Decision Making on Vehicles Relocations in Free-Flotation Car-Sharing Schemes
摘要
This paper addresses the Vehicle Relocation Problem (VReP) in free-floating car-sharing systems, focusing on real-time strategies to address demand-supply imbalances. We propose a solution combining AI-based demand and vehicle availability prediction with decision-making algorithms for staff-assisted relocations. Using data from Kraków, Poland, the service area is divided into zones to enable discrete optimization. Experiments with multiple algorithms show significant improvements in fleet management over non-optimized approaches. The results demonstrate the value of predictive models in improving the efficiency and competitiveness of car-sharing systems.