Knowledge Discovery for Sustainable Urban Mobility

Due to the rapid growth of urban areas, sustainable urbanization is an inevitable task for city planners to address major challenges in resource management across different sectors. Sustainable approaches of energy production, distribution, and consumption must take the place of traditional methods to reduce the negative impacts of urbanization such as global warming and fast consumption of fossil fuels. In order to enable the transition of cities to sustainable ones, we need to have a precise understanding of the city dynamics. The prevalence of big data has highlighted the importance of data-driven analysis on different parts of the city including human movement, physical infrastructure, and economic activities.

Sustainable urban mobility (SUM) is the problem domain that addresses the sustainability issues in urban areas with respect to city dynamics and people movements in the city. Hence, to realize an integrated solution for SUM, we need to study the problems that lie at the intersection of energy systems and mobility. For instance, electric vehicle invention is a promising shift toward smart cities, however, the impact of high adoption of electric vehicles on different units such as electricity grid should be precisely addressed.

In this dissertation, we use data analytics methods in order to tackle major issues in SUM. We focus on mobility and energy issues of SUM by characterizing transportation networks and energy networks. Data-driven methods are proposed to characterize the energy systems as well as the city dynamics. Moreover, we propose anomaly detection algorithms for control and management purposes in smart grids and in cities. In terms of applications, we specifically investigate the use of electrical vehicles for personal use and also for public transportation (i.e. electric taxis). We provide a data-driven framework to propose optimal locations for charging and storage installation for electric vehicles. Furthermore, adoption of electric taxi fleet in dense urban areas is investigated using multiple data sources.