How can we avoid traffic jams? design of on-demand traffic guidance systems

• Main Author: Xu, Muye.
• Other Authors: School of Electrical and Electronic Engineering
• Format: Final Year Project
• Language: English
• Published: 2012
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation::Control engineering
• Online Access: http://hdl.handle.net/10356/50095
• Institution: Nanyang Technological University

In most large cities, traffic congestion is quite common, especially at rush hours. Due to this reason, Intelligent Transportation Systems (ITS) are adopted with a growing popularity in those cities. ITS could collect on-site traffic data and information. Using these data, we could potentially develop a real-time traffic guidance system for individual drivers. By appropriately guiding drivers, traffic congestion may potentially be avoided or at least limited. In order to develop effective on-demand route guidance, we need to be able to track and predict the traffic flow in real-time. Indeed, if we can accurately predict how the traffic will evolve, we may be able to forecast potential traffic jams, and determine route guidance schemes to avoid them. In this research project, we have developed practical algorithms for tracking and predicting traffic flow in dynamic urban transportation networks in real-time. We developed algorithm at various stages, namely data acquisition/segmentation, traffic prediction and network optimization. At the initial phase, the mass data provided by various agencies will be treated in various ways respectively and the useful information is extracted from the segments. Prediction phase addresses the manner in which the traffic condition is predicted in advance of time and lastly, the network is optimized and optimum route will be provided. In all these phases, the raw data cannot be used directly, means it should be processed well to fulfill the basic requirement of data needed by each phases of the work. The major problem we noticed is missing data in the raw traffic data sets. It inspired us to conduct a extensive research in those missing data problem and the experiments and findings are explained in corresponding chapters.