Hybrid Control Scheme for More Efficient Charging of Two Electric Vehicle Batteries Simultaneously

The paper proposes a hybrid control scheme for more efficient charging of two electric vehicles (EV) simultaneously. In this work, a triple active bridge (TAB) is controlled using a hybrid scheme comprising of proportional-integral (PI) and deep reinforcement learning (DRL) algorithm to improve the charging efficiency of two electric vehicle batteries. The DRL part uses the Deep Q-Network (DQN) algorithm to balance the effects of varying parameters between the two EV batteries. The control model is trained following the Markov Decision Process (MDP). Battery parameters namely voltage, current, temperature, state of charge (SoC) and state of health (SoH) form the state-space (S
t
). Phase shift ratios, duty cycles and the switching frequency associated with the active bridges of the converter form the action space (a
t
). Charging efficiency and voltage stability form the reward function (r
t
). Simulation results indicate a slight increase in power efficiency of battery charging using the proposed method. The study contributes to the advancement of EV charging technologies, emphasizing cost savings, space savings, efficiency gains, and showing versatility with multiport charging.