CumInCAD is a Cumulative Index about publications in Computer Aided Architectural Design
supported by the sibling associations ACADIA, CAADRIA, eCAADe, SIGraDi, ASCAAD and CAAD futures

PDF papers
References

Hits 1 to 1 of 1

_id ddss2006-hb-151
id DDSS2006-HB-151
authors Jean-Marie Boussier, P. Estraillier, D. Sarramia, and M. Augeraud
year 2006
title Approach to Design Behavioural Models for Traffic Network Users - Choice of transport mode
source Van Leeuwen, J.P. and H.J.P. Timmermans (eds.) 2006, Innovations in Design & Decision Support Systems in Architecture and Urban Planning, Dordrecht: Springer, ISBN-10: 1-4020-5059-3, ISBN-13: 978-1-4020-5059-6, p. 151-166
summary Our research work concerns the development of a multimodal urban traffic simulator designed to be a tool of decision-making aid similar to a game wherein the user-player can test different scenarios by immersion in a 3D virtual city. Our approach is based on the activity-based model and the multi-agent technology. The implemented result is a hybrid simulator connecting numerical simulation and behavioural aspects coming from real data. This paper is focused on two points: firstly, we introduce how a final user (the traffic regulator) instantiates and assembles components so as to model a city and its urban traffic network; secondly, we present the use of Dempster-Shafer theory in the context of discrete choice modelling. Our approach manipulates input variables in order to test realistic representations of behaviours of agent categories in a decision-making process. The traffic modelling is based on a questionnaire elaborated from standard arrays of Taguchi. The significant variables and interactions are determined with the analysis of variance which suggests a reduced model describing the behaviour of a particular social category. The belief theory is used to take into account the doubt of some respondents as well as for the preferences redistribution if the number of alternatives changes. The effects of external traffic conditions are also quantified to choose a 'robust' alternative and to use the agents' memory.
keywords Urban traffic simulator, Virtual city, Multi-agent system, Behavioural model, Transport mean, Taguchi's method, Belief theory
series DDSS
last changed 2006/08/29 10:55

No more hits.

HOMELOGIN (you are user _anon_943319 from group guest) CUMINCAD Papers Powered by SciX Open Publishing Services 1.002