Nonlinear observer designs for state-of-charge estimation of Lithium-ion batteries
Nonlinear observer designs for state-of-charge estimation of Lithium-ion batteries
复制标题
DOI：
10.1109/acc.2014.6858766
复制
发表时间：
2014-06
期刊：
影响因子：
--
通讯作者：
Satadru Dey;B. Ayalew
中科院分区：
文献类型：
--
作者：
Satadru Dey;B. Ayalew
荷电状态（SOC）信息对于锂离子电池/蓄电池的控制、诊断和监测至关重要。与电池管理系统中经常采用的传统数据驱动或等效电路模型相比，电化学电池模型具有通过跟踪每个电极中的Li离子浓度来给出物理上准确的SOC信息的潜力。在本文中，两个非线性观测器的设计，估计锂离子电池的充电状态的基础上减少的电化学模型。第一个观测器设计使用一个恒定的增益Luenberger结构，而第二个通过加权增益与输出雅可比矩阵来改进它。对于这两个观测器的设计，李雅普诺夫直接法应用和设计问题转化为解决线性矩阵不等式。仿真结果证明了这两种观测器设计的有效性。
State-of-Charge (SOC) information is very crucial for the control, diagnostics and monitoring of Li-ion cells/batteries. Compared to conventional data-driven or equivalent circuit models often employed in battery management systems, electrochemical battery models have the potential to give physically accurate the SOC information by tracking the Li-ion concentration in each electrode. In this paper, two nonlinear observer designs are presented to estimate Li-ion battery State-of-Charge based on reductions of an electrochemical model. The first observer design uses a constant gain Luenberger structure whereas the second one improves it by weighing the gain with the output Jacobian. For both observer designs, Lyapunov's direct method is applied and the design problems are converted to solving LMIs. Simulation results are included to demonstrate the effectiveness of both observer designs.