Search

• Scenario-specific electric vehicle behavioral model. • Higher accuracy in representing electric vehicles' movements. • Applicability to timid, aggressive and normal vehicles. • Ability to represent traffic oscillations of electric vehicles. Electric vehicles (EVs) are deemed to be a solution for reducing air pollution and greenhouse gas emissions. As a result, the market share has increased exponentially in recent years. Despite their distinct vehicle dynamics and characteristics, movement simulation models dedicated to EVs are yet to be developed. In this research, a micro-traffic flow model for EVs by considering their unique acceleration/deceleration characteristics is proposed to represent and simulate the movements of EVs in traffic flow, especially in congested traffic. Car-following pairs where second car is an EV were collected from Longpan mid road, Nanjing, China in March 2019 for model calibration and verification. The results show that the proposed EV behavior model outperforms traditional behavior models for both timid and aggressive drivers. In assessing the predictive power of the movement simulation models, we compare their performance for collected car-following pairs. The R-squared values indicate that the performance of the EV behavior model is similar to that of the asymmetric behavior model under free-flow conditions, but substantially better for congested scenarios. With this model, we can better understand and reproduce the trajectories and energy consumption of EVs in complex traffic flow scenarios, and especially in congested traffic. [ABSTRACT FROM AUTHOR]

The widespread adoption of emerging connected and automated vehicles (CAVs) highlights the need for identifying the roadway capacity of mixed traffic flow with CAVs and human-driven vehicles (HDVs) for future traffic management. Previous studies focus on analyzing the impacts of CAV technologies on the mixed traffic capacity. However, research on how different car-following behaviors of HDVs to follow CAVs affect the mixed traffic capacity is still lacking. This study proposes an analytical formulation of the mixed traffic capacity based on the probability distributions of HDVs and CAVs and three types of behavior willingness (i.e., believer, neutral, and skeptic) of HDV-following-CAV in a mixed traffic environment. The sufficient conditions are derived for mixed traffic capacity to increase or decrease with the behavior willingness and CAV penetration rate. Based on that, an analytical lane management (LM) model is further proposed to determine the optimal number of CAV dedicated lanes (CDLs) and CAV non-dedicated lane (CNL) strategies that maximize the mixed traffic throughput, considering the mixed traffic demand, behavior willingness, and CAV penetration rate. The results from numerical experiments show that the proposed LM strategies can significantly improve mixed traffic throughputs by deploying optimal CDLs and CNL strategies, compared to one without the LM strategy. The mixed traffic throughputs improve and reduce as the trust willingness and skeptic willingness increase and decrease, respectively. Moreover, the increment in CAV penetration rate does not certainly correspond to the high mixed traffic throughputs. The behavior willingness and heterogeneous headways could be jointly considered to guarantee the positive influences of CAV penetration rate on the mixed traffic throughputs when implementing the LM strategy for the mixed traffic flow. • An mixed traffic flow capacity model considering different car-following behaviors developed. • Sufficient conditions for mixed traffic capacity to increase/decrease with parameters derived. • Mixed traffic throughput models considering the lane-level CAV penetration rate formulated. • Numerical analysis to verify the effectiveness of the lane management strategy conducted. [ABSTRACT FROM AUTHOR]