Traffic modeling: micro-analytical meets micro-simulation

Abstract

Powerful concepts have been developed and used in traffic modelling for a long time.  These apply to key elements of analytical models and are related to methods used in microsimulation models. This presentation recommends that formal procedures are developed for cross-verification (benchmarking) and calibration of analytical and microsimulation models using the well-established micro-analytical and microsimulation concepts. This effort should aim to provide model transparency which is essential for the model users to understand the model output that software generates. The presentation discusses a number of key concepts that apply to analytical and microsimulation models. These include the importance of lane-based modeling, lane use at intersection approaches and midblock lane changes, acceleration - deceleration, definitions and measurement methods for consistency of traffic performance variables, environmental objectives, capacity analysis, near-saturated and oversaturated conditions, queue discharge, gap-acceptance, arrival headway distributions and bunching models, and lane blockage and capacity constraint in networks.