An agent-based micro-simulation framework for modelling of dynamic activity–travel rescheduling decisions

The simultaneous implementation of daily activity–travel schedules of individuals in a given spatial environment generally gives rise to time- and location-varying congestion levels, which affect the conditions for subsequent activity and travel choices. Although such dynamics are commonly recognized, current activity-based models typically ignore the adaptive behaviour of individuals. In this article, we propose an agent-based simulation system that allows one to simulate, in addition to activity-scheduling behaviour, also the execution of schedules in space and time. Congestion levels at specific times and places emerge in the system and may lead to discrepancies between scheduled and actual activity and travel times. Agents respond to such unforeseen events by reconsidering an existing schedule (within-day re-planning) and by adapting their expectations about traffic conditions for subsequent days (learning). The system is illustrated using the activity–travel diary data collected in the Eindhoven region, the Netherlands, to better understand the choice of urban parks in the study area. We discuss the merits of the system for transport and spatial planning and identify avenues for future research.