Activity-based models: agent negotiation to cooperate for carpooling

Abstract

Carpooling is the co-travelling of people with similar destination in a similar time period and is considered to be an effective alternative transportation mode in order to counter alarmingly increasing traffic congestion and pollution emission levels. Carpooling can be very beneficial not only financially for the involved parties but also in a larger context of preserving environment and social welfare. However, according to different studies there is a very limited interest of people in carpooling especially for longer periods. This is mainly due to the lack of flexibility in people's daily schedule and tightness of the time intervals between certain fixed activities. As carpooling involves two or more individuals who have to negotiate and adapt their daily agenda in order to accommodate for cooperation; therefore, modelling the effects of carpooling mechanism is not a straight forward task. For this purpose, agent based models are used to simulate the carpooling behaviour that involves the interaction of autonomous entities termed as agents. A number of conceptual carpool models have been discussed in the literature review section of the report. A detailed study of the existing agent based carpool models reveals that the models are not accurate as they do not employ a behaviorally correct negotiation mechanism.Therefore, our model aims to extend the previous models by incorporating a more realistic departure time preference function for each agent by considering three different types of factors namely; (i) traveling f