Recursive Estimation of Battery Pack Parameters in Electric Vehicles

Battery management system (BMS) is an essential component for ensuring safety, reliability and battery life in Electric Vehicle (EV). Estimation of State-of-Charge (SOC) is one of the key functions of BMS in EV. This investigation combines the digital-twin approach and parameter estimation methods to estimate the battery pack's SOC online. The battery's digital-twin is an equivalent circuit model on which simulations are conducted by using current measurements. The model output is used as input to the parameter estimation algorithm which estimates the battery parameters. We provide a novel approach which combines off-line parameter estimation based on least-squares with recursive least squares to update the battery parameters when the vehicle is operating. This provides a way to monitor SoC and provide diagnostics based on current conditions. The proposed approach is illustrated using battery parameters obtained from a EV battery pack. Our results illustrates the efficacy of the approach to provide accurate estimates of parameter and SoC.