Optimal design for urban transit systems

Abstract

The adverse impacts of fossil fuels on the built and natural environment call for the adoption of a more sustainable urban transportation model. To that end, the design of urban transit systems must take into account environmental impacts associated with their operation, while also aim at increasing ridership and ensuring financial viability for transit operators. Recent advances in communication, data collection and wireless power transfer technologies shape a new reality for public transportation, creating the need for new decision support systems. In this context, the scope of this dissertation is to develop a modeling framework based on data science and artificial intelligence algorithms for planning and reshaping public transport systems exploiting recent advances in electric energy transmission technologies and Intelligent Transportation Systems. Specific objectives of the dissertation include: a.) the development of a comprehensive model for the design of a transit network, operated exclusively by an electric bus fleet under different charging infrastructure configurations, b.) the development of a framework for improving bus service planning, focusing on the problem of bus bunching, and c.) the development of appropriate tools and methods for exploiting Automatic Vehicle Location (AVL) data. The proposed framework is applied to established benchmark networks and data from the Athens Public Transport System, respectively.