A Comparative Study of the DEKF and DUKF for Battery SOC and SOH Estimation

The accurate estimation of the state of charge (SOC) and state of health (SOH) is essential for the safety and reliability of electric vehicle batteries. Conventional single-state Kalman filters are prone to parameter drift caused by cell aging, which leads to persistent SOC estimation errors. This study compares two dual-estimator methods, the Dual Extended Kalman Filter (DEKF) and the Dual Unscented Kalman Filter (DUKF), for simultaneous SOC and SOH estimation using a second-order equivalent-circuit model. The process and measurement covariance matrices were tuned through a structured optimization procedure to ensure consistent performance under different drive cycles and initialization errors. To mitigate the weak voltage sensitivity to capacity, synthetic SOC–capacity coupling was introduced to enhance SOH observability and accelerate convergence. Simulations conducted under the Urban Dynamometer Driving Schedule (UDDS) and a real-world CLUST7 profile demonstrated SOC root-mean-square errors near 2% for both filters. The DUKF achieved faster and smoother convergence than the DEKF but required roughly fivefold higher computational cost. These findings provide quantitative evidence supporting dual Kalman filtering as an effective framework for accurate and robust SOC/SOH estimation in production battery management systems.