Artificial intelligence based robust nonlinear controllers optimized by improved gray wolf optimization algorithm for plug-in hybrid electric vehicles in grid to vehicle applications

Abstract

The paper introduces a control strategy for a plug-in hybrid electric vehicle (PHEV) system, integrating a photovoltaic-based RES with battery and supercapacitor-based ESS. The objective is to efficiently utilize the PV system using a neural network to determine its maximum power points. To ensure robust control, the paper designs both a robust nonlinear higher-order super twisting sliding mode controller and an integral terminal sliding mode controller. One notable aspect of the proposed system is the implementation of a current-fed con verter for the PV and supercapacitor. Unlike SCR-based converters, current-fed con verters prevent device failure and core saturation, making them more reliable. This characteristic contributes to the overall robustness of the system. The integration of the neural network, along with the designed controllers and current-fed converter, enhances the efficiency and reliability of the plug-in hybrid electric vehicle system. This research holds promise for developing advanced and resilient energy management solutions for sustainable transportation systems in the future. Additionally, a power bi-directional converter is employed for the battery in the EV system. This converter enables the battery to both charge and discharge power as required by the system which facilitates the energy flow between the battery and the rest of the system, allowing efficient energy management and utilization within the PHEVs. The integration of ANN for PV with the above mentioned converters and controllers along with improved-gray wolf optimiza tion represents the novelty of this work. The paper confirms global asymptotic stability of the control method using Lyapunov stability analysis. It further demonstrates the proposed system’s behavior through a hardware-in-the-loop experiment.