An Ultra-Local Model-Based Control Method With the Bus Voltage Supervisor for Hybrid Energy Storage System in Electric Vehicles

Electric vehicles (EVs) with hybrid energy storage systems (HESSs) are getting popular as HESS can improve the battery's lifetime and thus reduce maintenance costs. In this article, an ultra-local model (ULM)-based method with the bus voltage supervisor (BVS) is proposed for the HESS in EVs with stable and quick dynamic performance. In the method, an integral sliding mode observer (ISMO) is designed for the disturbance term in ULM, and ULM generates the overall HESS current demand faster and more precisely than the conventional methods, which is then split into the transient and average components for driving the battery and supercapacitor (SC). Moreover, BVS generates extra current demand for SC during the transient process, realizing rapid recovery of bus voltage and decreasing the bus voltage spike. BVS will leave no degradation in steady-state performance with magnifying coefficients greater than one. Series simulations have been conducted to verify the dynamic performance and the effect of different curve functions on BVS. Consequently, a hardware-based experiment is conducted and the experimental results demonstrate the effectiveness of the proposed method with the 65%-69% faster dynamic performance and the 36%-41% lower voltage spike than the conventional PI methods.