Identifiability analysis of an electrochemical model of li-ion battery
Proceedings of the American Control Conference
The parameters of the electrochemical model of the battery bear important information about the condition of the battery. Therefore, their estimation not only improves the accuracy of state of charge estimation but can also play a substantial role in battery health monitoring and fault diagnostics. However, due to the lack of measurable system responses in the battery system, and too many model parameters, one question arises that whether the unknown model parameters can be uniquely estimated based on the measured output of the system. Identifiability analysis is the first step to address this important question. This paper applies sensitivity-based analysis to the battery system to study the identifiability of the battery parameters. The impact of different parameters such as input profile and step size in differentiation calculations are also studied on the identifiability of the parameters.
Samadi, M. Foad and Saif, Mehrdad. (2016). Identifiability analysis of an electrochemical model of li-ion battery. Proceedings of the American Control Conference, 2016-July, 3107-3112.