Factors that cause the self-discharge in valve-regulated sealed lead-acid batteries are discussed and measures to inhibit the self-discharge are put forward.
In the present study, the relationship between properties of different carbon materials and their impact on performance of VRLA （valve regulated lead acid） battery was studied. The material properties undertaken for...In the present study, the relationship between properties of different carbon materials and their impact on performance of VRLA （valve regulated lead acid） battery was studied. The material properties undertaken for the study are： surface area, conductivity and water absorption of the carbon. The electrode morphology revealed the uniform distribution of active material when high surface area carbon was added to NAM （negative active material）. The porosity of the plate also exhibited changes with respect to type of carbon materials added. The study further revealed that, the addition of high surface area carbon （-1,400 m^2/g） improves the charge acceptance of the battery with higher loading. Further improvement in charge acceptance was observed with addition of graphite to higher surface area carbon. Nevertheless, the float current of the battery got affected due to graphite loading and found there was no impact on shelf life of the battery in all the cases. The study demonstrates the need for customized ＂carbon formulation＂ to obtain the maximum performance out of the battery.展开更多
健康状态(state of health,SOH)是数据中心阀控式铅酸电池(value regulated lead acid,VRLA)容量及安全管理的关键指标,而常用的SOH测量方法因检测过程放电时间长、深度大无法满足运营需求,因此对数据中心VRLA电池的健康状态估计研究是...健康状态(state of health,SOH)是数据中心阀控式铅酸电池(value regulated lead acid,VRLA)容量及安全管理的关键指标,而常用的SOH测量方法因检测过程放电时间长、深度大无法满足运营需求,因此对数据中心VRLA电池的健康状态估计研究是非常必要的。针对SOH数据驱动建模存在的估计精度低的问题,提出一种基于时空注意力(spatio-temporal attention,STA)和长短期记忆(long short term memory,LSTM)网络的STA-LSTM深度学习模型。该模型用时空注意力机制在输入数据的特征和时间步上分配注意力权重从而生成新的输入,使用LSTM网络对新的输入进行编码以及实现SOH估计输出。基于电池放电深度50%的数据建模应用结果表明,STA-LSTM模型取得最优估计精度,注意力机制的引入提升黑箱模型的收敛速度、估计精度及物理可解释性。展开更多
文摘Factors that cause the self-discharge in valve-regulated sealed lead-acid batteries are discussed and measures to inhibit the self-discharge are put forward.
文摘In the present study, the relationship between properties of different carbon materials and their impact on performance of VRLA （valve regulated lead acid） battery was studied. The material properties undertaken for the study are： surface area, conductivity and water absorption of the carbon. The electrode morphology revealed the uniform distribution of active material when high surface area carbon was added to NAM （negative active material）. The porosity of the plate also exhibited changes with respect to type of carbon materials added. The study further revealed that, the addition of high surface area carbon （-1,400 m^2/g） improves the charge acceptance of the battery with higher loading. Further improvement in charge acceptance was observed with addition of graphite to higher surface area carbon. Nevertheless, the float current of the battery got affected due to graphite loading and found there was no impact on shelf life of the battery in all the cases. The study demonstrates the need for customized ＂carbon formulation＂ to obtain the maximum performance out of the battery.
文摘健康状态(state of health,SOH)是数据中心阀控式铅酸电池(value regulated lead acid,VRLA)容量及安全管理的关键指标,而常用的SOH测量方法因检测过程放电时间长、深度大无法满足运营需求,因此对数据中心VRLA电池的健康状态估计研究是非常必要的。针对SOH数据驱动建模存在的估计精度低的问题,提出一种基于时空注意力(spatio-temporal attention,STA)和长短期记忆(long short term memory,LSTM)网络的STA-LSTM深度学习模型。该模型用时空注意力机制在输入数据的特征和时间步上分配注意力权重从而生成新的输入,使用LSTM网络对新的输入进行编码以及实现SOH估计输出。基于电池放电深度50%的数据建模应用结果表明,STA-LSTM模型取得最优估计精度,注意力机制的引入提升黑箱模型的收敛速度、估计精度及物理可解释性。
