Validation of an agent-based travel demand model with floating car data

This paper compares the results of the agent-based travel demand model MITO (Microscopic Travel Demand Orchestrator) with Floating Car Data. MITO is developed using household travel survey data, and uses the traffic assignment model MATSim. The model estimates the travel demand for an average working day and is applied to the metropolitan area of Munich. In contrast to traditional approaches where travel demand models are validated using the local traffic counts, average travel speed from Floating Car Data (FCD) are used in this study. The main advantage of using FCD is that they cover extremely large parts of the network, whereas the local traffic counts are sparse and limited to a few major streets. The average link travel time and average speed between the model estimation and the FCD were compared with the goal of validating travel time calculations within an agent-based transport model.