Abstract

Projecte realitzat en el marc d’un programa de mobilitat amb Eidgenössische Technische Hochschule Zürich The objective of this Thesis is to understand the situation of a catalan highway with VSL strategies through empirical data of one day. It will be done using standard traffic tools to make the congested and uncongested states analysis. The Thesis will involve: visual identification of congested areas, bottleneck detection and diagnostic, identification of fundamental diagrams, lane density and speed distribution and comparison between speed limits and measured speeds. In addition, it will give recommendations and establish the basis for a future set of experiments to investigate VSL strategies in deep. The main findings are the detection of five bottlenecks, many lane changes near the ramps and the change of occupancy distribution between lanes in congested periods. After a careful and detailed analysis of the data we can conclude that VSL is not as effective as it should be. That is caused by the algorithm it uses and the fact that drivers disobey speed limits.

Abstract

Traffic management is being more important than ever, especially in overcrowded big cities with over-pollution problems and with new unprecedented mobility changes. In this scenario, road-traffic prediction plays a key role within Intelligent Transportation Systems, allowing traffic managers to be able to anticipate and take the proper decisions. This paper aims to analyze the situation in a commercial real-time prediction system with its current problems and limitations. We analyze issues related to the use of spatiotemporal information to reconstruct the traffic state. The analysis unveils the trade-off between simple parsimonious models and more complex models. Finally, we propose an enriched machine learning framework, Adarules, for the traffic state prediction in real-time facing the problem as continuously incoming data streams with all the commonly occurring problems in such volatile scenario, namely changes in the network infrastructure and demand, new detection stations or failure ones, among others. The framework is also able to infer automatically the most relevant features to our end-task, including the relationships within the road network, which we call as “structure learning”. Although the intention with the proposed framework is to evolve and grow with new incoming big data, however there is no limitation in starting to use it without any prior knowledge as it can starts learning the structure and parameters automatically from data. (Part of special issue: 20th EURO Working Group on Transportation Meeting, EWGT 2017, 4-6 September 2017, Budapest, Hungary) Peer Reviewed