Impact of Driver Anticipation on Macroscopic Traffic

Abstract

Accurate characterization of heterogeneous traffic flow is essential for realistic traffic modeling. Driver anticipation in heterogeneous traffic is impacted by lateral and forward distance, vehicle size, and effective angle of vision, however most existing macroscopic models neglect these factors. Larger vehicles require greater acceleration and deceleration, while smaller vehicles maneuver more easily, and reduced visibility or limited angle of vision compromises driver response, contributing to traffic accidents. This study develops a macroscopic traffic model that explicitly incorporates driver anticipation based on effective angle of vision, distance headways, and vehicle size. An anticipation term is introduced to adjust density and velocity according to realistic traffic conditions. Further, source term based on the gravitational impact is included, to predict the velocity behavior at uphill and downhill grades. The proposed model is assesed against the classical Payne Whitham model on a circular and straight road under varying lateral headways. Results demonstrate that the proposed approach has more realistic temporal and spatial evolution of density and velocity, capturing heterogeneous flow dynamics more accurately than conventional models. These findings highlight the importance of integrating human factors, road grade, and vehicle characteristics into macroscopic traffic modeling to improve the traffic prediction accuracy.