Low Temperature, Current Dependent Battery State Estimation Using Interacting Multiple Model Strategy

Lithium-ion battery State of Charge (SoC) estimation for Electric Vehicle (EV) applications must be robust and as accurate possible to maximize utilization ensure safe operation over a wide range operating conditions. SoC commonly utilizes filters such the Extended Kalman Filter (EKF) which rely on models, usually in form Equivalent Circuit Models (ECM). At low temperatures response current draw becomes increasingly non-linear, resulting amplified errors. In this study, dependent at temperature is proposed using an Interacting Multiple Model (IMM) filter with three ECMs covering C-rates. The IMM combined Smooth Variable Structure (SVSF) obtain estimates within error 2%.