Electrified transit
Bus and school-bus electrification
Fleet sizing, routing, timetabling and charging for battery-electric buses, from mixed-integer models and metaheuristics to multi-agent deep reinforcement learning that reacts to delays, demand and time-of-use prices in real time. Includes vehicle-to-grid revenue and battery sizing.
- MILP
- VNS metaheuristics
- PPO / MAPPO
- V2G
- Gurobi
- GAMS
Projects
- Multi-line electric bus scheduling and charging with deep RLUniversity at Buffalo (dissertation) · University at Buffalo · active
- Route Optimization, On-route Charging and V2G Integration (ASPIRE)NSF Engineering Research Center · Utah State University
- Increasing Affordability, Energy Efficiency, and Ridership of Transit Bus Systems through Large-Scale ElectrificationUS Department of Energy · Utah State University
- Alternative Fuel Vehicles in the UDOT Maintenance FleetUtah Department of Transportation · Utah State University