Identification of lithium-ion-battery-cells based on physically interpretable models for battery state estimation during vehicle operation

电池管理系统(BMS)是现代电动链的关键部件。除了安全功能外，有效的电池管理系统还可以完美地使用电池本身，大大延长车辆电池的使用寿命，从而降低电动汽车的总体成本。对于有效的BMS，电池模型必须基于(1)在线可用和(2)物理激励的模型参数。本项目的目的是将在线电缆的参数识别方法与产生物理理解的非参数方法结合起来。将开发的在线识别方法基于电池的分数模型，由于其特定的结构，允许进行物理解释，其参数通过新的识别程序进行识别。该方法基于“弛豫时间分布”的思想，已成功地应用于燃料电池。然而，需要用一种计算差动容量分布的新方法对其进行扩展。IRS在控制工程领域的能力和IWE在阻抗频谱领域的能力在这方面完美地互补。

A battery management system (BMS) is a key component of a modern electrified drive chain. In addition to safety functions an effective battery management system yields a perfect usage auf the battery itself and contributes significantly to a long lifetime of the vehicle’s battery and hence reduces the overall costs of an electric vehicle. For an effective BMS it is necessary that battery models base on (1) online available and (2) physically motivated parameters of the model. It is the aim of this project to couple parametric identification methods being online cable with non-parametric methods yielding a physical understanding. The online method identification to be developed bases on a fractional model of the battery, which allows a physical interpretation due to its specific structure, the parameters of which are identified by a new identification procedure. The procedure bases on the idea of “distribution of relaxation times” which has already been successfully applied to fuel cells. However it needs to be extended by a new method calculating the distribution of differential capacities. The competences of the IRS in the field of control engineering and the competences of the IWE in the field of impedance spectroscopy complement each other perfectly in this respect.