Adaptive Multi-Stage Charging, SiC MOSFET Inverter Design and V2G Optimal Scheduling for Electric Vehicle Grid Integration in Smart Urban Microgrids

India's electric vehicle market, having crossed 1.5 million annual unit sales in FY2024 driven by two-wheeler electrification and fleet operator adoption of electric buses and three-wheelers, is approaching the inflection point at which aggregated EV battery capacity becomes a significant distributed energy resource that smart grid operators can dispatch to meet peak demand, provide frequency regulation services, and reduce grid infrastructure investment. Vehicle-to-Grid (V2G) technology, which enables bidirectional power flow between EV batteries and the distribution grid, can unlock this potential — but requires co-optimisation of the battery charging protocol (to preserve cycle life under bidirectional cycling), the power electronics inverter design (to achieve high efficiency across the wide power range of V2G operation), and the grid-side scheduling algorithm (to optimise charge-discharge timing against time-of-use tariffs and grid ancillary service prices). This paper addresses all three dimensions in an integrated framework. An adaptive multi-stage charging algorithm that varies current profile based on real-time State-of-Health (SoH) estimation is compared against standard CC-CV and multi-stage fast charging on charge time, capacity retention after 1,000 cycles, and cell temperature rise. A 7.4 kW SiC MOSFET-based bidirectional on-board charger is designed and compared against an IGBT-based reference on switching losses, THD, and inverter efficiency. A Mixed-Integer Linear Programme (MILP) for V2G scheduling optimises charge-discharge decisions over a 24-hour horizon against the MSEDCL time-of-use tariff structure and real-time frequency regulation signals. Hardware-in-the-Loop (HIL) validation on a dSPACE MicroLabBox platform confirms the scheduling algorithm's real-time implementability.