Two Level Approach for Validation of Microscopic Simulation Models

New driver assistance systems can have consequential effects on traffic flow. Investigative simulations are necessary for the design, optimization and evaluation of these assistance systems. Previous calibration and validation methods utilized either microscopic or macroscopic measurement data. This paper's purpose is to argue that the formerly held calibration and validation perspectives with regard to traffic simulations are incomplete. Moreover, these assistance systems have their own set of particular requirements, and require the simultaneous consideration of microscopic and macroscopic system behavior. Therefore, this paper recommends that a measurement concept is needed to gain the required data necessary for proper calibration and validation. The concept presented in this paper advocates simultaneous measurements sourced in both a vehicle (microscopic) and overall traffic (macroscopic) perspective.