A Flexible Energy Management System for Solar Powered Electric-Bus Charging Stations

This paper presents a flexible energy management system to manage an electric bus charging station incorporated with solar power, energy storage system and the main grid. To account for solar power uncertainties, random forest regression is employed to forecast solar irradiation. Robust scheduling frameworks are proposed to optimize charging schedules while minimizing operational costs, incorporating day-ahead pricing (DAP), grid-side dynamics, solar generation, and energy storage system (ESS). Two case studies demonstrate the scalability of the proposed energy management system for large-scale real-world electric bus fleets. Moreover, the energy management system effectively manages varying state-of-charge (SoC) levels, allocate chargers based on parking schedules, and utilize bidirectional power transfer for grid support and ancillary services.