shown in the attached Figure 4 shown.

[0096] In this embodiment, Gaussian process regression is used to predict the remaining service life of the lithium-ion battery, which specifically includes the following steps:

[0097] Step 1 and step 2 are completely consistent with the steps in the embodiment, using the same data set and the same division of the training set and the test set.

[0098] Step 3: Build a lithium-ion battery life prediction model based on Gaussian process regression. The specific method is as follows: reduce the dimension of the two-dimensional data matrix obtained in step 2 and connect it with the one-dimensional data matrix and input it into the Gaussian process regression model. The output of the model is the estimated value of the remaining service life of the lithium-ion battery.

[0099] Step 4: Train the constructed Gaussian process regression model. When the loss of the training model no longer decreases continuously, extract the best performing model to predict the remaining service life of the lithium-ion battery. The life prediction value is as shown in the attached figure. Figure 5 , Attachment Figure 6 The error distribution diagram is shown in the attached Figure 7 shown.

[0100] According to the prediction results of the present invention and the comparative example, it can be clearly seen that the method for predicting the remaining service life of a lithium-ion battery proposed by the present invention significantly improves the accuracy and robustness of the prediction of the remaining service life of a lithium-ion battery. Features related to the remaining service life of the battery can be extracted more accurately, thereby accurately predicting the remaining service life of the battery, thereby assisting the battery management system to effectively manage and control the battery.

[0101] Those skilled in the art will appreciate that the embodiments of the present disclosure may be provided as methods, systems or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, the present disclosure may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0102] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product including an instruction device, which implements the functions specified in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0103] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0104] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure rather than to limit its protection scope. Although the present disclosure has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that after reading the present disclosure, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the disclosed claims to be approved.

Claims

1. A method for predicting the remaining service life of a lithium-ion battery based on a multidimensional data hybrid deep neural network, characterized in that: The method comprises: Step S1: Perform charge and discharge aging experiments on the battery to obtain performance indicators of various cycles of the battery and battery cycle observation data, establish an original aging data set for the lithium battery, and pre-process the aging data set; Step S2: Divide the preprocessed data set into a training set for model training and a test set for model evaluation, and organize and extract the data set through feature engineering; Step S3: Building a hybrid deep neural network model based on one-dimensional convolutional neural network, two-dimensional convolutional neural network and Transformer; Step S4: Use k-fold cross validation to train the hybrid deep neural network model. When the loss function of the training model no longer continues to decrease, save the model and record it as the optimal model of the fold; Step S5: After the k-fold cross-validation training is completed, the model is loaded and the output of each fold of the model on the test set data is calculated, and the average is taken to predict the remaining service life of the lithium-ion battery.

2. The method for predicting the remaining service life of a lithium-ion battery based on a multidimensional data hybrid deep neural network according to claim 1, characterized in that: The performance indicators of each battery cycle in step S1 include discharge capacity, charging time, and internal resistance; the battery cycle observation data includes voltage, current, temperature, and capacity during the battery charge and discharge cycle.

3. The method for predicting the remaining service life of a lithium-ion battery based on a multidimensional data hybrid deep neural network according to claim 1, characterized in that: The charge-discharge aging experiment in step S1 is a multi-stage charge-constant-current discharge process at a constant temperature, including: first charging the battery to a specific state of charge using a constant current, then charging the battery in the charged state to a cut-off voltage using another constant current, then charging at a constant voltage to the cut-off current of the battery, and discharging to the cut-off voltage with the same current.

4. The method for predicting the remaining service life of a lithium-ion battery based on a multidimensional data hybrid deep neural network according to claim 1, characterized in that: The preprocessing in step S1 includes: using a linear interpolation method to remove outliers and normalizing the data without outliers.

5. The method for predicting the remaining service life of a lithium-ion battery based on a multidimensional data hybrid deep neural network according to claim 1, characterized in that: The feature engineering in step S2 includes: The four curves of current, voltage, temperature and capacity of each cycle are linearly interpolated into four curves with 100 data points according to the time series. The four curves form a set of 4×100 two-dimensional data matrices; the three values ​​of discharge capacity, charging time and battery internal resistance of each cycle are combined into a 1×3 one-dimensional data matrix.

6. The method for predicting the remaining service life of a lithium-ion battery based on a multidimensional data hybrid deep neural network according to claim 5, characterized in that: The step S3 comprises: The one-dimensional data matrix is ​​input into a three-layer one-dimensional convolutional neural network to extract the battery aging information from the one-dimensional scalar features, and then the output of the one-dimensional convolutional neural network is processed through a fully connected layer; The two-dimensional data matrix is ​​input into a three-layer two-dimensional convolutional neural network connected front and back, and the output of the two-dimensional convolutional neural network is further extracted through the Transformer network; The output of the Transformer network and the fully connected layer are integrated and then output through two fully connected layers. The output obtained is the estimated value of the remaining service life of the lithium-ion battery. The processing flow of the Transformer network is: The data first passes through the position encoding module to preserve the time step information of the data; The time series data is processed through a multi-head self-attention mechanism to calculate the correlation of each time step in the sequence. The self-attention mechanism performs weighted aggregation of the input features of each time step through a weight matrix.

7. The method for predicting the remaining service life of a lithium-ion battery based on a multidimensional data hybrid deep neural network according to claim 1, characterized in that: The k-fold cross validation in step S4 refers to performing cross validation using 10 folds.

8. A lithium-ion battery remaining service life prediction system based on multi-dimensional data hybrid deep neural network, characterized in that: The system comprises: The charge-discharge aging experiment unit is used to perform charge-discharge aging experiments on the battery to obtain the performance indicators of each battery cycle and battery cycle observation data, establish the original aging data set of lithium batteries, and pre-process the aging data set; A data extraction unit is used to divide the preprocessed data set into a training set for model training and a test set for model evaluation, and to organize and extract the data set through feature engineering; Model building unit, used to build a hybrid deep neural network model based on one-dimensional convolutional neural network, two-dimensional convolutional neural network and Transformer; The training unit is used to train the hybrid deep neural network model using k-fold cross validation. When the loss function of the training model no longer continues to decrease, the saved model is recorded as the optimal model of the fold; The prediction unit is used to load the model and calculate the output of each fold model on the test set data after the k-fold cross-validation training is completed, and then take the average to predict the remaining service life of the lithium-ion battery.

9. A computer device, characterized in that: It includes a memory and a processor, wherein a computer program is stored in the memory. When the processor runs the computer program stored in the memory, the processor executes a method for predicting the remaining service life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, executes the steps of a method for predicting the remaining useful life of a lithium-ion battery based on a multi-dimensional data hybrid deep neural network as described in any one of claims 1 to 7.

Images