Accurate estimation of the State of Health (SOH) is critical for ensuring the safety and reliability of Battery Management Systems (BMS) in lithium-ion batteries. The ohmic internal resistance, a parameter in the Equivalent Circuit Model (ECM), serves as a key indicator of the aging degree during battery operation. Although data-driven methods for SOH estimation are highly flexible, their reliance on externally measured parameters often limits their ability to reflect complex internal battery dynamics, resulting in limited estimation accuracy. To address this issue, this study first identifies parameters of a first-order RC ECM using the Particle Swarm Optimization (PSO) algorithm, thereby extracting the key health feature—ohmic internal resistance ( R_0 ). Concurrently, Differential Thermal Capacity (DTC) analysis is employed to derive thermodynamic characteristics during constant current charging phases. An input dataset integrating these ECM parameters and DTC features is constructed to enhance the physical information density of the model inputs, thereby improving the data quality for the subsequent data-driven model. Furthermore, a hybrid model combining a Bidirectional Gated Recurrent Unit (BiGRU) neural network with the eXtreme Gradient Boosting (XGBoost) algorithm is developed to further advance the estimation accuracy. Validation on the Oxford battery aging dataset demonstrates the framework’s superior performance, achieving an average Root Mean Square Error (RMSE) of 0.26 R_0 feature reduces the estimation error by 36.7

Accurate estimation of the state of charge (SOC) in lithium-ion batteries is crucial for battery management systems. Addressing the issue of insufficient accuracy in conventional SOC estimation methods under dynamic operating conditions, this paper proposes a lithium-ion battery SOC estimation method incorporating stress characteristics. To compensate for SOC estimation deficiencies caused by fluctuations in voltage and current characteristics, this paper introduces battery stress variation as a novel feature to characterise SOC degradation. A stress testing experimental platform was constructed, with a series of stress characterisation experiments designed. Savitzky-Golay (SG) filtering was applied to the stress data to obtain optimised, high-quality stress features. Correlation analysis validated the strong relationship between these stress features and SOC. To address the degradation in SOC estimation accuracy caused by the plateau phase in stress variation curves during battery discharge, this paper proposes an LSSVM-AdaBoost-based SOC estimation model. By integrating the advantages of multiple LSSVM models and combining locally optimal models for different charge–discharge phases, this approach significantly enhances lithium-ion battery SOC estimation accuracy. SOC estimation was conducted under Federal Urban Driving State (FUDS) conditions using the constructed experimental platform. Experimental results demonstrate that the root mean square error (RMSE) and mean absolute error (MAE) of the proposed method’s SOC estimates are 0.225 and 0.186, respectively. To validate the effectiveness of stress features, the superiority of the LSSVM-AdaBoost model, and its generalisation capability across different operating conditions, a series of comparative experiments were designed. Results confirm that the LSSVM-AdaBoost model maintains relatively superior estimation accuracy across the entire SOC range. Consequently, the proposed method effectively enhances the accuracy and robustness of lithium-ion battery SOC estimation.

Accurately assessing the state of health (SOH) Retired lithium-ion batteries is crucial for their secondary use, and a key advancement has been the extension of SOH assessment from the automotive sector to secondary applications. To address the challenges of nonlinear capacity drop-offs at the end of the secondary life of retired batteries and the tendency of traditional data-driven models to diverge when dealing with cross-domain data, this paper has collected new battery ageing data and a dataset of retired battery ageing data to construct a full-lifecycle hybrid dataset, thereby identifying health features that are easy to obtain and highly correlated within the charging voltage range. By effectively combining physical information neural networks(PINN), Transformers, and bidirectional long short-term memory (LSTM) architectures, we have improved the accuracy of SOH estimation for end-of-life batteries using a hybrid dataset. Experiments have shown that, in secondary applications, the root mean square error (RMSE) of SOH estimates is consistently below 2%. Therefore, the model designed using this method and the construction of the dataset provide a new approach for high-precision battery health management in the secondary utilization of retired batteries.

Simultaneous recovery of waste heat and water from stripped gas streams in CO2 chemical absorption is vital for lowering energy penalties. This study reports a novel class of wettability patterned ceramic membranes fabricated via spray-coating commercial hydrophilic substrates with geometrically tailored “window-lattice” hydrophobic patterns and multi-segment wettability gradients. Topological optimization revealed that increasing the pattern count of hydrophobic domains, extending the hydrophilic-hydrophobic boundary length, and tailoring the hydrophobic pattern geometry were beneficial for enhancing heat recovery performance. The tailored membranes achieved a peak heat recovery flux of 22.1 MJ/(m2·h) and a specific heat recovery potential of 853.2 kJ/kg, outperforming the initial hydrophilic membrane by 15.6

Lithium-ion batteries, as critical energy storage components in electric vehicles and energy storage systems, require accurate state of health (SOH) estimation to ensure operational safety and extend service life. However, traditional data-driven methods often suffer from insufficient robustness when facing complex degradation behaviors, small-sample scenarios, and noisy environments. To address these challenges, this article proposes a dual-layer ensemble learning framework based on bidirectional long short-term memory-modified Newton-Raphson-based optimizer-convolutional neural network with dual-attention mechanism (BiLSTM [long short-term memory]-MNRBO-CNN-Dual Attention) for lithium-ion battery SOH prediction. First, multidimensional health factors (HFs) are extracted from four complementary characteristic curves derived from battery charge-discharge data. The Pearson correlation coefficient (PCC) and Spearman correlation coefficient (SCC) analyses are then employed to select degradation-sensitive features. Subsequently, multiple BiLSTM sub-models are constructed to capture temporal degradation patterns. An MNRBO is further introduced to adaptively determine optimal fusion weights. Finally, a CNN-Dual Attention meta-learning module achieves nonlinear feature mapping and improves prediction accuracy and model robustness. Experimental results on the Oxford, National Aeronautics and Space Administration (NASA), and laboratory aging datasets demonstrate that the proposed method achieves superior performance compared with conventional CNN, LSTM, and single-layer ensemble models. The proposed framework maintains both mean absolute error (MAE) and mean absolute percentage error (MAPE) within 1%, demonstrating high prediction accuracy and strong generalization capability.

A novel CO2 regeneration process was designed by coupling gas-assisted stripping to waste heat recovery (WHR) from the stripped gas, and this process was successfully implemented on a CO2 regeneration test rig. N2 was used as a model stripping gas to enhance CO2 regeneration and introduced through one of two modes: into the bottom of the CO2 stripper (the N2-bubbling mode) or at the side of the stripper (the N2-sweeping mode). Waste heat from the stripped gas, which was a mixture of H2O(g) and noncondensable gas, was also recovered and reintroduced into the stripper using the cold CO2-rich solvent bypassed from the main rich solvent stream. The experimental results verified that N2-assisted stripping could reduce the reboiler duty by enhancing the driving force for CO2 regeneration. Compared to the traditional thermal regeneration process, the reboiler duty could be reduced by 7.9% in the N2-assisted stripping regeneration process. Furthermore, when waste heat of the stripped gas was recovered using a transport membrane condenser in the N2-bubbling regeneration mode, the maximum reboiler duty saving could reach 26.8%. Such gas-assisted stripping is applicable in the WHR system to reduce energy consumption through dual pathways, namely upgraded driving force for CO2 regeneration and improved WHR. In future work, a more suitable stripping gas should be developed to facilitate CO2 separation from the stripped gas.

The state of health (SOH) of lithium-ion batteries is a crucial parameter for assessing battery degradation. The aim of this study is to solve the problems of single extraction of health features (HFs) and redundancy of information between features in the SOH estimation. This article develops an SOH estimation method for lithium-ion batteries based on multifeature fusion and Bayesian optimization (BO)-bidirectional gated recurrent unit (BiGRU) model. First, a total of eight HFs in three categories, namely, time, energy, and probability, can be extracted from the charging data to accurately describe the aging mechanism of the battery. The Pearson and Spearman analysis method verified the strong correlation between HFs and SOH. Second, the multiple principal components obtained by kernel principal component analysis (KPCA) can eliminate the redundancy of information between HFs. The principal component with the highest correlation with SOH is selected by bicorrelation analysis to be defined as the fused HF. Finally, to improve SOH estimation accuracy, the BO-BiGRU model is proposed. The proposed method is validated using battery datasets from NASA. The results show that the SOH estimation accuracy of the BO-BiGRU model proposed in this article is high, while mean absolute error (MAE) is lower than 1.2%. In addition, the SOH of the lithium battery is estimated using different proportions of test sets, and the results show that the root-mean-square error (RMSE) and the mean absolute percentage error (MAPE) of the SOH remain within 3%, with high estimation accuracy and robustness.

Accurate prediction of the SOH(State of Health) of lithium-ion batteries is essential for ensuring the safety and efficiency of new energy vehicles. To overcome the limitations of existing methods that rely on single-category HFs(Health features), this study proposes an SOH prediction approach based on feature-type analysis and a GAPSO-GCRN neural network. Multi-dimensional HFs are extracted from voltage, IC(Incremental Capacity), and DTV(Differential Thermal Voltammetry) curves during charging and discharging, covering voltage, time, temperature, and capacity dimensions. The Pearson-Spearman mixed correlation analysis, combined with feature evolution trends during aging, identifies three optimal indicators: voltage inflection point, capacity entropy change rate, and temperature rise rate. A GCRN(Graph Convolutional Recurrent Network) model is then developed, with a GAPSO(Genetic Algorithm-Particle Swarm Optimization) hybrid strategy employed for global hyperparameter optimization. Experimental results on the Oxford Battery Degradation Dataset show that the proposed method achieves MAE(Mean Absolute Error) and RMSE(Root Mean Square Error) within 0.4% for SOH prediction, demonstrating high accuracy, robustness, and strong generalization capability.

To improve the performance of lithium-ion battery pack balancing system, a modular balancing system based on fuzzy adaptive model predictive control (FAMPC) was proposed. Firstly, a dual-layer balancing topology structure was composed of an improved buck-boost circuit and a flyback transformer. Secondly, using the state of charge (SOC) at different levels of battery remaining capacities as inputs for the fuzzy logic algorithm, the constraints on the balancing current were adjusted. Then, based on FAMPC balancing control method, the duty cycle of the switching transistor was directly used as the system input. Finally, simulation experiments were conducted without employing additional current control mechanisms to change the battery pack state. The results show that compared with traditional fuzzy control methods, the proposed system increase the balancing speed by approximately 24.51% under normal conditions and can further increase the balancing speed to 34.48% under extreme conditions with low battery SOC. The proposed system combines the stability provided by fuzzy algorithms with the rapid response of model predictive control algorithms, ensuring safer and more stable operation of the battery pack, which can provide reference for the research of enhancing battery pack performance.

Charging strategy optimization for lithium-ion batteries is crucial to improve the efficiency of new energy devices. In this study, three conventional methods are compared: constant-current constant-voltage (CCCV) charging is the most efficient but slowest, pulse charging (PC) is the fastest but least efficient, and multistage constant-current (MSCC) is a compromise between speed and efficiency. To this end, we propose a dynamic optimal charging strategy based on model predictive control (MPC) that balances rapid-charging speed with battery safety. By integrating a low-order electrochemical-thermal-aging coupled model with real-time state estimation provided by an extended Kalman filter (EKF), a rolling-horizon framework is established to track both state-of-charge (SOC) and temperature reference trajectories. Experiments show that EKF has stronger initial error robustness (maximum deviation < 2%) than unscented Kalman filter (UKF) and unscented Kalman Bucy filter (UKBF), which provides reliable feedback for MPC. The new strategy achieves an optimal balance between charging efficiency and safety by dynamically adjusting the charging profile and significantly improves the charging speed under closed-loop control compared to the CCCV method, while controlling the temperature rise within 5 degrees C.

In the background of the rapid development of electric vehicles, accurate estimation of the State of charge (SOC) is essential to achieve battery management. However, there is a plateau problem with the voltage of LiFePO4 batteries. This creates challenges for state estimation. To address these problems. A SOC estimation method for LiFePO4 batteries based on a transportable force-electric dual-observation model is presented in this paper. In this paper, the expansion force signal and voltage signal are combined. The force-electric dual-observation model is obtained by training using a least squares support vector machine (LSSVM). The force-electric dual observation model is used as an observation equation for the adaptive untraceable Kalman filter algorithm (AUKF). The optimal weights for the force-electric dual-observation model are obtained by pre-experimentation. In addition, the force-electric dual-observation model for 12AH and 5AH LiFePO4 batteries was obtained using transfer learning (TL). Finally, the proposed method is fully evaluated by experiments under different constraints. The experimental results show that the average absolute errors of SOC estimation for the three batteries with different capacities are all within 1.2 %. The root mean square errors are all within 1.8 %. The robustness and accuracy of the proposed method are confirmed.

The battery balancing technology based on modular converters needs to solve the problem of how to make many modular converters in series and parallel work together stably. In this article, according to the characteristics of modular battery energy storage systems, the application form of droop control is improved, and a battery unit with converter (BUC) is designed by combining battery, modular converter, and droop control. The properties of parallel BUCs and series BUCs are analyzed and redesigned to make them suitable for series-parallel expansion and the output current sharing control of the batteries. On this basis, a construction method of energy storage systems based on BUCs is proposed. In addition, a two-layer composite control strategy based on improved droop control is proposed, including the lower control strategy and the upper control strategy. Both simulation and physical experiments show that the proposed scheme can realize the constant voltage or constant current control of the total output of the energy storage system, and it can make the output current of each battery shared according to a given ratio with high control accuracy, and it can tolerate the communication failure of up to 2.5 s.

Accurately predicting the future capacity and remaining useful life (RUL) of lithium-ion batteries is crucial for ensuring their safety and reducing the maintenance costs of related equipment. However, the aging data of lithium-ion batteries (LIBs) exhibit significant nonlinearity and are also affected by uncertainties such as capacity regeneration. To address this issue, this paper proposes an RUL prediction method based on a Convolutional Neural Networks-Attention Mechanism (CNN-Attention) combined with a Slime Mold Algorithm- Gaussian Process Regression (SMA-GPR) model. Firstly, SVMD is applied to extract capacity regeneration features and capacity decay features. Next, to solve the data dependency of single-model prediction, the SMA-GPR model is applied to improve the CNN-Attention prediction, thus solving the generalization problem of and obtaining an accurate RUL. Next, to verify the accuracy and robustness of the proposed method, the experiments involved long-term aging tests under multiple scenarios including three types of LiFeO4(LFP)batteries including 142Ah square cells,280Ah square cells and 10Ah pouch cells, and varying conditions including temperature, charge- discharge rates and pre-tightening forces. The prediction error based on experimental data applying the three batteries is within 1 %.

The state of health (SOH) of a battery is the main indicator of battery life. In order to improve the SOH estimation accuracy, a model framework for lithium-ion battery health state estimation with feature reconstruction and improved least squares support vector machine is proposed. First, the indirect health features (HF) are obtained by processing multiple health features extracted from the charging and discharging phases through principal component analysis to remove the information redundancy among multiple features. Subsequently, multiple smooth component subsequences of different frequencies are obtained by using variational modal decomposition to efficiently capture the overall downtrend and regeneration fluctuations of the data. Then, use the sparrow search algorithm to optimize the least squares support vector machine to build an estimation model, predict and superimpose the reconstructed fusion features of multiple feature subsequences. Finally, use the mapping relationship between the reconstructed HF and the SOH for the estimation. The NASA battery dataset and the University of Maryland battery dataset (CACLE) are used to perform validation tests on multiple batteries with different cycle intervals. The results show that the mean absolute error and root mean square error are less than 1% and the method has high-estimation accuracy and robustness.

In energy storage systems, multiple energy storage monomers are usually connected in series to obtain higher voltages, but the inconsistency of the voltage of each energy storage monomer will reduce the utilization of the storage unit. To address this problem, this article proposes a method for equalizing the voltage of series energy storage units based on LC resonant circuit. The equalization circuit consists of a switch array and an LC resonant converter, which can achieve energy transfer between any monomer and continuous multi‐monomer, and realize zero‐current conduction of the switch. The equalization circuit does not have a large number of magnetic components, and for each additional energy storage monomer, the circuit only needs to add a pair of switches, which has the advantages of high flexibility and expandability. Finally, the equalization simulation experiments are conducted on the Matlab/Simulink platform for the energy storage unit composed of four series‐connected energy storage monomers. The experimental results show that under static equalization, the voltage reaches equalization in about 25 min and the equalization efficiency reaches 96.64%, which greatly reduces the switching loss, improves the equalization speed and efficiency, and verifies the feasibility and effectiveness of the method.

To address the impact of wind-power fluctuations on the stability of power systems, we propose a comprehensive approach that integrates multiple strategies and methods to enhance the efficiency and reliability of a system. First, we employ a strategy that restricts long- and short-term power output deviations to smoothen wind power fluctuations in real time. Second, we adopt the sliding window instantaneous complete ensemble empirical mode decomposition with adaptive noise (SW-ICEEMDAN) strategy to achieve real-time decomposition of the energy storage power, facilitating internal power distribution within the hybrid energy storage system. Finally, we introduce a rule-based multi-fuzzy control strategy for the secondary adjustment of the initial power allocation commands for different energy storage components. Through simulation validation, we demonstrate that the proposed comprehensive control strategy can smoothen wind power fluctuations in real time and decompose energy storage power. Compared with traditional empirical mode decomposition (EMD), ensemble empirical mode decomposition (EEMD), and complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) decomposition strategies, the configuration of the energy storage system under the SW-ICEEMDAN control strategy is more optimal. Additionally, the state-of-charge of energy storage components fluctuates within a reasonable range, enhancing the stability of the power system and ensuring the secure operation of the energy storage system.

Precisely estimating the state of health (SOH) of lithium-ion batteries is essential for battery management systems (BMS), as it plays a key role in ensuring the safe and reliable operation of battery systems. However, current SOH estimation methods often overlook the valuable temperature information that can effectively characterize battery aging during capacity degradation. Additionally, the Elman neural network, which is commonly employed for SOH estimation, exhibits several drawbacks, including slow training speed, a tendency to become trapped in local minima, and the initialization of weights and thresholds using pseudo-random numbers, leading to unstable model performance. To address these issues, this study addresses the challenge of precise and effective SOH detection by proposing a method for estimating the SOH of lithium-ion batteries based on differential thermal voltammetry (DTV) and an SSA-Elman neural network. Firstly, two health features (HFs) considering temperature factors and battery voltage are extracted from the differential thermal voltammetry curves and incremental capacity curves. Next, the Sparrow Search Algorithm (SSA) is employed to optimize the initial weights and thresholds of the Elman neural network, forming the SSA-Elman neural network model. To validate the performance, various neural networks, including the proposed SSA-Elman network, are tested using the Oxford battery aging dataset. The experimental results demonstrate that the method developed in this study achieves superior accuracy and robustness, with a mean absolute error (MAE) of less than 0.9% and a root mean square error (RMSE) below 1.4%.

With the development of electric vehicles, the demand for lithium-ion batteries has been increasing annually. Accurately estimating the state of health (SOH) of lithium-ion batteries is crucial for their efficient and reliable use. Most of the existing research on SOH estimation is based on parameters such as current, voltage, and temperature, which are prone to fluctuations. Estimating the SOH of lithium-ion batteries based on electrochemical impedance spectroscopy (EIS) and data-driven approaches has been proven effective. In this paper, we explore a novel SOH estimation model for lithium batteries based on EIS and Convolutional Neural Network (CNN)-Vision Transformer (VIT). The EIS data are treated as a grayscale image, eliminating the need for manual feature extraction and simultaneously capturing both local and global features in the data. To validate the effectiveness of the proposed model, a series of simulation experiments are conducted, comparing it with various traditional machine learning models in terms of root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and coefficient of determination (R2). The simulation results demonstrate that the proposed model performs best overall in the testing dataset at three different temperatures. This confirms that the model can accurately and stably estimate the SOH of lithium-ion batteries without requiring manual feature extraction and knowledge of battery aging temperature.