Online SOH Estimation in Lithium-Ion Batteries Using Electrochemical Model and Unscented Particle Filter

The evaluation of the State of Health (SOH) is of paramount importance to ensure the safe and effective integration of lithium-ion batteries in electric vehicles (EVs) and energy storage systems. This paper presents an innovative methodology based on the electrochemical model and the Unscented Particle Filter (UPF) for SOH estimation in batteries. Specifically, the pseudo-two-dimensional (P2D) model is first employed to elucidate the internal battery mechanisms through a set of governing equations. Through a disassembly experiment, structural and performance parameters are determined using the P2D model and Genetic Algorithm, respectively. The stoichiometric ratio is utilized to calculate the battery capacity, forming the observation equation. Subsequently, a double exponential model is established as the state equation, creating a mathematical state-space representation to describe battery capacity degradation. The UPF method is then used to estimate the battery SOH by tracking the capacity state within the state-space representation. Comparative analysis with conventional data-driven methods, such as Gaussian Process Regression and Support Vector Regression, alongside the P2D model as a control group, demonstrates the superior accuracy of the UPF method based on the scheduled battery aging experiment.