Simulation and optimization of innovative urban transportation systems

Sammanfattning: The ongoing trends of urbanization and e-commerce continuously challenge the existing urban transportation systems. A steadily growing number of people traveling within urban areas, results in more trips taken with public transportation systems. Additionally, the constantly increasing number of urban logistic operations leads to more commercial vehicles in cities. These ongoing trends and the need for more sustainable operations require the design of robust and efficient transportation systems which additionally provide a high level of service for their users. In recent years, two innovative approaches have been proposed to overcome these challenges. That is, first, the use of autonomous buses as a replacement, or an addition to existing public transportation systems, and second, the consideration of consolidating multiple types of demand (i.e. passenger and freight) when planning and designing transportation systems. In this thesis, both approaches are studied and their impact on urban transportation systems is evaluated. This is achieved by developing novel simulation-based optimization models that consider technology-specific cost structures and capture the changed mode of operation for different vehicle technologies.In Papers I and II the deployment of autonomous buses on fixed-line public transportation networks is investigated. Changes in service frequency, vehicle capacity, and metrics corresponding to the level of service for public transportation users due to new vehicle technology are investigated. Furthermore, Paper I explores the transition from conventional public transportation systems to systems operated by autonomous buses, while Paper II investigates the changes in network design due to autonomous bus operations. The developed models are applied to case studies in Kista, Sweden, and Barkarby, Sweden. Two key results can be identified in these studies. First, autonomous bus deployment leads to an increase in service frequency, while waiting time for passengers can be reduced. Second, more passengers are attracted to autonomous bus lines by reducing the access walking distances and increased level-of-service. On more complex networks these trends are amplified. In each of Papers III and IV, a novel pickup and delivery model is proposed. The models consider vehicle concepts which allow for the consolidated transport of multiple demand types. In Paper III the vehicles can serve different types of demand by exchanging purpose-specific modules at dedicated service depots, while in Paper IV individual demand-specific vehicles can form platoons with modular length and varying configuration. The results of the extensive scenario studies and parameter analysis show that for multi-purpose vehicle operations (Paper III) the total costs can be reduced by an average of 13% and for platoon operations (Paper IV) the total costs are reduced by over 48%. In both models, the cost savings stem mainly from a reduction in fleet size, total vehicle trip duration, and the total distance traveled.