【编辑推荐】2024年2期科研论文、新能源

王岩松^1,2，陈顺^1,2，范国栋^1,2*，张希^1,2

（1.上海交通大学机械与动力工程学院，上海 200240； 2.上海交通大学国家汽车动力与智能控制工程技术研究中心，上海 200240）

摘要：锂离子电池由于高能量密度、低自放电率以及较长的生命周期，成为目前最为理想的储能设备之一，在电动汽车以及储能电站领域发挥着重要作用。随着锂离子电池的不断发展，对锂离子电池的能量密度的需求不断提高，使用三元正极材料的高能量密度电池越来越受欢迎。然而，三元锂离子电池虽然具有较高的能量密度，但材料稳定性及安全性相对较弱，容易引发热失控。热失控多是由于电池滥用造成，电滥用则是最为常见的滥用方式。

电动汽车的电池包是由电池单体串并联成组，通过主动或者被动均衡的方式保证各个电池单体的一致性。但是由于电池的不一致性、电池包均衡设计不足或者充电过程电池管理系统数据通讯故障等问题，导致部分电池单体在充电过程中超出其截止电压，出现不同程度的微过充现象。但微过充对于电池容量的影响尚未被完全揭示。

本文作者通过将三元镍钴铝（Li_xNi_0.80Co_0.15Al_0.05O₂,NCA）电池在不同最高截止电压（4.2 V、4.3 V和4.4 V）下进行恒流充电循环老化实验，采用差分电压分析、电化学阻抗（EIS）谱等原位测试方法，探究高截止电压循环对电池老化的影响。结果表明，随着充电最高截止电压的升高，电池老化速度加快，280次循环后（2.5 A 4.2 V、2.5 A 4.3 V、2.5 A 4.4 V），容量保持率分别为90.30%、88.19%和86.19%。差分电压分析进一步表明，容量衰减加速的原因在于正极活性物质损失和活性Li⁺损失。EIS测试表明，微过充导致固体电解质相界面（SEI）膜阻抗略微升高，电荷转移阻抗升高，且程度随截止电压的升高而增大。

Aging mechanism of ternary NCA Li-ion battery under slight overcharge

WANG Yansong^1,2, CHEN Shun^1,2, FAN Guodong^1,2∗, ZHANG Xi^1,2

(1. School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China; 2. National Engineering Research Center of Automotive Power and Intelligent Control, Shanghai Jiao Tong University, Shanghai 200240, China)

Abstract: Li-ion batteries, due to their high energy density, low self-discharge rates, and extended lifespan, stand out as one of the most promising energy storage devices, a crucial role in both electric vehicles and energy storage facilities. With the continuous evolution of Li-ion batteries, there is a growing demand for higher energy density, leading to increased popularity of high-energy-density batteries using ternary cathode materials. However, despite the higher energy density exhibited by ternary Li-ion batteries, their material stability and safety are relatively weaker and prone to thermal runaway incidents. Thermal runaway incidents are often triggered by battery abuse, with electrical abuse being the most common form of abuse.

The battery pack of electric vehicles is composed of individual batteries connected in series and parallel configurations, ensuring uniformity among the battery cells through active or passive balancing methods. However, due to cell disparities, inadequate battery pack balancing designs, or issues such as communication faults in the battery management system during the charging process, certain battery cells may exceed their cutoff voltage, resulting in varying degrees of mild overcharging during charging. However, the effect of mild overcharging on battery capacity has not been fully revealed.

Constant-current charging and cycling aging experiments on NCA(Li_xNi_0.80Co_0.15Al_0.05O₂) batteries at different maximum cutoff voltages (4.2 V, 4.3 V, and 4.4 V) are conducted and employed electrochemical performance testing methods such as differential voltage analysis and electrochemical impedance spectroscopy(EIS) to investigate the effects of high cut-off voltage cycling on battery aging. The results indicate that as the maximum cutoff voltage increased, the rate of battery aging accelerates. After 280 cycles（2.5 A 4.2 V、2.5 A 4.3 V、2.5 A 4.4 V）, the capacity retention is 90.30%, 88.19% and 86.19%, respectively. Differential voltage analysis further reveals that the accelerated capacity decay is due to the loss of cathode active materials and active Li⁺. EIS testing results disclose that mild overcharging led to a slight increase in the solid electrolyte phase interface (SEI) film impedance and an increase in the charge transfer impedance, the extent of which increases with the increase in the cut-off voltage.

柏帆¹, 王路达¹, 左红群¹, 谢长君^2∗

(1. 宁海县雁苍山电力建设有限公司,浙江 宁波 315600; 2. 武汉理工大学自动化学院,湖北 武汉 430070)

摘要：在“碳达峰”和“碳中和”的战略背景下，开发适应波动性输入的大规模储能技术是构建新型电力系统的重要途径。发展储能产业是实现“双碳”目标和解决太阳能和风能发电不连续问题的重要手段。质子交换膜燃料电池（PEMFC）是一种氢点转换装置，绿色节能、能量转换率高、低温下快速启动和噪声小等优点，但寿命短和维护成本高的问题限制了其大规模应用。准确估计PEMFC的退化趋势有利于电池运行与延寿，具有重要意义。

因此，提出由麻雀搜索算法（SSA）-优化回声状态网络（ESN）的储备池尺寸、泄漏率和正则化系数的组合算法，将燃料电池电堆电压、工作电流及其对应的时间作为老化数据，挖掘电堆在给定电流及老化时间下的变化规律，从而进行长期老化预测。SSA-ESN的结构图如图1所示。所提的SSA-ESN在法国燃料电池实验室的数据集下得到验证，静态工况下，训练集比例为30%、40%、50%和60%时，SSA-ESN相比于EKF，在误差指标RMSE上分别减少了82.48%、81.20%、12.88%和4.54%。在准动态工况下，在训练集比例为60%时，SSA-ESN算法的长期预测误差指标RMSE为0.018 9，相较于EKF算法提升了47.06%。因此，该方法的预测准确性得到验证，可实现高精度的PEMFC长期老化预测。

Long-term degradation prediction of proton exchange membrane fuel cell for energy storage

BO Fan ¹ ,WANG Luda ¹,ZUO Hongqun ¹ ,XIE Changjun ^2∗

(1. Ninghai County Yancangshan Electric Power Construction Co., Ltd., Ningbo, Zhejiang 315600,China; 2. School of Automation, Wuhan University of Technology, Wuhan, Hubei 430070, China )

Abstract: In the context of the strategies for “carbon peak” and “carbon neutrality,” the development of large-scale energy storage technologies that can adapt to fluctuating inputs is crucial for building a new type of power system. The growth of the energy storage industry is essential for achieving the “dual carbon” goals and addressing the intermittent nature of solar and wind energy generation. Proton exchange membrane fuel cells (PEMFCs) are hydrogen conversion devices known for their green energy efficiency, high energy conversion rates, rapid startup at low temperatures, and minimal noise. However, their short lifespan and high maintenance costs limit their large-scale application. Accurately estimating the degradation trend of PEMFCs is beneficial for battery operation and longevity, making it of significant importance.

Therefore, a combined algorithm that optimizes the reservoir size, leakage rate, and regularization coefficient of an Echo State Network (ESN) using the Sparrow Search Algorithm (SSA) is proposed. The aging data includes fuel cell stack voltage, operating current, and corresponding time. By mining the variation patterns of the fuel cell stack under given a current and aging time, long-term aging predictions can be made. The structure diagram of SSA-ESN is shown in Fig 1.

The proposed SSA-ESN has been validated using data from the French Fuel Cell Laboratory. Under static conditions, when the training set proportions are 30%, 40%, 50%, and 60%, SSA-ESN outperforms EKF in terms of the root mean square error (RMSE) by reducing it by 82.48%, 81.20%, 12.88%, and 4.54%, respectively. In quasi-dynamic conditions, with a training set proportion of 60%, the SSA-ESN algorithm achieves an RMSE of 0.018 9, representing a 47.06% improvement over the EKF algorithm. Consequently, this method’s predictive accuracy has been verified, enabling high-precision long-term aging predictions for PEMFCs.

图1 老化预测策略

Fig.1 The strategy of degradation prediction

杨瑞志¹，武鑫^1∗，熊星宇¹，胡亮²

[1. 华北电力大学能源动力与机械工程学院，北京 102206；2. 中国矿业大学（北京）机械与电气工程学院，北京 100083]

摘要：固体氧化物燃料电池（SOFC）电堆通常在700 ℃以上的高温下工作。电堆中的密封材料在高温下易退化失效，导致泄漏故障，引发电堆的热失控和损坏,影响系统运行稳定性。提出一种基于电堆温度和电压信号的经验小波变换（EWT）诊断方法。通过EWT分解温度和电压信号，得到多个多分辨率分析（MRA）分量，分析其中故障特征明显的MRA信号，求出时域特征方差，通过设定的方差阈值判断泄漏是否发生。通过千瓦级电堆实验平台的温度与电压数据，证明了EWT诊断方法可较好地检测电堆泄漏故障，如图1所示。相较于电压信号，通过分析温度信号可以提前约10 s检测到电堆泄漏故障，如图2所示。提出的方法，对于提前发现电堆泄漏故障具有指导意义。同时，对比集合经验模态分解方法与EWT方法，EWT方法可更快地检测到电堆泄漏故障。从快速诊断电堆泄漏故障的角度考虑，EWT方法的性能优于EEMD方法。

SOFC leakage fault diagnosis based on empirical wavelet transform

YANG Ruizhi¹, WU Xin^1∗, XIONG Xingyu¹, HU Liang²

(1. School of Energy Power and Mechanical Engineering, North China Electric Power University, Beijing 102206, China; 2. School of Mechanical and Electrical Engineering, China University of Mining and Technology-Beijing, Beijing 100083, China)

Abstract: Solid oxide fuel cell (SOFC) stacks usually operate at high temperatures above 700 ° C. The sealing material in the reactor is easy to degenerate and fails at high temperatures, leading to leakage failure, causing thermal runaway and damage to the reactor, and affecting the stability of the system. An empirical wavelet transforms (EWT) diagnosis method based on the temperature and voltage signals of the stack is presented. Multiple multi-resolution analysis (MRA) components are obtained through EWT decomposition of temperature and voltage signals. The MRA signals with obvious fault characteristics are analyzed, and the time-domain feature variance is obtained. The leakage is judged by the variance threshold set. Through the temperature and voltage data of the kilowatt-level reactor experiment platform, it is proved that the EWT diagnosis method can detect the reactor leakage fault well, as shown in Fig.1. Compared with the voltage signal, the reactor leakage fault can be detected about 10 s in advance by analyzing the temperature signal, as shown in Fig.2. The proposed method has a guiding significance for detecting leakage faults of the reactor in advance. At the same time, compared with the ensemble empirical mode decomposition method and EWT method, the EWT method can detect the reactor leakage fault faster. The EWT method is superior to the EEMD method in the rapid diagnosis of reactor leakage faults.

图1 故障电堆与正常电堆的温度方差对比

Fig.1 Comparison of temperature variance between faulty stack and normal stack

图2 EWT方法的故障电堆温度与电压方差对比

Fig.2 Comparison of temperature and voltage variance of fault stack using EWT method

吴志辉，钱晶，余沁芸，王亚珍^∗

(江汉大学光电材料与技术学院，光电化学材料与器件教育部重点实验室，湖北 武汉 430056)

摘要：双金属MOF-74具有独特的结构性能，是构建多孔材料的理想前驱体之一。以ZnCo-MOF-74作为前驱体，通过高温煅烧法制备衍生物ZnO/ZnCo₂O₄纳米球，对材料的结构和形貌进行分析，并探讨双金属比例对材料的电化学性能的影响。图1为不同比例的双金属前驱体煅烧后得到的衍生物在6 mol/L KOH溶液的三电极体系中的循环伏安曲线，在n(Zn)∶n(Co)=1∶1时得到的循环伏安曲线面积最大。恒流充放电结果也显示，在电流为1.0 A/g、n(Zn)∶n(Co)=1∶1时制得的ZnO/ZnCo₂O₄的比电容高达1 047.8 F/g，高于其他比例下所得到的混合金属氧化物。ZnO/ZnCo₂O₄的高比电容和良好的倍率性能归功于其较大的比表面积和双金属的协同效应。以n(Zn)∶n(Co)=1∶1合成的纳米球ZnO/ZnCo₂O₄作为正极材料、活性炭作为负极材料制备不对称超级电容器（ZnO/ZnCo₂O₄//AC），电化学性能测试显示，在0.5 A/g下比电容为74.3 F/g，在功率密度536.3 W/kg时，能量密度为26.4 W·h/kg。该简易器件还具有较好循环稳定性，经2 000次循环后，比电容仍保持81.3%。

Supercapacitor performance for ZnCo-MOF-74-derived nanospheres ZnO/ZnCo₂O₄

WU Zhihui, QIAN Jing, YU Qinyun, WANG Yazhen^∗

(Key Laboratory of Optoelectronic Chemical Materials and Devices of Ministry of Education, School of Optoelectronic Materials & Technology, Jianghan University, Wuhan, Hubei 430056, China)

Abstract: With the unique structural properties, bimetal MOF-74 are one of the ideal precursors for building porous materials. ZnCo-MOF-74 is used as a precursor to prepare derivative ZnO/ZnCo₂O₄ nanospheres by high-temperature calcination method, its structure and morphology are analyzed, and the influence of bimetallic ratio on the electrochemical performance of materials is explored. Fig.1 shows the cyclic voltammetric curve, tested in a three-electrode system with 6 mol/L KOH solution for the derivants calcining from precursors with different bimetallic proportions, and the area of cycilc voltammetric diagram is the largest for n(Zn)∶n(Co)=1∶1. The specific capacitance of ZnO/ZnCo₂O₄ is as high as 1 047.8 F/g with n(Zn)∶n(Co)=1∶1 at a current of 1.0 A/g, which is larger than those of the mixed metal oxides obtained at other ratios for galvanostat charge-discharge. The high specific capacitance and excellent multiplicative properties of ZnO/ZnCo₂O₄ are attributed to its large specific surface area and the synergistic effect of the bimetal. The asymmetric supercapacitor (denoted as ZnO/ZnCo₂O₄//AC) is prepared with ZnO/ZnCo₂O₄ as the cathode material and activated carbon as the anode material. The electrochemical results indicate a specific capacitance value of 74.3 F/g at 0.5 A/g. The corresponding energy density is 26.4 W·h/kg at a power density of 536.3 W/kg. The simple device has long-term cycling stability and the specific capacitance can still maintain 81.3% of its initial after 2 000 cycles.

图1 6种样品的CV曲线

Fig.1 CV curves of 6 samples

郝大程^∗，王城铭，李成勋，宋超

(大连交通大学环境与化学工程学院，辽宁 大连 116028)

摘要：无膜微流体燃料电池（MFC）利用微通道内多股流体的层流优势分离阳极和阴极，因此无需使用膜。此前报道的大多数MFC都需要外部泵来维持燃料电池中的流体流通，这不仅增加了系统的总体容量，而且限制了集成化和小型化。毛细管流动可以在微流体装置中承担泵送任务，由于多孔纸或纤维基材料中存在毛细作用驱动的层流，可以不使用外部泵，从而节约成本。目前报道以棉线为基底的MFC很少。搭建了以水合肼为燃料、石墨为电极以及棉线为流道材料的无催化剂微流体燃料电池。以水合肼为燃料，过氧化氢为氧化剂，NaOH和H₂SO₄分别为阳极及阴极的电解质作为微流体燃料电池的原材料，对电池性能进行测试。实验旨在探究能否在不使用催化剂及降低电池搭建成本的情况下，尽可能提高电池的功率密度，使其具有更好的应用前景。结果显示，石墨管阳极、石墨棒阴极组合的MFC电化学性能最佳，电池于75 mmol/L燃料浓度、1 mol/L电解质及4 mm电极距离下，取得了11.1 μW/cm²的最大功率密度和121.45 μA/cm²的电流密度。与文献报道的纸基肼MFC的性能相比，最大功率密度提升了7.5倍，极限电流密度提升了7.5倍。所搭建的MFC具有材料简单、成本较低且组装方便等特点，无需外部泵维持共层流，显示出作为便携式设备的集成化和小型化的潜力。

Cotton thread-based hydrazine microfluidic fuel cell with graphite electrodes

HAO Dacheng^∗, WANG Chengming, LI Chengxun, SONG Chao

(School of Environment and Chemical Engineering, Dalian Jiaotong University, Dalian, Liaoning 116028, China)

Abstract: Membraneless microfluidic fuel cell (MFC) uses the advantage of laminar flow of multiple fluids in microchannels to separate anode and cathode, so there is no need to use membrane. Most previously reported MMFC require external pumps to maintain fluid circulation in fuel cells. This not only increases the overall capacity of the system, but also limits the integration and miniaturization of cells. Capillary flow is a good substitute for pumps in microfluidic devices. Because of the capillary-driven laminar flow in porous paper or cotton-fiber-based materials, external pumps can be avoided, resulting in cost savings. There are few reports about MFC based on cotton thread. A catalyst-free MFC with hydrazine hydrate as fuel, graphite as an electrode, and cotton wire as a channel material is built. Hydrazine sulfate and barium hydroxide are used to prepare the fuel hydrazine hydrate, and hydrogen peroxide is used as the oxidant. NaOH and H₂SO₄ are used as electrolytes on the anode and cathode, respectively. The purpose of the experiment is to explore whether the power density of the cell can be improved as much as possible without using a catalyst and reducing the construction cost of the cell, to have a better application prospect. It is found that the electrochemical performance of MFC with a graphite tube anode and a graphite rod cathode is the best. The cell performance is optimized with 75 mmol/L fuel, 1 mol/L electrolyte, and 4 mm electrode spacing, and the maximal power density of 11.1 μW/cm² and a current density of 121.45 μA/cm² are achieved. Compare with the reported performance of hydrazine MFC, the maximum power density and the limiting current density have been increased by 7.5 times. The MFC constructed has the characteristics of simple materials, low cost, and convenient assembly, and no external pumps are required to maintain the co-laminar flow, showing potential of integration and miniaturization as portable devices.

图1  MFC的结构

Fig.1  Structure of MFC

梁宏毅¹，黎华玲²，甘友毅¹，邵丹^2∗

(1. 广汽本田汽车有限公司，广东 广州 510700；2. 广州能源检测研究院，广东省动力电池安全重点实验室，广东 广州 511447)

摘要：为探究动力锂离子电池的低温特性，对比制造工艺相同但低温性能差异较大（A组性能稍好）的不同批次的A、B两组三元动力锂离子电池在低温（-35 ℃）条件下的容量差异。建立“电池-材料”联动分析方法，从内部材料微观特征分析电池低温容量失效的原因。三元正极材料结构是影响电池低温容量的关键因素。与A组电池相比，B组电池的正极材料颗粒结构松散,经过低温放电后，颗粒表面出现裂纹，导致B组正极材料在低温放电过程中活性Li⁺损失、固体电解质相界面（SEI）膜阻抗增加、材料电导率差，继而导致低温容量失效。

Failure analysis of low-temperature capacity of ternary power Li-ion battery

LIANG Hongyi¹, Li Hualing², GAN Youyi¹, SHAO Dan^2∗

(1. GAC Honda Automobile Co., Ltd., Guangzhou, Guangdong 510700, China; 2. Guangdong Key Laboratory of Battery Safety, Guangzhou Institute of Energy Testing, Guangzhou, Guangdong 511447, China)

Abstract: In order to explore the low-temperature performance of power Li-ion battery, the capacity difference of different batches of two groups of A and B ternary power Li-ion batteries with the same manufacturing process but different low-temperature performance(the performance of group A is better) at low temperature (-35 ℃) is compared. A " battery-material" linkage analysis method is established to analyze the causes of low-temperature capacity failure of batteries from the microscopic characteristics of internal materials. The structure of ternary cathode material is the key factor affecting the low-temperature capacity of battery. Compared with group A, the cathode material particles of group B have looser particle-structure, cracks appeared on the surface of the particles after discharge, resulting in the loss of active Li ⁺ of the cathode materials of group B during the lowtemperature discharge process, the increase of solid electrolyte interface (SEI) film impedance, the poor conductivity of the material, which leads to the failure of low-temperature capacity.

图1 “电池-材料”联动分析流程图

Fig.1 “Battery-material” linkage analysis flowchart

图2 电池在不同温度条件下的EIS及等效电路图

Fig.2 EIS of batteries under different temperature conditions and equivalent circuit diagram

王迪¹，曹以龙²，杜君莉^3∗

(1. 郑州电力高等专科学校电力工程学院，河南 郑州 450000；2. 上海电力大学电子与信息工程学院，上海 200000；3. 国网河南省电力公司电力科学研究院，河南 郑州 450000)

Parameter identification of Li-ion battery model by MAFFRLS algorithm

WANG Di¹, CAO Yilong², DU Junli^3∗

(1. School of Electric Power Engineering, Zhengzhou Electric Power College, Zhengzhou, Henan 450000, China; 2. College of Electronics and Information Engineering, Shanghai University of Electric Power, Shanghai 200000, China; 3. State Grid Henan Electric Power Company Electric Power Science Research Institute, Zhengzhou, Henan 450000, China)

Abstract：The modeling method and the method of model parameter identification will affect the accurate estimation of the Li-ion battery state, especially under dynamic conditions. Therefore, the method of online identification of battery model parameters is very important.

An improved adaptive forgetting factor recursive least squares (MAFFRLS) algorithm is used to identify the parameters of the second-order RC equivalent circuit model of the battery and validated with the constant-current condition and dynamic condition data in the driving environment of an electric vehicle.

Under constant current condition, the results of terminal voltage error of the same type of algorithms under constant current charging and discharging conditions are not much different, and the error differences of the three algorithms under the same condition are not much different, which indicates that the three parameter identification algorithms mentioned above are close to each other under constant current condition. Under the DST condition, the MAE of the RLS, FFRLS and MAFFRLS algorithms are 0.010 2 V, 0.009 9 V and 0.004 6 V, respectively, and the RMSE is 0.015 5 V, 0.015 0 V and 0.006 8 V. The MAE and RMSE of the MAFFRLS algorithm are even smaller under the DST condition. The MAE and RMSE of the MAFFRLS algorithm are smaller in DST condition. Therefore, the superiority of the proposed MAFFRLS algorithm can be better demonstrated in DST condition.

The MAFFRLS algorithm can reflect the dynamic phenomenon of frequent current switching. The predicted terminal voltage errors are compared with the results of the RLS and FFRLS algorithms under constant-current and DST conditions. The results show that the voltage rms error values using the MAFFRLS algorithm are smaller than those using the RLS and FFRLS algorithms, and the response time of the MAFFRLS algorithm is also shorter, so the algorithm can be completed faster. Subsequently the algorithm can also be used for model-based estimation of charge state and health state.

田爱娜，潘壮壮，吴铁洲，姜久春^∗

(湖北工业大学太阳能高效利用及储能运行控制湖北省重点实验室，湖北 武汉 430068)

摘要：潜在的热失控风险，特别是过充电，可能会引发锂离子电池的安全事故，而目前电池故障诊断技术提取的外部特征信号无法及时反映电池内部状态以阻止安全事故的发生。以电化学阻抗谱（EIS）为基础，采用单频阻抗X_1Hz和X_400Hz实现电池内部阻抗的实时检测。在不同充电倍率、环境温度和健康状态下进行验证，特征阻抗表现出很好的稳定性。基于X_1Hz和X_400Hz的单频阻抗，提出一种热失控安全分级预警方法，并验证该方法的可靠性。该方法的最后一级预警可在热失控前5 min以上发出安全预警。

Thermal runaway graded warning of Li-ion battery based on single-frequency impedance

TIAN Aina, PAN Zhuangzhuang, WU Tiezhou, JIANG Jiuchun^∗

(Hubei Key Laboratory for High-Efficiency Utilization of Solar Energy and Operation Control of Energy Storage System, Hubei University of Technology, Wuhan, Hubei 430068, China)

Abstract： The potential risk of thermal runaway, especially overcharging, may lead to safety accidents in Li-ion batteries, while the external characteristic signals extracted by current battery fault diagnosis techniques cannot reflect the internal state of the battery in time to prevent safety accidents from occurring. Based on electrochemical impedance spectroscopy (EIS), the real-time detection of internal battery impedance is achieved by employing single-frequency impedance at X_1Hz and X_400Hz. After validation under different charging rates, environmental temperatures, and states of health, the characteristic impedance demonstrates remarkable stability. Utilizing single-frequency impedance at X_1Hz and X_400Hz, a thermal runaway safety classification and warning method is proposed, and its reliability is verified. The final-stage warning of this method can issue a safety warning more than 5 min before thermal runaway.