An Adaptive Observer Design for Real-Time Parameter Estimation in Lithium-Ion Batteries

This paper addresses the problem of parameter identification in an electrochemical model of a Lithium-ion battery. The starting point is the development of an extended model based on the absolute nodal coordinate formulation (ANCF-e), which provides high accuracy, yet low computational complexity. Using such a model, this paper proposes an adaptive observer to carry out real-time parameter estimation. The complex spatiotemporal relation between inputs, states, and outputs, and the combined presence of multiple nonlinearities related to the open circuit, electrolyte, solid, and overpotentials prevents the application of standard tools of parameter estimation and adaptive observers. These challenges are overcome by breaking the ANCF-e model into four subsystems which consist of key dynamic relationships between the molar flux and either measurable quantities or states of the solid and electrolyte dynamics, as well as a nonlinear subsystem that relates inputs, outputs, and states on one hand, and potentials on the other. Adaptive observers are proposed to identify parameters of each subsystem and are validated using simulation studies. Significant extensions of currently available adaptive observers are proposed for this purpose.