Battery State of Health Estimation Based on the Identification of Continuous-Time Impedance Models

The design of a state of health (SoH) observer for lithium-ion cells based on the identification of continuous-time impedance models is presented in this paper. For purely data-driven battery models, the use of continuous-time system identification methods enables a better interpretation of the model parameters, which is a key issue for estimating the SoH. Moreover, non-uniformly sampled data can be directly used for model construction. The performance of the proposed concepts is validated by means of real measurement data from accelerated ageing tests.