MATLAB Fall 2014 – Research Plan

• Group Name: Traffic Jammers Group
• Group participants names: Eric Hayoz, Janick Zwyssig
• Project Title: Modelling the phenomenon of congestion at Gotthard

General Introduction

Traffic is something that most of us face daily. Our group intends to model the traffic situation in front of the "Gotthard-Strassentunnel Nordportal" between Erstfeld and Göschenenen. Our goal is to predict an upcoming congestion and its length, by analysing the traffic situation at Erstfeld, Gurtnellen, Wassen and Göschenen.

The Model

Simulation of traffic flow from Erstfeld to Göschenen, including the 2-in-1-lane reduction in front of the “Gotthard-Strassentunnel”. Our model shall be based on the “Nagel-Schreckenberg-Modell”. The datasets that are available to us include traffic data of six full days.

The sets raised at Erstfeld, Gurtnellen, Wassen and Göschenen cover the full period, listed in 3 minute intervals:

• Vehicles

• Average speed

The following values were measured from tunnel entry northbound, at random times:

• Congestion in km
• Waiting time

We use datasets which were raised over three days to build and “train” our model, and then test it with the datasets of the other three days. We then compare the length of our simulated jam with the real one.

Fundamental Questions

How accurate can we predict a traffic congestion with its length at the “Gotthard-Strassentunnel” based on traffic data as listed above?

Expected Results

We believe that we can predict a traffic congestion and its length with an error of 30% or less.

References

[1] Andreas Schadschneider. Traffic flow modelling. http://www.thp.uni-koeln. de/˜as/Mypage/traffic.html. April 2000

[2] Torsten Held, Stefan Bittihn. Cellular automata for traffic simulation – NagelSchreckenberg model. March 2011

[3] K. Nagel and M. Schreckenberg. A cellular automaton model for freeway traffic. 1992

Research Methods

Nagel-Schreckenberg-Modell

Dataset provided by ASTRA

Other