Noise-Immune Model Identification and State of Charge Estimation for Lithium-Ion Battery Using Bilinear Parameterization

Accurate estimation of state of charge (SOC) is critical to the safe and efficient utilization of battery system. Model-based SOC observers have been widely used due to their high accuracy and robustness, but they rely on a well parameterized battery model. This paper scrutinizes the effect of measurement noises on model parameter identification and SOC estimation. A novel parameterization method combining instrumental variable (IV) estimation and bilinear principle is proposed to compensate for the noise-induced biases of model identification. Specifically, the IV estimator is used to reformulate an overdetermined system so as to allow co-estimating the model parameters and noise variances. The co-estimation problem is then decoupled into two linear sub-problems which are solved efficiently by a two-stage least squares algorithm in a recursive manner. The parameterization method is further combined with a Luenberger observer to estimate the SOC in real time. Simulations and experiments are performed to validate the proposed method. Results reveal that the proposed method is superior to existing method in terms of the immunity to noise corruption.