Life Prediction Of Lithium-ion Batteries Research Articles

Remaining useful life prediction of lithium-ion battery based on AConvST-LSTM-Net-TL

Abstract The accurate prediction of the Remaining Useful Life (RUL) of lithium-ion batteries can significantly enhance the safety and reliability of these batteries, thereby reducing operational risks. However, numerous existing methodologies operate under the assumption that both training and testing data adhere to the same distribution pattern, hindering the application of successful laboratory-based models to different target batteries. Hence, this study introduces a transfer learning model for predicting lithium-ion battery life, utilizing an attention mechanism-driven convolutional neural network (ACNN) in conjunction with a spatiotemporal Long Short-Term Memory network (ST-LSTM). Initially, the deep charging process data is leveraged for battery capacity estimation, where a Generalized Additive Model (GAM) is employed to delineate the nonlinear trend of battery capacity decay and characterize the capacity degradation. Following feature extraction, the Maximum Mean Discrepancy (MMD) value is computed to assess the distribution disparity between the two domains. Subsequently, the AConvST-LSTM-Net computes capacity estimations, loss functions for both source and target domains, merging these with the MMD value to formulate a comprehensive loss function. In conclusion, by employing transfer learning, the refined model is adapted to the target domain dataset, enabling the accurate prediction of the RUL of lithium-ion batteries. This approach is validated using the NASA lithium-ion battery dataset and the National Big Data Alliance for New Energy Vehicles (NDANEV), demonstrating the superior accuracy and robustness of the AConvST-LSTM-Net-TL model in comparisn to other prediction methods.

Remaining useful life prediction of lithium-ion batteries using a novel particle flow filter framework with grey model

Remaining useful life (RUL) prediction is a crucial aspect of the prognostics health management of lithium-ion batteries (LIBs). Owing to the influence of resampling technology, particle degradation is often observed in the particle filter-based RUL prediction of LIBs, resulting in a low prediction accuracy and large uncertainty. In this paper, a novel particle flow filter with the grey model method (GM-PFF) is proposed to forecast the RUL and state of health of batteries. First, the least squares method is employed to obtain the initial values for double exponential empirical model parameters. Subsequently, the grey model is used to predict the current cycle capacity of LIBs as an observation value for the particle flow filter, solving the inaccurate estimation problem of the state of particle flow filter observation values, and the particle flow filter method is employed to update model parameters. Finally, a test dataset is divided into early, middle, and late stages to predict the RUL of LIBs and obtain the probability distributions. On the CALCE and NASA PCoE LIB dataset, GM-PFF reduces RMSE by 1% compared to PFF, exhibiting a higher prediction accuracy and effectively addressing the particle degradation problem.

An enhanced deep learning framework for state of health and remaining useful life prediction of lithium-ion battery based on discharge fragments

An enhanced deep learning framework for state of health and remaining useful life prediction of lithium-ion battery based on discharge fragments

An Exponential Transformer for Learning Interpretable Temporal Information in Remaining Useful Life Prediction of lithium-ion Battery

An Exponential Transformer for Learning Interpretable Temporal Information in Remaining Useful Life Prediction of lithium-ion Battery

Deep learning-based Remaining Useful Life Prediction of Lithium-ion Battery Considering Two-phase Aging Process

Abstract The aging process of lithium-ion batteries is typically nonlinear, characterized by a knee point that divides it into two distinct phases: a slow aging phase and a rapid aging phase. Predicting the remaining useful life (RUL) in rapid phase is crucial for providing users with timely alerts to replace the battery before its end of life (EOL). A two-stage identification and prediction framework has been proposed to identify the current aging phase of a battery and predict its RUL if it is in the rapid aging phase, requiring data from only the most recent three cycles of the charge process. In the first stage, a 2D-ResNet is constructed for aging phase identification, offering strategies with high accuracy or low risk depending on user requirements. In the second stage, a Gated-ResNet-Attention model is innovatively constructed to filter out unreliable temperature features. The model achieved mean absolute error and root mean square error values of 6.94 and 8.12, respectively. The proposed framework can identify aging process and predict the RUL in rapid aging phase accurately with minimal input requirements. This approach reduces data storage requirements while providing users with sufficient indicative information for proactive battery maintenance.

Application of state of health estimation and remaining useful life prediction for lithium-ion batteries based on AT-CNN-BiLSTM

Ensuring the long-term safe usage of lithium-ion batteries hinges on accurately estimating the State of Health \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$(\ extrm{SOH})$$\\end{document} and predicting the Remaining Useful Life (RUL). This study proposes a novel prediction method based on a AT-CNN-BiLSTM architecture. Initially, key parameters such as voltage, current, temperature, and SOH are extracted and averaged for each cycle to ensure the uniformity and reliability of the input data. The CNN is utilized to extract deep features from the data, followed by BiLSTM to analyze the temporal dependencies in the data sequences. Since multidimensional parameter data are used to predict the SOH trend of lithium-ion batteries, an attention mechanism is employed to enhance the weight of highly relevant vectors, improving the model’s analytical capabilities. Experimental results demonstrate that the CNN-BiLSTM-Attention model achieves an absolute error of 0 in RUL prediction, an \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$R^{2}$$\\end{document} value greater than 0.9910 , and a MAPE value less than 0.9003 . Comparative analysis with hybrid neural network algorithms such as LSTM, BiLSTM, and CNN-LSTM confirms the proposed model’s high accuracy and stability in SOH estimation and RUL prediction.

Data-Driven Feature Design Framework for Extreme Early Cycle Life Prediction for Lithium-Ion Batteries

Numerous publications describe the construction of data-driven models for cycle life prediction for lithium-ion batteries. The models with the highest predictive capability have been constructed using user-defined features, which are transformations of the raw input data, before applying feature selection and regression. The performance and interpretability of the results depend on the design of the features, especially when data are limited. For example, Severson et al. [1] designed the Var(ΔQ 100-10) feature, which was very successful for cycle life prediction for lithium-ion phosphate batteries. A linear regression model constructed from only that one nonlinear feature, computed from the raw voltage-capacity data collected during discharge, was able to predict cycle life from the first 100 cycles. The same feature was able to classify cells into short and long-lived using only the first 5 cycles. A variation of that feature, the ΔV feature, was later shown to apply to other battery chemistries (e.g., [2]), although not with the same degree of accuracy.Following this work, there was much research focused on using various transformations of electrochemical and thermal input data for performing data-driven lithium-ion battery lifetime prognostics. These works show the interplay between mechanistic understanding and data-driven approach: mechanistic understanding guides the design of promising features, which results in more accurate and interpretable data-driven models.However, it is challenging to design effective features for applications where less physics is known, such as performing lifetime prognostics right after the formation step of the lithium-ion battery (i.e., extreme early cycle life estimation) [3]. Performing extreme early cycle life estimation is important to (1) pre-screen battery cells with poor quality SEI layers before the product is shipped to the customer, (2) speed up optimization of the formation protocol without the need for extensive tests of aging cycles, and (3) reveal uninvestigated physics of formation. To achieve these goals, we have developed a systematic data-driven framework that guides the design of such promising features without requiring in-depth physics-based models and understanding.In this presentation, a data-driven feature design framework is described which is composed of (1) downselection of input data segments from fused LASSO regression coefficients [4], (2) design of feature candidates based on learned coefficients and similar applications, and (3) evaluation of each feature candidate based on the predictive performance and the shape of the coefficients [4]. This framework is applied to the NMC532-graphite dataset with 62 different formation protocols and identical aging protocols which captures the sole impact of formation protocol on the cycle life. The data-driven model using the features designed from our framework is compared with the agnostic model using formation protocol parameters and an automated machine learning (aka autoML) model using the features generated from the ‘tsfresh’ package.

A residual service life prediction of lithium-ion batteries based on decomposition algorithm and fully connected neural network

A residual service life prediction of lithium-ion batteries based on decomposition algorithm and fully connected neural network

Remaining useful life prediction of lithium-ion batteries based on DBO[sbnd]CNN-DSformer

In order to improve the accuracy of predicting RUL of lithium-ion batteries, a lithium-ion battery RUL prediction method based on the DBOCNN-DSformer model is proposed. Firstly, the health characteristics of the battery are extracted and the local information of health features is mined using CNN. DSformer is utilized for global information, local information, and variable correlation learning of battery aging characteristics. The DBO is used to optimize the super-parameters of the CNN-DSformer model and build the DBOCNN-DSformer model. Finally, the battery aging data set was used for verification. The results show that DBOCNN-DSformer, which sets different prediction starting points, can extract sequence information from input data and establish long-term dependencies between sequences. The maximum average MRE error in the NASA data set was 0.05, the maximum average MAE was 0.018, and the maximum average AE error was within 5. The maximum average MRE error of the CALCE data set was 0.37, the maximum average MAE was 0.014, and the maximum average AE prediction error was within 10. Compared with LSTM, RNN, and Transformer models, it was found that DBOCNN-DSformer showed high prediction accuracy and good robustness.

Remaining useful life prediction of lithium-ion batteries based on autoregression with exogenous variables model

The gradual decrease capacity serves as a pivotal health indicator, reflecting the condition of lithium-ion batteries. Accurate forecasting of capacity can ascertain the remaining lifespan of these batteries at any given cycle, which is crucial for managing batteries in electric vehicles. This paper proposes an Autoregression with Exogenous Variables (AREV) model, which continually updates itself through a sliding window, offering predictions of battery state of health and remaining useful life, which extends battery prognostics at a fixed operating condition to different operating conditions. In addition, unlike most models that require multiple battery data of the same type for training, the proposed model only requires the use of fragmented data of the target battery with length around 30-50 cycles for capacity prediction and determines battery life based on battery failure thresholds. The above two points enable this model to be updated online without the need for any offline training. Finally, four different types of battery dataset , with different chemical substances and different charge and discharge conditions (especially dataset that follows random walk discharging profile to stimulate the real power consumption process) , are applied to verify the effectiveness and robustness of proposed RUL prediction approach. It shows that the proposed model can accurately predicting future capacity values. Timely warning signals can be issued before the end of life of battery, thereby ensuring the safe driving of electric vehicles and timely battery replacement.

Remaining Useful Life Prediction of Lithium-Ion Batteries Based on Simplified Electrochemical Model and TSO-TCN

Accurate prediction of the remaining useful life (RUL) of lithium-ion battery is critical in practical applications, but is challenging due to the presence of multiple aging pathways and nonlinear degradation mechanisms. In this paper, a method for RUL prediction is proposed combined with battery capacity aging mechanism based on transient search optimization (TSO)-temporal convolutional network (TCN) algorithm. First, the particle swarm optimization algorithm is used to derive three health indicators directly related to capacity loss from a simplified electrochemical model. Then, the TCN parameters are optimized with transient search algorithm to obtain the optimal prediction model. Finally, the RUL prediction are compared with other typical algorithms, and the results show that the proposed method can accurately predict the RUL of lithium-ion battery, and the life prediction error is within 10 cycles. Compared to TCN, the prediction results remain accurate even with less training data, and the error metrics are reduced by about 50% with the maximum error only 7 cycles from the 250th charge/discharge cycle.

Remaining Useful Life Prediction of Lithium-Ion Batteries With Limited Degradation History Using Random Forest

Remaining Useful Life Prediction of Lithium-Ion Batteries With Limited Degradation History Using Random Forest

TPANet: A novel triple parallel attention network approach for remaining useful life prediction of lithium-ion batteries

Accurate remaining useful life (RUL) prediction for lithium-ion batteries (LIBs) is important for proper equipment operation. Among the numerous existing studies on battery research, data-driven approaches have gained significant attention because they obviate the need for complex chemical and physical modeling of battery processes. However, in most current data-driven methods, the sliding window size values are determined empirically, which affects both the time required for model prediction and the prediction accuracy. To address this issue, this paper proposes a two-stage prediction method for battery capacity aging trajectories. In the first stage, a false nearest neighbor (FNN) technique is utilized to infer the sliding window size of the battery, reducing the error associated with manually selecting the sliding window size. In the second stage, a novel triple parallel attention network (TPANet) based on convolutional neural networks (CNNs) and attention units is developed, which extracts the input battery capacity degradation features through a parallel mechanism. The introduction of attention units in each parallel branch allows the model to enhance the capture of capacity regeneration phenomena as well as long-term dependence on input data. The proposed approach is validated on two publicly available datasets as well as a battery developed by ourselves. Diverse simulation results demonstrate the ability of the proposed approach to accurately predict the RUL of LIBs in a short time with broad generalization capabilities.

Capacity fading knee-point recognition method and life prediction for lithium-ion batteries using segmented capacity degradation model

Aiming at the problem that the cycle life of the battery is hard to be accurately estimated, a segmented capacity degradation model is established based on the trend of capacity degradation rate. By applying a genetic algorithm (GA) to optimize the model parameters, GA-Weibull and GA-SVR (support vector regression) degradation models and their corresponding segmented models are established, and the battery life is calculated. Furthermore, a method taking the RMSE as the target is put forward to calculate the knee-point of capacity degradation. Finally, the life is predicted by segmented models. The results show that compared with unsegmented models, describing the process of capacity degradation by segmented models reduces the error notably, makes the determination coefficient closer to 1, and improves the model accuracy effectively. Later, the proposed method makes the fitting error of the segmented model smaller than that of Bacon-Watts method, thus obtaining a more accurate knee-point. In addition, it is found that there exists linear relationship between the estimated knee-point and the battery life. Finally, the prediction results based on segmented model show that the prediction accuracy of segmented GA-SVR model is higher, and the MAPE is only 4.32%. The relevant results can provide some guidance for the reliability evaluation and product quality management of lithium-ion batteries, which is conducive to establishing more accurate models for battery life prediction.

A novel LFSSA-LSTM algorithm for the remaining useful life prediction of lithium-ion batteries in whole life cycle

The accurate prediction of the remaining useful life (RUL) of lithium-ion batteries is essential to improve the safety of battery systems and optimize battery management strategies. Therefore, a new algorithm combining Sparrow search algorithm (SSA) improved by the Levy flight method and long short-term memory neural network (LSTM) is proposed in this paper. By extracting the National Aeronautics and Space Administration (NASA) battery aging datasets, the battery correlated health characteristics are obtained, and the deterioration trend of battery performance is effectively described. The sparrow search algorithm improved by the Levy flight method enhances the ability of the algorithm to jump out of the optimal solution. By using the improved algorithm, the hyper-parameter optimization problem of the long short-term memory network is solved, and the accuracy of the algorithm is improved. Finally, the probability density function is used to predict the remaining useful life of lithium-ion batteries, which effectively prevents the uncertainty caused by a single prediction result. Compared with long short-term memory neural network and the one optimized by a traditional sparrow search algorithm, the mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) of the proposed algorithm are kept at 1.77 %, 2.24 %, and 1.25 %, respectively. The results indicate that the algorithm has better accuracy and fitting performance.

Remaining useful life prediction of lithium-ion batteries based on performance degradation mechanism analysis and improved Deep Extreme Learning Machine model

Remaining useful life prediction of lithium-ion batteries based on performance degradation mechanism analysis and improved Deep Extreme Learning Machine model

Remaining Useful Life Prediction of Lithium-Ion Batteries Based on Neural Network and Adaptive Unscented Kalman Filter

Remaining Useful Life Prediction of Lithium-Ion Batteries Based on Neural Network and Adaptive Unscented Kalman Filter

A novel multiple kernel extreme learning machine model for remaining useful life prediction of lithium-ion batteries

The safety and decommissioning of lithium-ion batteries (LIBs) are closely related to the remaining useful life (RUL). Accurate life prediction of LIBs can prevent safety issues and recycling at the end of life. This paper proposes a method based on the multiple kernel extreme learning machine (MKELM) to improve the stability and accuracy of LIBs life prediction. According to the measured cyclic aging data of the 18650 battery, in-depth analysis of charging curves, extract the equal voltage rise charging time, the constant voltage charging time and the equal time voltage drop in the static state as the health factors, using MKELM to predict the remaining cycle life of LIBs, and using the whale algorithm to optimize the nuclear parameters of MKELM. The proposed method is simulated and verified by the actual LIBs data. When the training set is the first 50% of the data, the whale optimization algorithm-MKELM (WOA-MKELM) algorithm predicts the average error cycle number of the three batteries is 15.6 cycles. After adding 10% of the training set, the average number of forecast error cycles reduced to 5 cycles, and the results show that the method has high stability and accuracy.

An Improved Transformer Model for Remaining Useful Life Prediction of Lithium-Ion Batteries under Random Charging and Discharging

With the development of artificial intelligence and deep learning, deep neural networks have become an important method for predicting the remaining useful life (RUL) of lithium-ion batteries. In this paper, drawing inspiration from the transformer sequence-to-sequence task’s transformation capability, we propose a fusion model that integrates the functions of the stacked denoising autoencoder (SDAE) and the Transformer model in order to improve the performance of RUL prediction. Firstly, the health factors under three different conditions are extracted from the measurement data as model inputs. These conditions include constant current and voltage, random discharge, and the application of principal component analysis (PCA) for dimensionality reduction. Then, SDAE is responsible for denoising and feature extraction, and the Transformer model is utilized for sequence modeling and RUL prediction of the processed data. Finally, accurate prediction of the RUL of the four battery cells is achieved through cross-validation and four sets of comparison experiments. Three evaluation metrics, MAE, RMSE, and MAPE, are selected, and the values of these metrics are 0.170, 0.202, and 19.611%, respectively. The results demonstrate that the proposed method outperforms other prediction models in terms of prediction accuracy, robustness, and generalizability. This provides a new solution direction for the daily life prediction research of lithium-ion batteries.

State of Health Estimation and Remaining Useful Life Prediction of Lithium-Ion Batteries by Charging Feature Extraction and Ridge Regression

The state of health (SOH) and remaining useful life (RUL) of lithium-ion batteries are critical indicators for assessing battery reliability and safety management. However, these two indicators are difficult to measure directly, posing a challenge to ensure safe and stable battery operation. This paper proposes a method for estimating SOH and predicting RUL of lithium-ion batteries by charging feature extraction and ridge regression. First, three sets of health feature parameters are extracted from the charging voltage curve. The relationship between these health features and maximum battery capacity is quantitatively evaluated using the correlation analysis method. Then, the ridge regression method is employed to establish the battery aging model and estimate SOH. Meanwhile, a multiscale prediction model is developed to predict changes in health features as the number of charge-discharge cycles increases, combining with the battery aging model to perform multistep SOH estimation for predicting RUL. Finally, the accuracy and adaptability of the proposed method are confirmed by two battery datasets obtained from varying operating conditions. Experimental results demonstrate that the prediction curves can approximate the real values closely, the mean absolute error (MAE) and root mean square error (RMSE) calculations of SOH remain below 0.02, and the maximum absolute error (AE) of RUL is no more than two cycles.