电动汽车电池的充电状态和健康状况评估：关键问题和挑战

使用电动汽车（EV）进行运输被视为实现可持续发展和解决环境问题的必要组成部分。当前对环境的关注,例如化石燃料的快速消耗,空气污染的增加,能源需求的加速增长,全球变暖和气候变化,为交通运输部门的电气化铺平了道路。电动汽车可以解决上述问题。电源已成为电动汽车发展的关键,尤其是锂离子（Li-ion）电池。由于其能量密度、功率...     

Review on state of health estimation of retired lithium-ion batteries

由于电池组中电池单体之间存在性能差异,退役锂离子电池在投入梯次利用前需要借助健康状态（SOH）评估技术进行电池单体的分类与配组。健康状态评估系统的构建涉及电池建模、电池测试、数据处理、算法开发等各种技术问题。目前通过基于模型的参数识别与直接提取健康因子是构建SOH评估体系的两种主要思路。在电池模型的简化、测试工况的设计、健康因子的选择和算法的应用与优化等方面已经有了很多研究。如何在缩短电池测试时间的同时提高评估系统的泛化能力是目前该研究领域的主要问题,这些问题的解决对于SOH评估系统真正在梯次利用锂离子电池的产业化中发挥作用至关重要。在未来的研究中通过优化测试工况和数据融合等技术,有望开发出性能更好的SOH评估系统。     

Discrimination of Li-ion batteries based on Hamming network using discharging-charging voltage pattern recognition for improved state-of-charge estimation

Differences in electrochemical characteristics among Li-ion batteries and factors such as temperature and ageing result in erroneous state-of-charge (SoC) estimation when using the existing extended Kalman filter (EKF) algorithm. This study presents an application of the Hamming neural network to the identification of suitable battery model parameters for improved SoC estimation. The discharging-charging voltage (DCV) patterns of ten fresh Li-ion batteries are measured, together with the battery parameters, as representative patterns. Through statistical analysis, the Hamming network is applied for identification of the representative DCV pattern that matches most closely of the pattern of the arbitrary battery to be measured. Model parameters of the representative battery are then applied to estimate the SoC of the arbitrary battery using the EKF. This avoids the need for repeated parameter measurement. Using model parameters selected by the proposed method, all SoC estimates (off-line and on-line) based on the EKF are within ±5% of the values estimated by ampere-hour counting.     

Use of lithium-ion batteries in electric vehicles

An account is given of the lithium-ion (Li-ion) battery pack used in the Northern Territory University's solar car, Fuji Xerox Desert Rose, which competed in the 1999 World Solar Challenge (WSC). The reasons for the choice of Li-ion batteries over silver–zinc batteries are outlined, and the construction techniques used, the management of the batteries, and the battery protection boards are described. Data from both pre-race trialling and race telemetry, and an analysis of both the coulombic and the energy efficiencies of the battery are presented. It is concluded that Li-ion batteries show a real advantage over other commercially available batteries for traction applications of this kind.     

Illustration of experimental,machine learning,and characterization methods for study of performance of Li-ion batteries

The development of fault diagnosis of Li-ion batteries used in electric vehicles is vital. In this perspective, the present work conducted a comprehensive study for the evaluation of coupled and interactive influence of charging ratio, number of cycles, and voltage on the discharge capacity of Li-ion batteries to predict the life of battery. The charging-discharging experimental tests on Li-ion batteries have been performed. The data such as charging ratio, number of cycles, voltage, and discharge capacity of Li-ion batteries are measured. Machine learning approach of neural networks is then applied on the obtained data to compute the effects, normal distribution, parametric analysis, and sensitivity analysis of the input parameters on the capacity of battery. It can be noticed that discharge capacity increased with an increase in full voltage. Further, it has been observed from the sensitivity analysis that the full voltage is most relevant parameters to the capacity of the battery. Additionally, scanning electron microscopy/energy dispersive spectroscopy (SEM/EDS) of the electrodes before and after experiments have been performed, to investigate the elemental dissolution due to the charging/discharging cycles. The findings and analysis from the proposed study shall facilitate experts in making decisions on the remaining life and charging capacity of the battery.     

Li-ion battery state of health estimation through Gaussian process regression with Thevenin model

This paper proposes a model-based and data-driven joint method to estimate the state of health of Li-ion batteries. To accurately quantify battery degradation, a novel resistance-based aging feature is defined from the Thevenin model, and the defined aging feature is approximately linear with capacity degradation. An orthogonal experimental design and a two-way analysis of variance are used to validate the robustness of the defined aging feature. Considering the influence of temperature on battery performance, Box-Cox transformation is introduced to improve the aging feature linearity at low temperatures. Then, an estimator for state of health is established by using Gaussian process regression. Battery aging experiments are conducted to illustrate the estimation effect of the proposed method. The experimental results show that the proposed method has high estimation accuracy at different temperatures. Using the same aging feature, the backpropagation network and support vector regression are implemented to verify the generality of the estimation framework.     

基于深度神经网络的梯次利用电池健康状态评估

随着大量退役电池梯次利用,对退役动力电池健康状态的准确估计是保障电池梯次利用安全高效运行的前提。针对上述问题,提出基于深度神经网络学习的梯次利用电池健康状态评估方法。根据不同循环次数下梯次利用电池充放电性能的差异性,从梯次利用电池物理特性角度挖掘影响梯次利用电池老化特征的主要参数,利用皮尔逊法计算电池老化特征与梯次利用电池健康状态的相关系数,选取较高相关度特征作为深度神经网络的输入,建立基于深度神经网络学习的梯次利用电池健康状态评估模型。通过美国国家航空航天局Ames卓越预测中心的锂离子电池测试数据仿真实例验证了该文方法的有效性。仿真结果表明,与传统神经网络相比,深度神经网络学习可明显提高梯次利用电池健康状态的预测精度,为退役动力电池健康状态评估提供理论依据。     

Investigating electrical contact resistance losses in lithium-ion battery assemblies for hybrid and electric vehicles

Lithium-ion (Li-ion) batteries are favored in hybrid-electric vehicles and electric vehicles for their outstanding power characteristics. In this paper the energy loss due to electrical contact resistance (ECR) at the interface of electrodes and current-collector bars in Li-ion battery assemblies is investigated for the first time. ECR is a direct result of contact surface imperfections, i.e., roughness and out-of-flatness, and acts as an ohmic resistance at the electrode-collector joints. A custom-designed testbed is developed to conduct a systematic experimental study. ECR is measured at separable bolted electrode connections of a sample Li-ion battery, and a straightforward analysis to evaluate the relevant energy loss is presented. Through the experiments, it is observed that ECR is an important issue in energy management of Li-ion batteries. Effects of surface imperfection, contact pressure, joint type, collector bar material, and interfacial materials on ECR are highlighted. The obtained data show that in the considered Li-ion battery, the energy loss due to ECR can be as high as 20% of the total energy flow in and out of the battery under normal operating conditions. However, ECR loss can be reduced to 6% when proper joint pressure and/or surface treatment are used. A poor connection at the electrode-collector interface can lead to a significant battery energy loss as heat generated at the interface. Consequently, a heat flow can be initiated from the electrodes towards the internal battery structure, which results in a considerable temperature increase and onset of thermal runaway. At sever conditions, heat generation due to ECR might cause serious safety issues, sparks, and even melting of the electrodes.     

Online state of health prediction method for lithium-ion batteries,based on gated recurrent unit neural networks

Online state of health (SOH) prediction of lithium-ion batteries remains a very important problem in assessing the safety and reliability of battery-powered systems. Deep learning techniques based on recurrent neural networks with memory, such as the long short-term memory (LSTM) and gated recurrent unit (GRU), have very promising advantages, when compared to other SOH estimation algorithms. This work addresses the battery SOH prediction based on GRU. A complete BMS is presented along with the internal structure and configuration parameters. The neural network was highly optimized by adaptive moment estimation (Adam) algorithm. Experimental data show very good estimation results for different temperature values, not only at room value. Comparisons performed against other relevant estimation methods highlight the performance of the recursive neural network algorithms such as GRU and LSTM, with the exception of the battery regeneration points. Compared to LSTM, the GRU algorithm gives slightly higher estimation errors, but within similar prediction error range, while needing significantly fewer parameters (about 25% fewer), thus making it a very suitable candidate for embedded implementations.     

Autonomous health management design for lithium-ion battery in middle and high orbit satellites

针对中高轨卫星蓄电池组在轨管理问题,在分析某型号在轨使用需求和锂离子蓄电池特性的基础上,提出了锂离子蓄电池在轨自主健康管理系统的设计,并在某MEO轨道卫星上进行了验证,通过在轨数据分析,验证了所设计系统的有效性,为后续型号锂离子蓄电池组在轨自主管理提供经验。