Andreas Pottmeier

A Microscopic Simulator for Freeway Traffic

In this contribution a micro-simulator for freeway traffic which is coupled to a stream of on-line data stemming from inductive loops is presented. In order to provide network-wide information about the current traffic state a cellular automaton traffic flow model is combined with the measured data. The framework is applied to the freeway network of North Rhine-Westphalia (NRW), where data from about 3,500 inductive loops are provided on-line minute by minute. Special attention is paid to the technical aspects of the simulation like the network structure. Results are presented and discussed.

Simulation Framework for the Autobahn Traffic in North Rhine-Westphalia

In the last decade there has been a continuous progress in the development of cellular automata models of vehicular traffic. The most recent models are able to reproduce free flow, spontaneous jam formation, synchronized traffic, as well as meta-stability. However, these models have been developed and tested on topologically simple road networks and the translation to large and topologically complex real road networks is non-trivial. In this paper we describe the cellular automaton model we use to simulate the traffic on the autobahn network in North Rhine-Westphalia and discuss some of the challenges that arise when using this model on such a huge and topologically complex network.

Mechanical Restriction Versus Human Overreaction: Accident Avoidance and Two-Lane Traffic Simulations

Lee et.al. [1] have proposed a cellular automaton model that emphasizes the conflict between human overreaction and limited mechanical capabilities as the origin of congested traffic states. The limited acceleration and deceleration capabilities lead to a rather different approach to realize realistic traffic modeling. But the original model lacks the robustness and usability for more complicated and flexible simulations. In order to allow an extension of the model to two-lane traffic a modification of the original single-lane model is presented that ensures the absence of any collisions in the model dynamics.

Simulation of the Autobahn Traffic in North Rhine-Westphalia

The amount of vehicular traffic has reached the capacity of road networks in many densely populated regions world-wide. Especially in North Rhine-Westphalia, situated in the western part of Germany, growing traffic demand gives rise to more and more congestion on the autobahn network. Therefore, the need for intelligent information systems has become increasingly important. Here we present a combination of real time traffic data analysis and microscopic traffic simulations as the basis of an online-tool that provides full information of the traffic state on the complete autobahn network. Using a java applet the results are made topical minute by minute in the internet (www.autobahn.nrw.de).

TRAFFIC FORECAST IN LARGE SCALE FREEWAY NETWORKS

Traffic flow in large and complex freeway networks is a highly nonlinear phenomenon, which makes traffic forecast a difficult task. In this article an approach to traffic forecast is presented, whi...

OLSIM: inter-urban traffic information

Nowadays detailed and reliable information about current and future traffic states is a crucial requirement not only for each modern traffic control centre, but also for driver accessible traffic information applications. In this work the novel traffic information system OLSIM is presented that is based on three components. The first is a highly realistic traffic flow model with which all vehicles in a large scale network are simulated. The second component consists of efficient data processing and forecast algorithms that form heuristics from a large database that is fed every minute by traffic data of 4,000 inductive loop detectors across the road network. The third is a graphical user interface which can be accessed at www.autobahn.nrw.de. More than 200,000 users each day indicate the importance of such a system.

On-Line Simulation of Large Scale Networks

In this contribution a micro-simulator for freeway traffic in large scale networks is presented. Based on a cellular automaton traffic flow model the simulation, which is coupled with measured data from induction loops etc., can provide network-wide information about the current traffic state. The simulation of the whole network is performed in multiple real-time. The framework is applied to the freeway network of North Rhine-Westphalia (NRW), where traffic data from about 3,500 inductive loops are available and provided on-line minute by minute. Technical aspects of the simulation, e.g., the network structure and the incorporation of the on-line data, are illustrated. This is followed by a discussion of the efficiency of the simulator. Furthermore, the results of the on-line simulation, i.e., daily traffic demands and the traffic state of NRW are given.