Calibration and Validation of Microscopic Models of Traffic Flow

Abstract

Because microscopic models are heavily used in applications, the appropriate calibration and validation of these models have been a recent concern. The contribution of this paper is to compare some of these models by calibrating and validating them with data from double loop detectors on a multilane freeway. To simplify this task, the models were tested by simplifying the multilane reality to a simulation of only a single lane. The results show that by simulating the multilane road with single-lane models, calibration errors (Theil's U-value, or the root mean square error) of 14% to 16% were obtained. A validation of the models, which was done by taking the calibrated parameters of one data set to reproduce the other data sets, gives additional errors of about 0.5 to 2.5 percentage points. This is in good agreement with other calibration and validation approaches performed recently.

The principles of calibrating traffic microsimulation models

“…GEH \ 5 indicates a good fit Hourdakis et al (2003) Theil's bias proportion (Um) Hourdakis et al (2003), Barcelo and Casas (2004), Brockfeld et al (2005) A high value implies the existence of systematic bias. Um = 0 indicates a perfect fit, Um = 1 indicates the worst fit measures uses a power higher than 2.…”

“…The fact that in each dimension the measures of fit have a different meaning is most apparent when examining variations between the measurements. For instance, calibrating a TMM by focusing on estimates of variation of the travel speed over different time Hourdakis et al (2003), Barcelo and Casas (2004), Brockfeld et al (2005) A high value implies that the distribution of simulated measurements is significantly different from that of the observed data. Us = 0 indicates a perfect fit, Us = 1 indicates the worst fit…”

“…Uc = 1 indicates a perfect fit, Uc = 0 indicates the worst fit. r is the correlation coefficient (2002), Hourdakis et al (2003), Toledo and Koutsopoulos (2004), Barcelo and Casas (2004), Brockfeld et al (2005) Combines effects of all 3 Theil's error proportions (Um, Us, Uc). U = 0 indicates a perfect fit, U = 1 indicates the worst fit…”

Microscopic driving theory with oscillatory congested states: Model and empirical verification

“…However, this model does not belong to this class since the fundamental hypothesis of KTPT is missing. Moreover, empirical calibration and validation results in a higher accuracy than that of previously investigated models (Brockfeld et al, 2005;Wagner, 2010). To these ends, Section 2 analyses the US-101 trajectory datasets on a single freeway lane, away from lane changes and the influence of bottlenecks.…”

Traffic monitoring with serial images from airborne cameras

“…Empirical studies, measurements, modeling attempts, and simulation programs are numerous and diverse. The challenge is to develop robust methods for predicting, visualizing and modeling of complex traffic events (Brockfeld et al, 2004). Only on the basis of reliable data, the simulation, control and planning of traffic systems can be optimized and, thereby, contribute to the efficient use of existing infrastructure, the reduction of emissions, and the increase of transport safety.…”