Minimizing Energy Use of Mixed-Fleet Public Transit for Fixed-Route Service

Published:

Recommended citation: Sivagnanam, A., Ayman, A., Wilbur, M., Pugliese, P., Dubey, A., & Laszka, A. (2021, May). Minimizing energy use of mixed-fleet public transit for fixed-route service. In Proceedings of the AAAI Conference on Artificial Intelligence (Vol. 35, No. 17, pp. 14930-14938)

📌 Key Contributions

  • Formulated a mathematical model to optimize energy consumption for public transit agencies operating mixed fleets of Electric Vehicles (EVs) and Internal Combustion Engine Vehicles (ICEVs)
  • Transformed the model into an Integer Programming (IP) formulation to obtain exact optimal solutions for small to medium-sized problem instances
  • Developed a scalable solution approach combining heuristics and metaheuristics (e.g., local search, genetic algorithms) to efficiently solve large-scale instances in polynomial time
  • Reduced annual energy costs by $140K for the Chattanooga public transit agency through the proposed approach

📝 Publications

The work is published in: Proceedings of the AAAI Conference on Artificial Intelligence (AAAI21) “Minimizing Energy Use of Mixed-Fleet Public Transit for Fixed-Route Service” [AAAI-21].

Other relavent publication: ACM Transactions on Internet Technology (TOIT) “Data-driven prediction and optimization of energy use for transit fleets of electric and ICE vehicles” [ACM]


💻 Code & Data

The implementation and sample data can be found in the following [Code & Data].