Pengfei Tao

In complex traffic environment, it is still a great challenge for Autonomous Vehicles (AVs) to understand the surrounding drivers’ Lane-Changing Decision (LCD) intention accurately. The LCD intention is affected by Driver’s Psychology (DP) and Driving Style (DS). But few LCD studies considered DP and DS simultaneously. We previously proposed a LCD model, by combing DP and DS, termed as DP&DS-LCD model. Nevertheless, there are some factors not fully considered in this model, including the driver’s Visual Attention (VA) in DP quantification and the influence of driving state on DS. Therefore, an enhanced LCD Model is developed, by integrating the DP under VA (DPVA) and DS Layering (DSL), named as DPVA&DSL-LCD model. In the model, a psychological field model coupling driver’s VA mechanism is established to represent the surrounding vehicles’ influence on the driver. Then, a DSL framework is proposed by adding the influence of driving state on DS. The Gaussian Mixture Model (GMM) clustering and Support Vector Machine (SVM) classifier are respectively adopted in training and recognition phases to identify the current driving style. Finally, integrating the DPVA and DSL, the Light Gradient Boosting Machine (LightGBM) algorithm is used to train the LCD model. In experiments, the open I-80 database from Next Generation Simulation (NGSIM) is adopted to train the DPVA&DSL-LCD. And compared with other three LCD models, the prediction performance of DPVA&DSL-LCD model achieved the best. Therefore, the DPVA&DSL-LCD model is effective and could provide support for the decision-making of AVs by predicting surrounding vehicles’ LCD intention.

Unraveling the determinants of travelers’ parking behavior intentions is critical to the widespread adoption of smart parking systems (SPSs), which hold the promise of greatly enhancing parking efficiency and optimizing resource allocation within urban spaces. Our study pioneers the use of an integrated methodology combining structural equation modeling (SEM) and hierarchical regression modeling (HRM) to dissect the complex interplay of these determinants. We found that, in the structural equation model, social influence notably stood out as having the most significant impact on the intention to utilize SPSs. Notably, while perceived privacy concerns may have ranked lower in terms of influence among these factors, their role was relatively crucial, particularly given the contemporary emphasis on data security. Moreover, within the hierarchical regression model, driving experience was found to play a crucial role in determining the intention to use SPSs. Equally important, our research revealed a divergence in parking intentions between individuals with children and those without. This points towards the imperative need for personalized strategies that can cater to the diverse requirements of different user demographics. This research offers guidance for operators of SPSs aiming to formulate targeted approaches.

Pavement segregation reduces pavement strength and causes early pavement damages. As a convenient and nondestructive subsurface detection technology, Ground Penetrating Radar (GPR) is widely applied to detect segregation by manual interpretation. This induces labor-intensive and time-consuming. And then the segregation is interpreted automatically by image processing technology based on GPR B-scan images. However, this ignores the intrinsic physical properties of segregation. To address the problem, a new method for automatically detecting segregation is proposed to interpret segregation by combining visual features with intrinsic physical properties of segregations. In the method, GPR data preprocessing is carried out to remove the subgrade layer disturbance and enhance GPR signal. And then the candidate segregation locating is put forward by segregation representation index and graph cuts algorithm according to segregation visual and statistical characteristics. Segregation visual features and intrinsic physical properties are represented for recognition according to gray level co-occurrence matrix and dielectric properties. The extreme gradient boosting is employed as a classifier to recognize segregation. In experiments, experiment samples are collected by 3D GPR testing vehicle and coring experiments, and comparison experiments are carried out to validate the proposed method by analysis on feature selection, segregation localization and segregation recognition. The results show the proposed method has a better performance than the other methods, such as multi-grained cascade forest, deepLab v3+ and SegNet. As a result, the proposed method will provide a new reference for automatic interpretation GPR application, especially, on the combining visual features with intrinsic physical properties.

Currently, many types of sensors, such as cameras and radars have been widely deployed on road networks according to historical accident data for traffic accident detection and prevention. However, the uncertainty of traffic accidents and finite historical accident data will induce the bias of sensor deployment. As a result, engineers have to adjust the locations of sensors many times to pursue better accident detection and prevention performance in practice. This will cause a large number of engineering costs and can hardly realize the optimal reallocation of sensors. To address the problem, an optimal reallocation method of heterogeneous sensors is proposed. In the method, to reduce the influence of accident uncertainty under finite historical accident samples, the optimal transport is introduced to estimate the stable spatial distribution of accident risk. Then the sensor reallocation problem is reduced as a coverage enhancement problem on the stable spatial distribution of accident risk. The optimal reallocation model of heterogeneous sensors is constructed to optimize the type, number, and location of reallocated sensors by maximizing the coverage quality with the constraint of cost-effectiveness, the number of existing heterogeneous sensors, and so on. And the bisection method and AO-GI-MO algorithm are combined to solve it. In experiments, stability, consistency, reliability, and comparison experiments are carried out to validate the proposed method by using international open traffic accident data. The results show the proposed method could improve the detection performance of traffic accidents effectively.

In order to take more active measures to prevent and control secondary accidents, it is necessary to describe risk propagation process after the accident. To this end, this paper deeply analyzed the risk propagation mechanism between vehicles and proposed a novel secondary rear-end collision accident risk propagation model, which could real-time evaluate vehicle rear-end collision risk. The research scene of the paper is a single-lane road rear-end collision scene, so a driving risk field model suitable for this scene is first established. Based on this, the vehicle operation interaction risk field force is calculated. Then, the risk field force is converted into risk probability through the hyperbolic tangent function, and the vehicle operation interaction risk model is obtained. In addition, a risk propagation framework based on dynamic Bayesian network is constructed to describe the propagation process of rear-end collision risk from the accident vehicle to following vehicles. Finally, according to the probabilistic reasoning process of the framework, combined with the accident vehicle risk, a secondary rear-end collision accident risk propagation model is established. Simulation experiments show that the paper model can describe the evolution trend of rear-end risk and the risk assessment results are more accurate. And after the accident, the rear-end collision risk propagation speed will increase with the increase of traffic flow, and the number of secondary rear-end vehicles will also increase with the increase of traffic speed. These conclusions are of great significance in formulating vehicle anti-collision strategies and deploying risk management and control facilities.

The traffic state of the bus line is the information basis for the bus company to make bus dis-patch and travel time prediction. However, the bus GPS data is severely sparse in time and space coverage of traffic state, due to the long data sampling time interval and low bus departure frequency. Because of ignoring the severe sparseness of the bus data, the existing traffic state methods cannot reconstruct the traf-fic state accurately. To deal with this problem, a new traffic state estimation method for the bus line, named GANBS, is proposed. First, an improved generative adversarial network (GAN-I) is used to generate rea-sonable bus data. GAN-I aims to find the probability space of the data distribution under sparse sampling. And to reduce the size of the latent space of data, the traffic knowledge is introduced as prior information layers. Then, a traffic adaptive bilateral smoothing method (BS) is used to map discrete bus data into the continuous traffic state. The BS convolves data with a bilateral kernel, which multiplies the local action kernel with a mask of traffic state similarity. Therefore, the BS can maintain transitions between different traffic patterns while separating noise from traffic state. Finally, a set of numerical experiments are per-formed on the real bus data set in Changchun. The results show that the GAN-I can accurately reproduce the traffic state when the missing rate of data exceeds 50%. And the BS can eliminate the noise better com-pared with other methods.

Lane-Changing Decision (LCD) behavior is complex, dangerous, and varied by the Driver’s Psychology (DP) and Driving Style (DS). For lacking the consideration of DP and DS, the existing LCD models cannot predict the LCD process varying with the drivers and traffic conditions accurately. To deal with the problems, a new LCD model coupling DP and DS, named DP&DS-LCD, is put forward. In the model, a psychological field model is constructed to represent the DP effect on the scene vehicles. And then, K-means and K-Nearest Neighbor (KNN) algorithm are respectively adopted in learning and recognition phases to recognize the current driving style pattern. Finally, based on the DP and DS, the multi-Grained Cascade Forest (gcForest) algorithm is applied to predict the LCD behavior. In experiments, DP&DS-LCD is compared with other three LCD models by using the opening I-80 database from Next Generation Simulation project (NGSIM). And the results showed that the DP&DS-LCD model achieved the best performance. Therefore, the DP&DS-LCD model is effective and could provide support for the decision of autonomous vehicles by predicting the surrounding vehicles’ Lane-Changing (LC) behavior.

Bicycle traffic has heavy proportion among all travel modes in some developing countries, which is crucial for urban traffic control and management as well as facility design. This paper proposes a real‐time multiple bicycle detection algorithm based on video. At first, an effective feature called multiscale block local binary pattern (MBLBP) is extracted for representing the moving object, which is a well‐classified feature to distinguish between bicycles and nonbicycles; then, a cascaded bicycle classifier trained by AdaBoost algorithm is proposed, which has a good computation efficiency. Finally, the method is tested with video sequence captured from the real‐world traffic scenario. The bicycles in the test scenario are successfully detected.

The conventional car-following theory is based on the assumption that vehicles will travel along the center line of lanes. However, according to the field survey data, in complex traffic conditions, a lateral separation between the leader and the follower frequently occurs. Accordingly, by taking lateral separation into account, we redefined the equation of time-to-collision (TTC) using visual angle information. Based on the stimulus-response framework, TTC was introduced into the basic General Motors (GM) model as a stimulus, and a non-lane-based car-following model of steady-state traffic flow was developed. The property of flow-density relationship was further investigated after integrating the proposed car-following model with different parameters. The results imply that the property of steady-state traffic flow and the capacity of each lane are highly relevant to the microscopic staggered car-following behavior, and the proposed model significantly enhances the practicality of the human driving behavior model.