Article
Energy & Fuels
Dao Zhao, Zhijie Zhou, Peng Zhang, Yijun Zhang, Zhichao Feng, Yang Yang, You Cao
Summary: To accurately assess the health condition of satellite li-ion battery pack and ensure the safety and service life of satellite power supply, a novel indicator system integrating battery inconsistency and pack performance indicators is built based on the performance requirement of battery pack. Equilibrium and bypass state are employed as factors to make the assessment result more accurate and comprehensive, using a method based on ER and BRB for inconsistencies fusion and rule-based assessment. The results can be expressed as belief degree distribution or score, and the proposed method is proven effective by assessing the health condition of a certain type of in-orbit satellite's li-ion battery pack using telemetry data.
JOURNAL OF ENERGY STORAGE
(2023)
Article
Multidisciplinary Sciences
Huiqiao Liu, Qian Xiao, Yu Jin, Yunfei Mu, Jinhao Meng, Tianyu Zhang, Hongjie Jia, Remus Teodorescu
Summary: This study introduces an improved LightGBM-based framework for battery RUL prediction, which achieves enhanced accuracy and speed by adaptively adjusting to multiple HIs and improving the loss function, validated with database testing for optimal performance.
Article
Engineering, Aerospace
Raul Llasag Rosero, Catarina Silva, Bernardete Ribeiro
Summary: Predictive Maintenance (PM) strategies are of interest in the aviation industry to reduce maintenance costs and Aircraft On Ground (AOG) time. This paper proposes the integration of a physics-based model with a data-driven model to predict the Remaining Useful Life (RUL) of aircraft cooling units. The results show that the cooling units experience a normal degradation stage before an abnormal degradation that occurs within the last flight hours of useful life.
Article
Engineering, Multidisciplinary
Chun Su, Hongjing Chen, Zejun Wen
Summary: This study proposes a novel method for Li-ion battery's online RUL prediction based on multiple health indicators, which effectively estimates the remaining capacity using GRNN and predicts RUL using NAR, while reducing interference through wavelet denoising. The method was demonstrated to be effective in obtaining Li-ion batteries' RUL through a NASA dataset case study.
EKSPLOATACJA I NIEZAWODNOSC-MAINTENANCE AND RELIABILITY
(2021)
Article
Automation & Control Systems
Jianghong Zhou, Yi Qin, Jun Luo, Tao Zhu
Summary: Facing the gap in the unsupervised construction of health indicator (HI) with a uniform failure threshold, a new approach is developed by estimating the distribution of the raw vibration signal using the Gaussian mixture model and designing a distribution contact ratio metric (DCRM). A distribution contact ratio metric health indicator (DCRHI) is constructed to represent the degradation process and obtain a uniform failure threshold. Furthermore, a novel consolidated memory gated recurrent unit (CMGRU) is proposed to slow down the forgetting speed of important trend information and improve the prediction ability. The proposed methodology shows great application value in the RUL prediction.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Energy & Fuels
Lingfeng Fan, Ping Wang, Ze Cheng
Summary: This paper introduces a remaining capacity prediction technique for lithium-ion batteries based on partial charging curve and health feature fusion, with a battery aging model established by Gaussian process regression. The reliability and accuracy of the proposed method are validated on six battery data sets from NASA and the University of Oxford.
JOURNAL OF ENERGY STORAGE
(2021)
Article
Nanoscience & Nanotechnology
Chunli Liu, Dezhi Li, Licheng Wang, Liwei Li, Kai Wang
Summary: In this study, a temporal convolutional network is used to predict the remaining useful life of supercapacitors. The model's stability and accuracy were verified by using residual blocks, early stopping technology, and the Adam algorithm for optimization. The simulation demonstrated the robustness and accuracy of the model in predicting the remaining useful life of supercapacitors.
Article
Energy & Fuels
Jinsong Yang, Weiguang Fang, Jiayu Chen, Boqing Yao
Summary: This paper proposes an integrated lithium-ion battery RUL prediction method based on a particle resampling strategy and unscented particle filter. The method improves prediction accuracy and reliability by generating proposal distribution of particle weights using unscented Kalman filter and employing optimal combination strategy for resampling.
JOURNAL OF ENERGY STORAGE
(2022)
Article
Energy & Fuels
Shaotang Cai, Jun Hu, Shuoqi Ma, Zhenning Yang, Hao Wu
Summary: This paper proposes an adaptive prediction method for the remaining useful life (RUL) of electric vehicle (EV) power battery based on the whale swarm algorithm-long short-term memory (WSA-LSTM) algorithm. By screening health indicators and globally optimizing parameters, the method achieves fast and accurate prediction of battery life under DC fast charging conditions.
Article
Green & Sustainable Science & Technology
Jiwei Wang, Zhongwei Deng, Kaile Peng, Xinchen Deng, Lijun Xu, Guoqing Guan, Abuliti Abudula
Summary: This paper proposes a method for predicting the future health of a lithium-ion battery pack based on the aging data of the battery cells and early cycling data of the pack. It constructs degradation models of health indicators and uses a data-driven model to predict the health of the pack. Experimental results validate the satisfactory accuracy of the proposed method.
Article
Computer Science, Information Systems
Maximilian Benker, Artem Bliznyuk, Michael F. Zaeh
Summary: The quality of RUL estimation is crucial for predictive maintenance strategies, with deep learning methods often relying on large amounts of failure data. This paper introduces a data-efficient approach that can estimate RUL without the need for complete failure sequence data.
Review
Engineering, Multidisciplinary
Ming-Feng Ge, Yiben Liu, Xingxing Jiang, Jie Liu
Summary: This review discusses the importance of estimating the SOH and predicting the RUL of lithium-ion batteries, summarizes the current research status, methods classification, advantages and limitations, as well as future development trends and challenges.
Article
Energy & Fuels
Jiabo Li, Min Ye, Yan Wang, Qiao Wang, Meng Wei
Summary: This research paper proposes a Gaussian process regression (GPR) based framework for predicting the remaining useful life (RUL) of battery energy storage systems (ESS), using automatic stack autoencoder (SAE) and improved whale optimization algorithm (WOA). The SAE and gray relation analysis are used to extract indirect health indicators (HIs) from battery degradation data. The GPR approach combined with the WOA method is utilized to establish the RUL prediction framework. The simulation results show that the proposed algorithm can predict RUL with an error control of less than 2% and has good advantages and prediction ability.
JOURNAL OF ENERGY STORAGE
(2023)
Article
Energy & Fuels
Xinghong Zhang, Yi Xu, Zehao Gong
Summary: As the use of lithium-ion batteries in energy storage systems increases, there is a growing concern about accidents caused by battery performance degradation. To ensure the steady operation of these systems, it is important to accurately forecast the remaining useful life (RUL) of lithium-ion batteries. This study proposes a prediction framework that combines incremental capacity analysis (ICA) and electrochemical impedance spectroscopy (EIS) to address RUL issues. The framework uses convolutional neural network (CNN) and improved long short-term memory network (ILSTM) to establish a mapping association between fusion features and RUL. The parameters of the improved particle swarm optimization algorithm (IPSO) are optimized to enhance its optimization ability. The applicability and efficacy of this method are confirmed through numerical experiments using NASA PCoE datasets.
Article
Energy & Fuels
Xinghong Zhang, Yi Xu, Zehao Gong
Summary: With the increasing percentage of lithium-ion batteries in energy storage systems, accidents caused by the deterioration of battery performance continue to occur. Learning how to accurately forecast the remaining useful life (RUL) of lithium-ion batteries is crucial for ensuring the steady operation of these systems. This study proposes a prediction framework that combines incremental capacity analysis (ICA) and electrochemical impedance spectroscopy (EIS) to address RUL issues.
Review
Chemistry, Multidisciplinary
Yunhong Che, Xiaosong Hu, Xianke Lin, Jia Guo, Remus Teodorescu
Summary: This paper provides a comprehensive review of aging mechanisms and health prognostic methods for lithium-ion batteries, and discusses the main challenges and research prospects. The complex relationships between aging mechanisms, modes, factors, and types are summarized. Prognostic methods are categorized based on time scales and objectives, followed by detailed reviews and comparative evaluations. Key challenges are presented and potential solutions are discussed. Future trends and new ideas for battery health prognostics are proposed.
ENERGY & ENVIRONMENTAL SCIENCE
(2023)
Article
Thermodynamics
Yue Wu, Zhiwu Huang, Yusheng Zheng, Yongjie Liu, Heng Li, Yunhong Che, Jun Peng, Remus Teodorescu
Summary: This paper proposes a novel velocity prediction method for the full driving cycle of electric vehicles based on spatial-temporal commuting data, and applies the predicted velocity to predictive energy management. A real-time two-stage full driving cycle prediction method is proposed, and a multi-horizon model predictive control method is established to optimize power distribution. Compared with traditional MPC, MH-MPC can reduce 4.2% battery degradation cost with real-time computation requirements satisfied.
ENERGY CONVERSION AND MANAGEMENT
(2023)
Article
Automation & Control Systems
Kai Zhang, Lulu Jiang, Zhongwei Deng, Yi Xie, Jonathan Couture, Xianke Lin, Jingjing Zhou, Xiaosong Hu
Summary: This article proposes a fault diagnosis method for the early detection and assessment of soft internal short-circuit faults in lithium-ion battery packs, ensuring the safe operation of electric vehicles. Fault features are extracted from the data using the incremental capacity curve, making them easier to identify than small voltage differences. The local outlier factor method is then used to detect the early soft internal short-circuit fault by calculating the local outlier factor value of each cell within the battery pack.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Article
Chemistry, Analytical
Sabir Hossain, Xianke Lin
Summary: This paper presents a strategy to obtain a real-time pseudo point cloud from image sensors instead of LiDARs, achieving better performance in terms of real-time operation and accuracy. Different depth estimators are used to generate pseudo point clouds similar to LiDAR, outperforming existing approaches in terms of both depth estimation and point cloud generation speed.
Article
Energy & Fuels
Lin Wang, Xiaowei Zhao, Zhongwei Deng, Lin Yang
Summary: This paper focuses on the accurate estimation of State of Charge (SoC) for electric vehicles and hybrid electric vehicles. A new model updating strategy based on electrochemical impedance spectroscopy (EIS) is proposed, which effectively enhances the SoC estimation accuracy by modifying model parameters and capacity according to the change rate of ohmic impedance.
JOURNAL OF ENERGY STORAGE
(2023)
Article
Automation & Control Systems
Yunhong Che, Daniel-Ioan Stroe, Xiaosong Hu, Remus Teodorescu
Summary: This article proposes a novel semi-supervised self-learning method for battery lifetime prediction. Health indicators are extracted and used to reconstruct historical capacities for self-training of the lifetime model. The self-trained lifetime model is able to predict future degradation and provide probabilistic predictions of future capacities.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2023)
Article
Engineering, Electrical & Electronic
Nan Zhao, Fengqi Zhang, Yalian Yang, Serdar Coskun, Xianke Lin, Xiaosong Hu
Summary: This article proposes a predictive energy management strategy for connected plug-in hybrid electric vehicles (PHEVs) based on real-time dynamic traffic prediction. The strategy involves predicting future traffic information using a wavelet neural network (WNN) and optimizing the parameters of the network using a particle swarm optimization (PSO) algorithm. A long short-term memory-based velocity predictor is also proposed for the strategy. The performance of the strategy is verified using actual traffic data and results show improvements in fuel economy of 17.57% and 28.19% respectively.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Article
Chemistry, Applied
Yunhong Che, Yusheng Zheng, Xin Sui, Remus Teodorescu
Summary: This paper proposes a self-supervised learning framework to improve the performance of battery SoH estimation by using limited labeled data. It solves the challenges of poor generalization, lack of labeled data, and unused measurements during aging.
JOURNAL OF ENERGY CHEMISTRY
(2023)
Article
Chemistry, Applied
Jia Guo, Yunhong Che, Kjeld Pedersen, Daniel-Ioan Stroe
Summary: In this paper, the impedance spectrum of a battery is predicted using the battery charging voltage curve and optimized based on electrochemical mechanistic analysis and machine learning. The internal electrochemical relationships between the charging curve, incremental capacity curve, and the impedance spectrum are explored, improving the physical interpretability and defining the proper partial voltage range for machine learning models. The experimental results show that the predicted errors for impedance spectrum are less than 1.9 mO with the proper partial voltage range selected by the corelative analysis of the electrochemical reactions.
JOURNAL OF ENERGY CHEMISTRY
(2023)
Article
Engineering, Civil
Xiaolin Tang, Guichuan Zhong, Shen Li, Kai Yang, Keqi Shu, Dongpu Cao, Xianke Lin
Summary: In this paper, the fully parameterized quantile network (FPQN) is utilized to estimate the full return distribution and generate uncertainty-aware driving behavior. The proposed method outperforms the baseline methods in terms of safety and can make reasonable decisions in challenging driving cases in the presence of uncertainty.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Engineering, Electrical & Electronic
Yunhong Che, Yusheng Zheng, Yue Wu, Xianke Lin, Jiacheng Li, Xiaosong Hu, Remus Teodorescu
Summary: This paper proposes an online end-to-end state monitoring method based on transferred multi-task learning for battery management systems. The method improves accuracy and computational efficiency under various application scenarios. Experiments show that the proposed method has better accuracy and can be used for monitoring batteries in electric vehicles throughout their entire lifecycle.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2023)
Editorial Material
Chemistry, Physical
Yunhong Che, Xiaosong Hu, Remus Teodorescu
Summary: Lithium-ion batteries are crucial for advancing clean energy globally, especially in power grids and electrified transportation. However, accurately estimating battery health under real-world conditions, particularly in aging mode diagnosis, is challenging due to complex usage conditions and a lack of precise measurement. In a recent issue of Nature Communications, Dubarry et al. address this issue by utilizing machine learning and battery digital twins to diagnose aging modes in photovoltaics-connected batteries that have been operating for 2 years and experienced over 10,000 degradation paths under various seasonal and cloud shading conditions.
Article
Energy & Fuels
Yunhong Che, Soren Byg Vilsen, Jinhao Meng, Xin Sui, Remus Teodorescu
Summary: A novel method for end-to-end sensor-free differential temperature voltammetry reconstruction and state of health estimation is proposed in this paper, which can be used for battery management to ensure safe and optimal usage. The method reconstructs the differential temperature curve using the partial charging or discharging curve, eliminating the need for temperature sensor measurement. The reconstructed differential temperature curve and the partial differential capacity curve are then used for end-to-end state of health estimation.
Review
Thermodynamics
Yusheng Zheng, Yunhong Che, Xiaosong Hu, Xin Sui, Daniel-Ioan Stroe, Remus Teodorescu
Summary: This paper provides a comprehensive review of temperature estimation techniques in battery systems, discussing potential metrics, different estimation methods, and their strengths and limitations in battery management. The challenges and future opportunities in battery thermal state monitoring are also identified and discussed.
PROGRESS IN ENERGY AND COMBUSTION SCIENCE
(2024)
