Optimizing Urban Mobility in Multi-Mode Transportation Systems

Abstract

The transportation system needs innovative schemes and applications to facilitate mobility in cities. These schemes should be user-friendly, easy, enjoyable, and convenient according to citizens' constraints. Three dynamic multi-mode transportation models are developed in which the passenger, for his/her trip, can use one transportation mode or a combination of transportation forms such as a car, a bus, or a bicycle. The first model is based on the particle swarm optimization (PSO) algorithm and the genetic algorithm (GA), and it is called dynamic mobility traffic (DMT). The second and third models are both based on the Stackelberg game theory with the PSO and GA, but one is a centralized infrastructure called the game theory multi-mode transport (GT-MMT) model, and the second is a community-based infrastructure called the decentralized game-theoretic (DGT) model. All three models are implemented through a realistic scenario in a specific city using OMNET++, OpenStreetMap, Inet, and Veins software tools. We proposed a new Dynamic Mobility Traffic(DMT) scheme that combines public buses and car ride-sharing. The main objective is to improve transportation by maximizing the riders' satisfaction based on real-time data exchange between the regional manager of the transportation modes, the public buses, the car ride-sharing mode, and the riders. The GT-MMT and DGT models, which involve multiple modes of transportation, allow the user to be an active prosumer who can travel in the city using public and private forms and make decisions about trip costs. In the two models, the passenger is the leader, and the rest of the modes are followers. The utility functions for each player are different depending on their goals. While the passenger aims to reach the destination in the shortest possible time and for the lowest price, the bus and the car try to have the seats vacant as often as possible, and the bicycle tries to increase the availability for the passenger. We propose these two models to manage passenger needs, public bus interests, car ride-sharing, and bicycle constraints. The analytical and simulation results prove the effectiveness of the proposed schemes. The effectiveness of the decentralized game-theoretic transportation model appears more clearly when compared with the centralized multi mode dynamic approach in [1], as it gives much better optimization results.