Does anisotropy hold in mixed traffic conditions?

Traditionally, car-following models have focused on the interaction between leader and follower vehicles. However, some researchers have explored the significance of incorporating upstream vehicle information and considering non-anisotropic traffic flow patterns for homogeneous traffic conditions. This paper presents a comprehensive investigation into the impact of upstream vehicle information on car-following models for mixed traffic conditions. We employ three different approaches, macroscopic, microscopic, and nonparametric modeling, using empirical data to examine the evidence for the existence of non-anisotropy in mixed traffic conditions. We investigate the macroscopic evidence of fundamental diagrams, revealing complex interactions among heterogeneous traffic. The propagation of congestion can exceed the travel speed of the traffic stream itself, emphasizing the intricate nature of traffic dynamics. Our findings demonstrate that incorporating upstream vehicle information significantly enhances the prediction accuracy of car-following models. Particularly in scenarios where heavy vehicles are present upstream, the inclusion of upstream information proves to be crucial. Further, we conduct a microscopic analysis using triplet trajectories involving heavy vehicles to validate our findings, utilizing Edie’s definition. The microscopic validation supports the presence of non-anisotropic fundamental diagrams and reinforces the need to consider upstream interactions in car-following models.