As the accident-prone sections and bottlenecks, highway weaving sections will become more complicated when it comes to the mixed-traffic environments with connected and automated vehicles (CAVs) and human-driven vehicles (HVs). In order to make CAVs accurately identify the driving behavior of manual-human vehicles to avoid traffic accidents caused by lane changing, it is necessary to analyze the characteristics of the mandatory lane-changing (MCL) process in the weaving area. An analytical MCL method based on the driver’s psychological characteristics is proposed in this study. Firstly, the driver’s MLC pressure concept was proposed by leading in the distance of the off-ramp. Then, the lane-changing intention was quantified by considering the driver’s MLC pressure and tendentiousness. Finally, based on the lane-changing intention and the headway distribution of the target lane, an MLC positions probability density model was proposed to describe the distribution characteristics of the lane-changing position. Through the NGSIM data verification, the lane-changing analysis models can objectively describe the vehicle lane-changing characteristics in the actual scenarios. Compared with the traditional lane-changing model, the proposed models are more interpretable and in line with the driving intention. The results show significant improvements in the lane-changing safe recognition of CAVs in heterogeneous traffic flow (both CAVs and HVs) in the future.

Aimed at the vehicle/pedestrian visual sensing task under low-light conditions and the problems of small, dense objects and line-of-sight occlusion, a nighttime vehicle/pedestrian detection method was proposed. First, a vehicle/pedestrian detection algorithm was designed based on You Only Look Once X (YOLOX). The model structure was re-parameterized and lightened, and a coordinate-based attention mechanism was introduced into the backbone network to enhance the feature extraction efficiency of vehicle/pedestrian targets. A feature-scale fusion detection branch was added to the feature pyramid, while a loss function was designed, which combines Complete Intersection Over Union (CIoU) for target localization and Varifocal Loss for confidence prediction to improve the feature extraction ability for small, dense, and low-illumination targets. In addition, in order to further improve the detection accuracy of the algorithm under low-light conditions, a training strategy based on data domain transfer was proposed, which fuses the larger-scale daylight dataset with the smaller-scale nighttime dataset after low-illumination degrading. After low-light enhancement, training and testing were performed accordingly. The experimental results show that, compared with the original YOLOX model, the improved algorithm trained by the proposed data domain transfer strategy achieved better performance, and the mean Average Precision (mAP) increased by 5.9% to 82.4%. This research provided effective technical support for autonomous driving safety at night.

To enable efficient data gathering in wireless sensor networks, a novel partial matrix completion algorithm is proposed by introducing historical data to the original data matrix. The recovery accuracy is further improved by making full use of the band-limited feature of the sensory data. We analyzed the recovery error of the proposed algorithm and derived a theoretical error upper bound. Experimental results on two real-world data sets show that the proposed algorithm achieves good recovery accuracy even when the sampling ratio is very low, which outperforms the state-of-the-art methods.

Modern complexities associated with an arterial traffic makes existing safety prediction methods insufficient to meet desired standards required by recent developmental needs. This paper proposes an enhanced active safety prediction method based on big-data approach and Stacked AutoEncoder-Gated Recurrent Unit. Firstly, the big-data technology is used to construct a dynamic identification model to recognize real-time operation state and risk state. Secondly, the Stacked AutoEncoder-Gated Recurrent Unit is used to predict a level of safety based on associated recognition results. This paper uses data from working days of Sunset Boulevard, California, from January (1st), 2020, to February (28th), 2020. The results of analysis show that the accuracy of the proposed dynamic recognition model reaches 98.92%, which is better than existing models such as random forest, K-nearest neighbor, and naïve Bayes models. In addition, it is found that the Stacked AutoEncoder-Gated Recurrent Unit can achieve a prediction accuracy of 95.157% and has significant advantages in terms of efficiency. The proposed methods will provide feasible solutions for actively monitoring safety levels.

With the exponential growth of traffic data and the complexity of traffic conditions, in order to effectively store and analyse data to feed back valid information, this paper proposed an urban road traffic status prediction model based on the optimized deep recurrent Q-Learning method. The model is based on the optimized Long Short-Term Memory (LSTM) algorithm to handle the explosive growth of Q-table data, which not only avoids the gradient explosion and disappearance but also has the efficient storage and analysis. The continuous training and memory storage of the training sets are used to improve the system sensitivity, and then, the test sets are predicted based on the accumulated experience pool to obtain high-precision prediction results. The traffic flow data from Wanjiali Road to Shuangtang Road in Changsha City are tested as a case. The research results show that the prediction of the traffic delay index is within a reasonable interval, and it is significantly better than traditional prediction methods such as the LSTM, K-Nearest Neighbor (KNN), Support Vector Machines (SVM), exponential smoothing method, and Back Propagation (BP) neural network, which shows that the model proposed in this paper has the feasibility of application.

As the emergence of the connected and autonomous vehicles (CAVs), the platooning technology is believed to play a key role in the future intelligent transportation system. However, current studies mainly focus on the beneficial sides of CAV platoons, and less attention is given to their negative effects. This study develops a mixed equilibrium model for CAV platoons and human-driven vehicles (HDVs), which consider both the positive and negative sides of CAV platooning. On the positive side, CAV platoons are assumed to follow user equilibrium (UE) route choice for their information advantages, while HDVs to follow stochastic user equilibrium (SUE). CAV platoons are also presumed to improve the road capacity. On the negative side, the speed of CAV platoons is slower than that of HDVs for safety stakes, which will impede the latter to overtake. The HDVs is split up into overtaking and nonovertaking flows with different speeds. Furthermore, the model is built up as a mixed UE-SUE equilibrium problem and reformulated as a nonlinear complementarity problem. In addition, an optimal path platooning method is proposed to reduce the negative effects, by integrating travel costs of both CAV platoons and HDVs into its objective function. Numerical results show that the introduction of CAV platoons may increase the travel cost at the initial stage, and the proposed method can effectively reduce the platooning disturbance, thus helps promoting the wider applications of CAV platoons.