Abstract

Novo Mesto is a small town in the Slovenian countryside which is undergoing a new project of road safety. New traffic signals are to be placed in bus stops to give drivers awareness of children standing waiting for the bus. Our mission is to use an eye tracker (the Dikablis device by Ergoneers) to study the reaction of three test drivers against these new signals; if they do detect them or not and if they modify their driving behavior once noticed them. The methodology is to do a first test before the signal placements and another one after, and afterwards compare both situations. Results are satisfying, with high values of reliability, and make us think the new signals will be useful and will contribute to make the road a safer place for children. Outgoing

Abstract

Thesis (MTransportSysEngineering)--University of South Australia, 2005. Transport systems can be modelled and analysed with two techniques namely analytical and simulation. Until recently, aaSIDRA was the leader in the traffic modelling and analysis software based on analytical theory at micro scale. However, it cannot be used for more than one intersection. Meanwhile, the PARAMICS, AIMSUN NG and VISSIM are micro simulation software which are widely used in modelling and analysing transport systems. Further, CUBE Dynasim is also micro simulation software. It is created by Dynalogic in France and integrated in CUBE by Citilabs, UK. However, this software is relatively new compared to PARAMICS, AIMSUN NG or VISSIM. It is able to model multi nodes and multi modals transport system. The version of software used in this research is CUBE Dynasim 1.4.9 released in 2004 aaSIDRA 2.0 version 2.0.0.205.

Abstract

Excessive traffic in our urban environments has detrimental effects on our health, economy and standard of living. To mitigate this problem, an adaptive traffic lights signalling scheme is developed and tested in this paper. This scheme is based on a state space representation of traffic dynamics, controlled via a dynamic programme. To minimise implementation costs, only one loop detector is assumed at each link. The comparative advantages of the proposed system over optimal fixed time control are highlighted through an example. Results will demonstrate the flexibility of the system when applied to different junctions. Monte Carlo runs of the developed scheme highlight the consistency and repeatability of these results. peer-reviewed