基于ESO-UKF的动力电池内部温度在线估计  

Battery Internal Temperature Online Estimation Method through ESO-UKF

作　　者：史永胜[1] 左玉洁 符政 刘博亲 王凡[1] JAMSHER Ali SHI Yongsheng;ZUO Yujie;FU Zheng;LIU Boqin;WANG Fan;JAMSHER Ali(College of Electrical and Control Engineering,Shaanxi University of Science&Technology,Xi'an Shaanxi 710021,China)

机构地区：[1]陕西科技大学电气与控制工程学院,陕西西安710021

出　　处：《电子器件》2022年第1期160-166,共7页Chinese Journal of Electron Devices

基　　金：国家自然科学基金项目(61871259);陕西省科技厅工业科技攻关计划项目(2019GY-175)。

摘　　要：准确的内部温度估计对动力电池的安全使用至关重要,为了在线获取准确的电池内部温度,提出一种基于温度估计模型的ESO-UKF电池内部温度估计方法。其中温度估计模型由Bernardi生热模型与热路传热模型组成,生热模型中端电压由神经网络获取,传热模型参数由递推最小二乘法辨识得到;该算法利用温度估计模型的离散状态空间描述,提出ESO-UKF进行电池内部温度的在线估计,将影响估计精度的传感器偏差视为扩展状态与原状态一起估计,实现了对不确定状态的估计;测试验证表明该算法的估计误差小于1℃,能够实现多工况下内部温度的在线估计,估计精度高、适应性强。Accurate internal temperature is very important for the safe use of battery. In order to obtain the battery internal temperature online, an ESO-UKF internal temperature estimation method based on temperature estimation model was bulit. The temperature estimation model combines the Bernardi heat generation model and the heat transfer model. In the heat generation model, the terminal voltage is obtained by the neural network, and the heat transfer model parameters are identified by the recursive least-squares. Using the discrete state space description of the temperature estimation model, ESO-UKF algorithm was proposed to estimate the internal temperature of the battery online. The sensor bias is treated as an extended state to be estimated together with the original state together. The test results show that the estimation error of the algorithm is less than 1 ℃,which can realize the online estimation of internal temperature under multiple working conditions. The estimation accuracy is high and the adaptability is strong.

关 键 词：动力电池 内部温度估计 无迹卡尔曼滤波 扩展状态观测器 

分 类 号：TM911[电气工程—电力电子与电力传动]