Adaptive estimation-based hierarchical model predictive control methodology for battery active equalization topologies: Part II - equalizer control

In the first paper, a model predictive control (MPC) balancing strategy as the superior control is developed based on a universal equalization plant with balancing efficiency. Herein, two typical active equalization topologies, i.e., bidirectional adjacent, and bus-based, are considered to design the low-level controller, and the bidirectional flyback converter is selected as an individual cell equalizer (ICE). First, the extended Kalman filter (EKF) state-of-charge (SOC) estimation is presented to develop a unified ICE model. Then, an adaptive estimation-based control is proposed to track the optimal balancing reference current yielding from the superior control MPC formulation under the charge/discharge disturbance and parameters uncertainties. The current/voltage adaptive sliding-mode tracking control of the ICE is derived through the defined Lyapunov function, and the results are compared with a PI controller. Combined with the MPC balancing strategy, the EKF and adaptive-based controls are utilized to perform a balancing task for a 5-series battery cells module, which reduced the unbalanced SOC from 13% to 1%. To conclude, with the proposed superior control MPC strategy yielding optimal current targets, the low-level adaptive estimation-based controller for the ICE can effectively balance the inconsistent series-connected cells in typical active equalization systems by taking disturbance and uncertainty into consideration.