Observability and Estimability for State and Parameter Estimation in Li-Ion Battery Systems Under Different Parameter Configurations

This paper addresses the state and parameter estimation problem in lithium-ion batteries, focusing on an equivalent-circuit model (ECM). The ECM, consisting of a resistor and resistor-capacitor (RC) pairs to capture Ohmic and polarization behaviors, is widely used in battery management system design. While the ECM structure is well-studied, its integration into an estimator allows flexibility in selecting parameter sets to represent its characteristics. Parameters such as resistance, capacitance, time constant, and decay factor convey information about the battery system. However, the choice of parameters significantly influences estimation convergence, accuracy, and robustness. This study examines various parameter set selections, evaluates their performance through observability and estimability analyses, and implements state-of-charge (SOC) and parameter estimation using extended Kalman filters. The SOC and parameter estimation results are consistent with the theoretical observability and estimability analyses.