Narrow Results

In this paper, we proposed a model based on the connected vehicles to control the traffic flow at the intersection. To evaluate this model, we studied its impact on the flux and the probability of accidents at the intersection. On the one hand, simulation results showed that the increase in the number of connected vehicles decreases the congestion coefficient and enhances the traffic situation in the system. On the other hand, connected vehicles reduce the probability of accidents at the intersection. In this way, the vehicle intersection (central controller) communication can decrease the congestion and enhance road safety, especially in the intermediate and high traffic conditions.

Connected vehicles (CVs) have led to innovative changes in the transportation field. In this wise, an extended CA model was established incorporating the CVs and regular vehicles to investigate a roundabout traffic flow under the open boundary condition. The throughput and the phase diagrams of the system in the (${\alpha }_1,{\alpha }_2$) area is constructed in both cases: with and without CVs. The density profile and the capacity are also calculated. The simulation findings show that CVs presentation changes the traffic dynamics, enhancing roundabout capacity in accordance with the CVs penetration rate. Furthermore, it is found that roundabout size plays a dominant role in its performance where it suggests that the midsized roundabouts operate more optimally. Another relevant aspect is the short sized of Splitter Island shows a better performance. Moreover, the traffic dynamics is negatively (resp. positively) correlated to the probability of displaying indicator $P_{\mathrm{\text{ind}}}$ as well as the probability of aggressive driving behavior of regular vehicles (resp. the cooperation probability $P_{\mathrm{\text{cop}}}$).

This paper explored the impacts of vehicle-to-infrastructure (V2I) communication on the mixed traffic flow consisting of connected vehicles (CVs) and human-driven vehicles (HVs). We developed a cellular automaton model for mixed flow at the signalized intersection. In addition to considering the motion characteristics of CVs and the influence of HVs on the motion behavior of CVs, the model also considered the influence of signal lights. CVs determine their velocities via V2I communication in order to pass the signal light with less delay and avoid stopping. Through simulations, we found that the presence, frequency and range of V2I communication all make a difference in the mixed flow. Also, 1-Hz communication reduces the number of vehicles within 300 m before the red light from 36 to 26, and the 10-Hz communication reduces one more; 1-Hz communication increases the number of accelerations, but when the frequency increases to 10 Hz, the number of accelerations decreases to the same value as without V2I communication, but the value of number of accelerations increases monotonously with the frequency; traffic delay decreases and capacity increases as the frequency increases. However, as the communication range increases, except that the number of accelerations first decreases and then increases, other traffic characteristics remain unchanged. The number of accelerations reaches a minimum at about 500 m.

Under the Vehicle-to-Vehicle (V2V) environment, connected vehicles (CVs) can share the traveling information with each other to keep the traffic flow stable. However, the open network cooperation environment makes CVs vulnerable to cyberattacks, which leads to changes in driving behavior. The existing theories divide cyberattacks into three types: bogus information, replay/delay and collusion cyberattacks. In addition, the mixed flow consisting of truck and car is a common form of road traffic. In order to clarify the potential impact of cyberattacks on mixed traffic flow, this paper proposes an extended car-following model considering cyberattacks under CVs environment. Subsequently, the stability of the model is analyzed theoretically, and the stability condition of the model is obtained. The numerical simulation is carried out and the result shows that the cyberattacks lead to different degrees of traffic behavior hazards such as queue time extension, congestion and even rear end collision. Among them, cooperative attack is the most serious.

This study proposes a data fusion and deep learning (DL) framework that learns high-level traffic features from network-level images to predict large-scale, multi-route, speed and volume of connected vehicles (CVs). We present a scalable and parallel method of processing statewide CVs’ trajectory data that leads to real-time insights on the micro-scale in time and space (two-dimensional (2D) arrays) on graphics processing unit (GPUs) using the Nvidia rapids framework and dask parallel cluster, which provided a 50× speed-up in the data extraction, transform and load (ETL). A UNet model is then applied to perform feature extraction and multi-route speed and volume channels over a multi-step prediction horizon. The accuracy and robustness of the proposed model are evaluated by taking different road types, times of day and image snippets and comparing the model to benchmarks: Convolutional Long–Short-Term Memory (ConvLSTM) and a historical average (HA). The results show that the proposed model outperforms benchmarks with an average improvement of 15% over ConvLSTM and 65% over the HA. Comparing the image snippets from each prediction model to the actual image shows that image textures were highly similar in UNet to the benchmark models used. UNet’s dominance in performing image predictions was also evident in multi-step forecasting, where the increase in errors was relatively minimal over longer prediction horizons.