Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{204521,
        author = {S. Arya and Dr. J Arul Linsely},
        title = {AI-Enabled Power Management in Multi-Port Converter for Renewable Integrated EV Charging Systems},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {no},
        pages = {203-208},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=204521},
        abstract = {The rapid proliferation of electric vehicles (EVs), coupled with the large-scale deployment of intermittent renewable energy sources, has exposed a critical gap in existing charging infrastructure: the inability to intelligently coordinate multiple energy inputs under real-time variability constraints. This paper presents an Adaptive Hybrid AI-Controlled Multi-Port Power Converter (MPPC) architecture designed specifically for grid-integrated, solar-powered EV charging stations. The proposed system incorporates a four-port silicon carbide (SiC)-based converter topology interfacing a photovoltaic (PV) array, battery storage, utility grid, and EV output through a shared 400 V DC bus. At the core of the control framework lies a Long Short-Term Memory (LSTM) neural network that forecasts short-term solar irradiance and EV demand with a mean absolute error of 6.3%, enabling anticipatory rather than purely reactive power dispatch. The LSTM forecasts are fed into a Model Predictive Controller (MPC) that solves a cost-minimisation scheduling problem every 10 seconds, while a Deep Q-Network (DQN)-based reinforcement learning agent handles sub-second port switching decisions in real time. A dedicated AI fault detection module isolates port faults within 2 ms, ensuring continuous EV charging without service interruption. Simulation results obtained in MATLAB/Simulink over a full 24-hour cycle demonstrate a peak efficiency of 96.2%, grid total harmonic distortion (THD) of 1.8%, power factor of 0.99, and a 41% reduction in operational cost compared to conventional grid-only charging systems. These outcomes confirm that tightly coupling predictive LSTM-based forecasting with multi-objective optimisation and reinforcement learning yields a substantially more capable and resilient EV charging solution than any of these strategies applied in isolation.},
        keywords = {Multi-port converter, EV charging, renewable energy, artificial intelligence, predictive energy management, grid integration, LSTM, MPC.},
        month = {June},
        }