基于模型的锂离子电池<strong>SOC</strong>估计方法综述

基于模型的锂离子电池SOC估计方法综述

谭必蓉, 杜建华, 叶祥虎, 曹馨, 瞿常

Overview of SOC estimation methods for lithium-ion batteries based on model

Birong TAN, Jianhua DU, Xianghu YE, Xin CAO, Chang QU

表1 常见锂离子电池等效电路模型对比

Table 1 Comparison of common equivalent circuit models of lithium ion batteries

模型名称	状态方程	优点	缺点
Rint模型	U_t =U_OC-i_tR₀	结构简单	精度低，应用性差
Thevenin模型	$\{\begin{array}{l} U_{t} = U_{O C} - i_{t} R_{0} - U_{1} \\ \frac{d U_{1}}{d t} = \frac{i_{t}}{C_{1}} - \frac{U_{1}}{R_{1} C_{1}} \end{array}$	结构较简单，考虑了极化现象，应用性好	对电池特性描述存在局限
双极化模型	$\{\begin{array}{l} U_{t} = U_{O C} - i_{t} R_{0} - U_{1} - U_{2} \\ \frac{d U_{1}}{d t} = \frac{i_{t}}{C_{1}} - \frac{U_{1}}{R_{1} C_{1}} \\ \frac{d U_{2}}{d t} = \frac{i_{t}}{C_{2}} - \frac{U_{2}}{R_{2} C_{2}} \end{array}$	能较好模拟电池动态特性，应用性较好	模型结构较复杂，计算成本较高
PNGV模型	$\{\begin{array}{l} U_{t} = U_{O C} - i_{t} R_{0} - U_{1} - U_{d} \\ \frac{d U_{1}}{d t} = \frac{i_{t}}{C_{1}} - \frac{U_{1}}{R_{1} C_{1}} \\ \frac{d U_{d}}{d t} = \frac{i_{t}}{C_{d}} \end{array}$	考虑了SOC值对开路电压的影响，考虑了电池动态特性	模型结构复杂，计算成本高，应用性一般
GNL模型	$\{\begin{array}{l} U_{t} = U_{O C} - i_{t} R_{0} - U_{1} - U_{2} - U_{d} \\ \frac{d U_{d}}{d t} = \frac{i_{t}}{C_{d}} + \frac{U_{O C}}{R_{d} C_{d}} - \frac{U_{d}}{R_{d} C_{d}} - \frac{U_{1}}{R_{d} C_{d}} - \frac{U_{2}}{R_{d} C_{d}} \\ \frac{d U_{1}}{d t} = \frac{i_{t}}{C_{1}} + \frac{U_{O C}}{R_{d} C_{1}} - \frac{U_{d}}{R_{d} C_{1}} - \frac{U_{1}}{R_{d} C_{1}} - \frac{U_{2}}{R_{d} C_{1}} - \frac{U_{1}}{R_{1} C_{1}} \\ \frac{d U_{2}}{d t} = \frac{i_{t}}{C_{2}} + \frac{U_{O C}}{R_{d} C_{2}} - \frac{U_{d}}{R_{d} C_{2}} - \frac{U_{1}}{R_{d} C_{2}} - \frac{U_{2}}{R_{d} C_{2}} - \frac{U_{2}}{R_{2} C_{2}} \end{array}$	能准确模拟电池动态特性，且考虑了自放电	模型结构过于复杂，应用性差