Comparing traffic state estimators for mixed human and automated traffic flows

Abstract

This article addresses the problem of modeling and estimating tra c streams with mixed human operated and automated vehicles. A connection between the generalized Aw Rascle Zhang model and two class tra c flow motivates the choice to model mixed tra c streams with a second order tra c flow model. The tra c state is estimated via a fully nonlinear particle filtering approach, and results are compared to estimates obtained from a particle filter applied to a scalar conservation law. Numerical studies are conducted using the Aimsun micro simulation software to generate the true state to be estimated. The experiments indicate that when the penetration rate of automated vehicles in the tra c stream is variable, the second order model based estimator o↵ers improved accuracy compared to a scalar modeling abstraction. When the variability of the penetration rate decreases, the first order model based filters o↵er similar performance.

Influence of CAV clustering strategies on mixed traffic flow characteristics: An analysis of vehicle trajectory data

“…Hence evaluation of multi-lane scenarios provide more realistic estimation, and they can be found in Arnaout and Arnaout (2014); Lee et al (2014); Songchitruksa et al (2016) The heterogeneous traffic flow attributed to partial market penetration has been examined in recent years. Wang et al (2017a) proposed a second-order traffic flow model for mixed traffic streams with HVs and AVs. Xiao et al (2018) studied the traffic flow characteristics using a multiregime car-following model that is capable of switching among human-driven, ACC, and CACC modes.…”

A multiscale model for traffic regulation via autonomous vehicles

“…Autonomous Vehicles (briefly AVs) represent the most disruptive technology for traffic regulation [15,25,32,33,35]. The effect of AVs in terms of influencing bulk traffic has been studied in-silico [7,14,18,30,34], artificial environment [17] and also in experiments [29].…”

Traffic state estimation using stochastic Lagrangian dynamics

“…For example, stochastic extensions of the cell transmission model [5,6] have been proposed [37,18]; other approaches have extended the link transmission model [44], both at the individual link level and the network level [33,32,34,26]. In general, there still remain issues related to the physical accuracy of the sample paths of existing stochastic traffic models, particularly those developed for purposes of traffic state estimation (see [36,39] for recent reviews). The main culprit is the dominance of time-stochasticity (or noise) in the stochastic models, mostly developed in Eulerian coordinates [15,38,28,16,40,2,41,43,10,37,1,19], but also in Lagrangian coordinates [46,45,3].…”