基于优化ANFIS的鲁棒非线性控制在太阳能离网电动汽车充电站中的应用

This paper attempts to improve the performance of an off-grid electric vehicle charging station powered by photovoltaic (PV) panels and batteries. To ensure optimal performance of the PV panels, maximum power point tracking (MPPT) is implemented using an adaptive neuro-fuzzy inference system. A robust control technique is designed for regulating the currents and voltages within the system. Conventional sliding mode controllers (SMCs) are prone to chattering, and while the inclusion of a super-twisting component mitigates this issue, it introduces windup problems. To address these challenges, we propose a conditioned-based super-twisting SMC (CST-SMC), which effectively resolves the windup issue. The controller gains are optimized using the grey wolf optimization algorithm, and global stability is demonstrated through Lyapunov stability analysis. The proposed system is simulated in MATLAB and experimentally validated using a Delfino F28379D-based hardware-in-the-loop setup. Numerical results show that the proposed CST-SMC has a faster rise time (0.0564 sec), less overshoot (0.00389), and a shorter settling time (0.4256 sec) compared to the conventional ST-SMC, which has a slower rise time (0.137 sec), greater overshoot (0.38095), and a longer settling time (3.8071 sec). Thus, the system’s dynamic response is improved. Furthermore, the proposed CST-SMC controller provides smoother regulation during the constant current and constant voltage stages, unlike the ST-SMC, which exhibits chattering. The improved performance makes the CST-SMC more reliable for EV charging.