Calibration of Platoon Dispersion Parameters on the Basis of Link Travel Time Statistics

Abstract

A calibration technique for platoon dispersion parameters for the widely used TRANSYT platoon dispersion model is presented. This technique calibrates platoon dispersion factor, travel time factor, and smoothing factor directly from the average link travel time and its standard deviation and can capture practically all of the roadway and traffic conditions in the field such as road grades, curvature, parking, opposing flow interference, traffic volume, and other sources of impedance. The technique is especially suited for applications in advanced traffic management system networks in which the required link travel time data could be obtained on a real-time basis. TRANSYT’s implementation of two scenarios is examined. The first scenario inputs the calibrated platoon dispersion parameter, with the result being that the smoothing factor used by TRANSYT is different from the calibrated parameter. The second scenario inputs a revised platoon dispersion factor, which is designed to make the smoothing factor used by TRANSYT identical to the calibrated parameter. This examination induces a recommendation that the TRANSYT input card or its internal calculation procedure for platoon dispersion be revised so that the average link travel time in the geometric distributed platoon dispersion model is consistent with those from the same model. The calibration of platoon dispersion parameters with field-collected link travel time data shows that platoon dispersion parameters are different for different standard deviations of link travel times even on the same street, and, therefore, the platoon dispersion parameters must be calibrated on a site-specific basis.