摘要
提出以对电池衰减老化敏感的电压增量作为特征量,使用铅酸电池全寿命周期数据,利用深度神经网络算法(DNN)建模,进行健康状态(SOH)在线估算,以提高预测的准确性。该方法能准确预测电池的SOH。在不同循环倍率(0.10C、0.15C和0.20C)下,以容量作为估算方法,SOH估算平均误差小于1.0%,最大误差不超过3.0%;在不同(60%、80%和100%)放电深度(DOD)下,以放电截止电压作为估算方法,SOH估算平均误差小于0.5%,最大误差不超过3.0%。
To improve the estimation accuracy,a voltage increment that was sensitive to battery decay and aging was proposed as a characteristic quantity,a deep neural network(DNN)model was built through the full life periodic data of the lead-acid battery to realize the online estimation of state of health(SOH).The SOH of the battery could be estimated accurately through that method.Under different cycle rates(0.10C,0.15C and 0.20C),with capacity as the estimation method,the average error and maximum error of SOH estimation was less than 1.0%and 3.0%respectively.While under different(60%,80%and 100%)depth of discharge(DOD),with discharge cut-off voltage as the estimation method,the average error and maximum error of SOH estimation was less than 0.5%and 3.0%,respectively.
作者
胡晨
金翼
崔邴晗
杜春雨
HU Chen;JIN Yi;CUI Bing-han;DU Chun-yu(National Key Laboratory on Operation and Control of Renewable Energy and Energy Storage,China Electric Power Research Institute,Beijing 100192,China;Institute of Advanced Chemical Power Sources,School of Chemistry and Chemical Engineering,Harbin Institute of Technology,Harbin,Heilongjiang 150001,China)
出处
《电池》
CAS
北大核心
2021年第1期63-67,共5页
Battery Bimonthly
基金
国家电网公司科技项目(DG71-17-009)。
