AI-Driven IoT-Enabled Smart Charging and Energy Management Framework for Electric Vehicles in Vehicle-to-Grid Networks

Abstract

The rapid growth of electric vehicle (EV) use means that the requirement for intelligent charging systems that respond to renewable generation, electricity tariffs, grid loading and battery constraints is even greater. This paper proposes the smart charging and energy management framework for electric vehicles under a vehicle-to-grid environment with artificial intelligence-enabled Internet of Things integration. The framework integrates the features of real-time sensing, cloud-assisted data exchange, photovoltaic generation, bidirectional charging, battery state-of-charge supervision and adaptive power control. The core of this architecture is a scheduling layer that determines whether connected vehicles should charge, discharge, or remain idle based on continuously communicated information on vehicle availability, energy demand, departure time, and battery level. The goals are to minimize energy expenses from the grid, restrict peak power demand from the grid, maximize the photovoltaic self-consumption ratio, maintain the DC-link steady-state, and avoid severe battery cycling. Simulation results suggest that coordinated operation can reduce peak grid demand from 312.18 kW to 208.38 kW and daily grid energy costs from INR 24,357.78 to INR 20,320.49. Photovoltaic utilization increases from 79.98% to 98.40%, and direct current-link voltage root-mean-square error decreases from 8.77 V to 1.89 V, while the final average battery state of charge achieves a value of 75.35% without violating the previously specified limits of operation (20-90%). These results demonstrate that intelligence, communication and bidirectional power exchange integrations enable cost-effective, stable and grid-friendly EV charging under realistic daily operating conditions.