Battery degradation diagnosis with field data, impedance-based modeling and artificial intelligence
Weihan Li, Jue Chen, Katharina Lilith Quade, Daniel Luder, Jingyu Gong, Dirk Uwe Sauer, Energy Storage Materials, Vol. 53, December 2022.
By collecting battery data from the field and building up the battery digital twin in the cloud, the degradation of batteries can be monitored online on the electrode level and the information regarding the degradation modes can be extracted from the data. Here, we present a degradation diagnosis framework for lithium-ion batteries by integrating field data, impedance-based modeling, and artificial intelligence, revolutionizing the degradation identification with accurate and robust estimation of both capacity and power fade together with degradation mode analysis. By integrating an impedance-based model and an open-circuit voltage reconstruction model, the hybrid model consists of parameters representing the change of impedance in a wide frequency domain and the change of open-circuit voltage during degradation. Based on the field data with low and high dynamics, the data-driven parameter identification method using a multi-step cuckoo search algorithm considering parameter sensitivity differences shows high accuracy and robustness in aging parameter estimation and degradation mode identification even under sensor noise. Furthermore, the data requirement for the battery digital twin in the sense of sampling rate was investigated considering degradation identification accuracy, computational cost, and data storage cost. This work highlights the opportunity in online electrode-level degradation diagnosis in the field through battery modeling and artificial intelligence.
Free download: ScienceDirect