Fleet and charging infrastructure decisions for fast-charging city electric bus service. (November 2021)

• Record Type: Journal Article
• Authors: Guschinsky, Nikolai; Kovalyov, Mikhail Y.; Rozin, Boris; Brauner, Nadia
• Is Part Of: Computers & operations research. Volume 135(2021)
• Publication Date: 2021-11
• Subjects: Electric bus -- Fast-charging -- Optimal planning -- Randomized heuristic -- Particle Swarm Optimization -- Scheduling; Operations research -- Periodicals; Electronic digital computers -- Periodicals; 004.05
• DOI: 10.1016/j.cor.2021.105449
• Languages: English
• ISSNs: 0305-0548

Abstract

Abstract: Decision aspects concerning the introduction of fast-charging city electric buses are studied in this paper. The main studied problem consists of determining a fleet of electric buses and their charging infrastructure such that a social-ecological value is maximized. In the most representative time period, all the electric buses should be available to drive, the required inter-bus interval should be maintained, the output power of any charging station and transformer should not be exceeded, and the total capital cost and the total annual operating, depreciation and energy cost should not exceed certain maximum thresholds. The total passenger demand satisfied by the electric buses can be considered as the value to be maximized. A secondary problem consists of finding a passenger load balanced schedule of the vehicles on the same route. Mathematical models for these two problems are proposed. A randomized heuristic algorithm combined with the Particle Swarm Optimization is developed for the main problem, and a known polynomial time algorithm is adapted for the secondary problem. A case study for the city of Minsk (Belarus) and computer experiments with random instances are provided. The proposed approach delivered solutions with values deviating at most 12% on average and 24% in the worst case from the upper bounds obtained as optima of a relaxed problem. Highlights: Primary problem is to determine fleet of electric buses and charging infrastructure. Secondary problem is to find passenger-load balanced schedule of vehicles. Randomized heuristic algorithm combined with PSO is developed for primary problem. Polynomial time algorithm is developed for secondary problem. Case study and computer experiments with random instances are provided.