Design and Implementation of a UKF-based SOC Estimator for LiMnO2 Batteries Used on Electric Vehicles

To ensure safe and reliable battery operations, an accurate battery state of charge (SOC) estimation is critical for the battery systems used in electric vehicles and hybrid electric vehicles because of the arduous operation conditions. This paper presents a SOC estimator designed based on the unscented Kalman filter (UKF), which is very popular in the state estimation in non-linear systems. The dynamic characteristics of the battery are modeled with an equivalent circuit, which is composed of two capacitors, three resistors and a voltage source to simulate the equilibrium open circuit voltage (OCV). To relieve the computation requirement of the original UKF, an efficient implementation using a Cholesky factorization is investigated, and thereby a SR-UKF based SOC estimator is proposed. Experiment results shows that the model proposed can track the dynamic behavior of the battery very well and the UKF-based SOC estimator has a good performance in the state estimation, and a comparison with EKFbased estimator also shows that a better accuracy can be got by the proposed UKFbased estimator. Streszczenie. Artykuł prezentuje system kontroli SOC (state of charge – stan naładowania) baterii używanych w pojazdach elektrycznych. System bazuje na filtrze Kalmana typu UKF. Własności dynamiczne baterii modelowane są przy pomocy odpowiedniego obwodu elektrycznego zastępczego. System może śledzić właściwości dynamiczne baterii i badać jej stan naładowania. (Projekt i zastosowanie estymatora stanu naładowania baterii LiMnO2 używanych w pojazdach elektrycznych)