Automatic calibration of agent-based public transit assignment path choice to count data

摘要

This work describes the calibration of a schedule-based transit assignment inside an iterative microscopic agent-based simulation. The calibration challenge implies that the behavioral rules should be modified in order to move the simulation closer to observed passenger counts. First, route choice set of agents is enriched with travel parameter utilities randomization. Secondly, the calibration interacts directly into the performance evaluation of individual daily plan of activities, so that the plan is also evaluated for its contribution to the count reproduction. In this way, appropriate plans from the calibration perspective can persist along simulation iterations. The Berlin public transport system with day-based counts is used as test scenario. The results show that the calibration approach can work with large scale scenarios, and that it is able to deal with the inter-temporal aspects implied by counts. (C) 2016 Elsevier Ltd. All rights reserved.