Auflistung nach Schlagwort "Traffic Simulation"
1 - 2 von 2
Treffer pro Seite
Sortieroptionen
- KonferenzbeitragAthos: An Extensible DSL for Model Driven Traffic and Transport Simulation(Modellierung 2020, 2020) Hoffmann, Benjamin; Urquhart, Neil; Chalmers, Kevin; Guckert, MichaelMulti-agent systems may be considered appropriate tools for simulating complex systems such as those based around traffic and transportation networks. Modelling traffic participants as agents can reveal relevant patterns of traffic flow. Upsurging traffic in urban areas increases the relevance of such simulations and the insight they provide into reducing congestion and pollution. Developing multi-agent traffic simulations is a challenging task even for professional software developers. In contrast, domain experts need tools that can be quickly adapted to new questions emerging in their research without potentially error-prone communication with software developers. There is a need for simulation tools that are intuitive to domain experts yet flexible and adaptable by software developers as required. A model driven approach with an extensible domain specific language delivers an answer for both of these opposing requirements. The modeller is relieved from implementing time consuming programming details and can focus on the application itself. We present the domain specific language Athos that allows to create simulations of traffic and transport related problems declaratively. The models are platform independent and executable code can be generated for appropriate multi-agent platforms. The language is flexible and can be easily extended by exploiting the structure of the problem domain itself. In this paper, we present Athos and focus on how it can be extended by arbitrary traffic and routing algorithms through an annotation-based extension mechanism.
- KonferenzbeitragWhat does it Take to Test a Bicycle Simulator for Realism? A Discussion of the Challenges and Possible Evaluation Methods(Proceedings of Mensch und Computer 2024, 2024) Trefzger, Mathias; Raschke, Michael; Fath, Michael; Eckart, JochenIn recent years, bicycle simulators have become increasingly important as a research tool, resulting in a large number of new prototypes at various academic institutions. In order for the results of studies to be transferred from a simulator to reality, it is necessary to ensure that simulators reproduce realistic traffic behaviour. Most bicycle simulators are evaluated using cycling parameters. However, the (combined) use of other methods is necessary to ensure realism. In this paper we present a variety of evaluation methods and discuss why we consider them relevant for evaluating the realism. We present the challenges and potential and the study design of our current study comparing cycling behaviour in reality and in simulation. Our study design should serve as a reference for other researchers and facilitate their own research. This paper contributes to the motivation of new and future work in the still young field of cycling simulation.