State estimation of solid-state batteries via a distributed parameter electrochemical model

Solid-state Lithium-ion (Li-ion) batteries have emerged to be one of the leading candidates for future energy storage systems owing to their exceptional high specific energy and safety benefits. Accurate electrochemical information enables safe and optimal energy management. This paper presents a provably convergent partial differential equation (PDE) based state estimation framework utilizing an electrochemical solid-state battery model, by leveraging voltage and current measurements only. Notably, this work promptly considers various crucial nonlinear effects and electrolyte dynamics. In particular, the nonlinear generation/recombination process of Li-ions in the solid electrolyte is pivotal in predicting electrolyte Li-ion distributions. We develop a robust state observer by exploiting concepts from semilinear PDE backstepping technique without relying on prior model discretization and reduction. The estimation error stability and input-to-state stability are also rigorously demonstrated.