Using a MultiAgent Transport Simulation to Study the Impact of the Preference for Joint Activities on Leisure Trip Distance Distribution

Abstract

Since the beginning of transportation planning, researchers have made constant effort to improve the accuracy of forecasts. Though much progress have been done since the first commuter models, current approaches still have their flaws.

A particularly difficult problem is the accurate estimation of leisure trip distance distribution, and the changes of this property with changes in the transport supply. A particular difficulty arises from the fact that leisure travel is in essence social.

This works includes social networks and joint decision processes in a multi-agent simulation framework, and studies how good it allows to represent the specific features of leisure activities observed in data, namely trip distance distribution and high shares of joint travel.