Search

For on-demand dynamic ride-pooling services, e.g., Uber Pool and DiDi Pinche, a well-designed vehicle dispatching strategy is crucial for platform efficiency and passenger experience. Most existing dispatching strategies overlook incoming pairing opportunities, therefore suffer from short-sighted limitations. In this paper, we propose a forward-looking vehicle dispatching strategy, which first predicts the expected distance saving that could be brought about by future orders and then solves a bipartite matching problem based on the prediction to match passengers with partially occupied or vacant vehicles or keep passengers waiting for next rounds of matching. To demonstrate the performance of the proposed strategy, a number of simulation experiments and comparisons are conducted based on the real-world road network and open trip data from Haikou, China. Results show that the proposed strategy outperforms the baseline strategies by generating approximately 31% more distance saving and 18% less average passenger detour distance. It indicates the significant benefits of considering future pairing opportunities in dispatching, and highlights the effectiveness of our innovative forward-looking vehicle dispatching strategy in improving system efficiency and user experience for dynamic ride-pooling services.

A time-space (TS) traffic diagram, which presents traffic states in time-space cells with color, is an important traffic analysis and visualization tool. Despite its importance for transportation research and engineering, most TS diagrams that have already existed or are being produced are too coarse to exhibit detailed traffic dynamics due to the difficulty of collecting high-fidelity traffic data. To increase the resolution of a TS diagram and enable it to present ample traffic details, this paper introduces the TS diagram refinement problem and proposes a multiple linear regression-based solution. Data collected at different times, in different locations and even in different countries are employed to thoroughly evaluate the accuracy and transferability of the proposed model. Two tests, which attempt to increase the resolution of a TS diagram 4 and 16 times, are carried out to evaluate the performance of the proposed model. In the increase-4-times test, the errors represented by Mean Absolute Percentage Error are all less than 0.1, and in the increase-16-times test all less than 0.17. Model comparison demonstrates that the proposed model outperforms the classic adaptive smoothing method in refining TS diagrams. All the strict tests with diverse data show that the proposed model, despite its simplicity, is able to refine a TS diagram with promising accuracy and reliable transferability. The proposed refinement model will “save” widely existing TS diagrams from their blurry “faces” and enable TS diagrams to show more traffic details.

A speed perturbation model is proposed in this paper for the heterogeneous platoon to measure the interaction between vehicles and investigate the propagation laws of perturbations. Then, a ratio is defined to quantify the relationship between the communicating information impact (CII) and the car-following behaviour impact (CFI). A modified IDM is employed to evaluate the role of communicating information in perturbation propagation. Results show that communicating information can suppress the amplification of perturbation. Under the stable environment, the interaction between vehicles decreases with frequency. There is a critical frequency $ {\omega _c} $ ωc(value of 0.55 in this case) that distinguishes the relationship between the CII and CFI. When the frequency is smaller than $ {\omega _c} $ ωc, the CFI is larger than CII; otherwise, the CII dominates the interactions between vehicles. Under the unstable environment, the interaction between vehicles increases first and decreases then, where the vehicles are mainly affected by the car-following behaviours.

One of the most challenging problems in human-machine co-work is the gap between human intention and the machine's understanding and execution. Large Language Models (LLMs) have been showing superior abilities in solving such issue. In this article, we design a universal framework that embeds LLMs as a vehicle “Co-Pilot” of driving, which can accomplish specific driving tasks with human intention satisfied based on the information provided. Meanwhile, a utilization workflow is defined to handle the interaction between humans and vehicles, and memory mechanism is introduced to organize the information involved in the tasks. Expert Oriented Black-Box tuning is proposed to improve the performance of the Co-Pilot without finetuning or training the LLMs. In the experiment, the Co-Pilot is applied to two different tasks, i.e., path tracking control and trajectory planning. The Co-Pilot adjusts vehicle operating conditions by selecting a proper controller or planning a certain trajectory to fit human intentions. Simulation tests are conducted to evaluate the performance and generality of the proposed module. The results show that the Co-Pilot can accomplish most of the tasks based on only natural language processing, although it is not flawless. Finally, a discussion about human-machine hybrid intelligence and further applications of LLMs in autonomous driving is made. We believe that such a framework has promising potential in further applications in the field of automous vehicles.

Autonomous intersection management (AIM) guides vehicles to pass through signal-free intersections individually. Since vehicles are permitted to turn in any direction from any lane in AIM, it is desired to erase lane changes on the entry lane to achieve high traffic efficiency. However, erasing lane changes exposes intersections to complex conflicts. To study the trade-off between lane changes on the entry lanes and conflicts inside the intersection, this paper makes a comprehensive comparison of two kinds of intersection systems. First, all-direction turn lanes (ADTL) and specific-direction turn lanes (SDTL) are designed to distinguish lane change behaviors. Second, a two-stage method is proposed to manage the oncoming vehicles at the intersection. The first stage is timing schedule optimization and the second stage is trajectory optimization. Then, a method based on Monte Carlo Tree Search (MCTS) is designed to solve the timing schedule optimization model at high traffic demands. Finally, two parts of simulation experiments are conducted in this paper. The first part verifies the performance of the MCTS-based method. The second part compares ADTL and SDTL in multiple scenarios. The results show that ADTL outperforms SDTL in the efficient system, but the opposite conclusion appears in the fault-tolerant system. Besides, combining ADTL with lane changes may bring out the full value of ADTL.

A major shortcoming of the conventional car-following models is that these models only consider the current spacing and speeds of the target vehicle and its immediate leading vehicle, without taking into account prior driving actions, even for those from the same driver. In other words, the numerous prior experiences have no influence in predicting vehicular movements for the next time step. In this research, we propose a machine-learning-based data-driven methodology that is able to take advantage of the high-resolution historical traffic data in the current data-rich era, to predict vehicular movements in an accurate manner with high computational efficiency. The proposed car-following model has a simple model structure based on a fixed-radius near neighbors (FRNN) search algorithm and it can be applied to high-resolution, real-time vehicle movement prediction, modeling, and control. A comprehensive performance comparison is also conducted among the proposed car-following model, another similar data-driven model, and two conventional formula-based models. The results indicate that the FRNN algorithm-based car-following model is superior to all other three models in terms of prediction accuracy and is more computationally efficient compared to its data-driven-based counterpart. Some extensive applications of the proposed car-following model are also discussed at the end of this article.

As one of the most frequently occurred driving behaviors, the lane-changing’s impact on surrounding traffic is important. From the aspect of macro-traffic flows, its impact on traffic efficiency and safety has been investigated. However, as a non-essential lane-changing type, a single discretionary lane-changing’s impact on surrounding traffic is still difficult to be quantified by real-word traffic data. To solve this problem, this paper first proposes a trajectory data-based analytic method to calculate response time of the lane-changing vehicle and its following vehicles, then develops a time sequence data processing method based on feature points to analyze the change pattern of the variable minimum space, and then a lane-changing’s temporal and spatial on surrounding traffic can be captured by the decreasing-increasing pattern of minimum space. Besides, a high-precision trajectory data set, called Zen-traffic, is employed to demonstrate the effectiveness of the proposed methods and reach conclusions. Given the condition that the speed of a discretionary lane-changing vehicle is between 6 to 20 m/s, i.e., the traffic condition at which lane changes more likely occur, the results of the paper include (1) on average, 4 or 5 vehicles are usually impacted by a discretionary lane change; (2) the average impact time is 12 to 13 seconds and it has little relationship with the distance between the lane-changing vehicle and the following vehicle on the target/initial lane; (3) the lateral movement direction (i.e., changing to left or to right) of a lane change has a significant impact on the number of the vehicles affected.

The road-crossing of pedestrians at unsignalized crosswalks is a major concern for road safety. Previous studies focused on explaining of the mechanism underlying this behavior, but a framework of prediction is missing. To predict this behavior, only variables measured before the decision is made should be considered. To explore whether historical data is able to predict the behavior, this paper investigates pedestrians’ wait-or-go (WOG) behavior based on trajectory data and a machine learning method, both of which have been rarely applied by previous studies. The use of trajectory data enables the analysis of several influential factors related to moving characteristics, which are critical for pedestrians’ decision making. The framework based on machine learning, combined with trajectory data, achieves good explanatory power and predictability of pedestrians’ WOG behavior. Moreover, a possible application of this study is the prediction of pedestrian road-crossing intention in the context of autonomous cars.

This paper presents a multi-agent dynamic lane-changing (LC) trajectory planning method for CAV. In this method, a decision module is constructed by means of a potential field to determine the LC starting point. Then a series of trajectories is generated in the trajectory generation module. A cost function is constructed for searching for the corresponding optimal trajectory for both the subject vehicle and the participants. The simulation results indicate that the proposed model improves the LC success rate and reduces duration. Differing from the traditional model, we consider the cooperation feature of CAV’s LC and satisfy the subject vehicle’s demand as well as minimizing its impact on the other participants. Moreover, the driving environment including mesoscale information is considered to improve the LC success rate, which provides a new strategy for optimizing LC decision. Additionally, the method can also be applied to simulate CAVs’ LC behaviour.

This paper proposes an analyzable agent-based route choice modeling framework with good theoretical properties. This modeling framework allows heterogeneous individual learning rules and learning rates. As long as travelers' route choice behaviors conform to the framework, even though their learning rules and learning rates are heterogeneous, the network flows can be proven to be with asymptotically stable fixed points. An approximation for network flow distribution is proposed from the perspective of the stochastic process. Some phenomena observed in laboratory experiments are well captured by the agent-based framework. Many existing network-level day-to-day dynamic models can be regarded as special cases of the framework by setting the concrete learning rules and learning rates of the agents. Numerical simulations are used to show model properties. This study can deepen our understanding of the behavioral mechanism of individual-level day-to-day route choice and network-level day-to-day traffic flow dynamics.

ABSTRACTVehicle to anything (V2X), especially vehicle-to-vehicle (V2V) technology, is regarded as a promising technology to improve traffic efficiency and safety. Under V2V environment, drivers can communicate with each other, and obtain the real-time information of followers, besides of that from the preceding vehicle. According to the condition of sending a request of cooperative driving to the preceding vehicle, drivers are categorized into three sensitivity types, i.e. aggressive, normal and conservative. The mechanism of the driver's reaction to the backward request is also presented. Then, a safety rule-based cellular automaton (CA) model is developed, which involves bi-directional vehicular information. Finally, through numerical experiments, the influences of cooperative driving by V2V on traffic efficiency and safety are analyzed and concluded as follows. 1) The cooperative driving is conducive to improving the efficiency of vehicles to achieve their desired states when the speed constraint (of head vehicle) is serious; 2) The higher average danger coefficient is produced in the more congested vehicular fleet; 3) With regard to homogeneous vehicle fleets, more aggressive driving behaviors with V2V communication results in the lower safety performance, but can increase the traffic flux at synchronized flow and congested state; 4) As for heterogeneous vehicle fleets, the stimulation type, the sensitivity type, and the orders of sensitivity types all have an important effect in the safety performance.

AbstractIt is foreseeable that connected vehicles (CVs) and regular vehicles (RVs) will be operated together on roads in the upcoming decades. The appearance of CVs provides us an opportunity to improve traffic efficiency by guiding drivers to properly steer their CVs. To take advantage of this opportunity, this paper proposes two cooperative driving strategies for CVs that move in heterogeneous RV and CV traffic, namely, CDS-L and CDS-G. CDS-L is a local strategy that considers only the traffic surrounding the subject CVs. CDS-G is a global strategy that considers the conditions of all CVs on a road. The basis of the strategies is to incorporate the spatial and temporal averages of the driving status of surrounding CVs. A parameter that is a function of the traffic density is introduced to balance traffic efficiency and fluctuations. Simulation results demonstrate that the proposed strategies well suppress traffic fluctuations and thus promote traffic efficiency. The proposed strategies are dedicated to traffic congestion mitigation by using low-level configuration CVs.

Owing to the developments of wireless communication technologies, mobile health (M-Health) has been postulated as a promising means to improve healthcare quality and save lives in emergencies. However, using wireless communications in M-Health faces the following challenges. First, the wireless transmission may generate electromagnetic interference (EMI) and trigger critical malfunctions to the medical devices. Second, different types of M-Health applications and time-varying patient conditions require dynamic quality of service (QoS). Third, energy efficiency (EE) should be optimized to guarantee the reliability of M-Health services, considering the limitations of the existing battery technologies. To address these challenges, this paper investigates the joint channel and power resource allocation problem for vehicular M-Health communications. The problem is formulated as a mixed-integer nonlinear program (MINLP) to maximize the system EE with the consideration of EMI constraints on medical equipment and dynamic QoS requirements of medical users. To find possible solutions, we reformulate the MINLP problem by relaxing the integer variables and transforming the objective to convex forms. Based on the dual-decomposition method, we first obtain the optimal power allocation and then recover the channel allocation variables to integers. To satisfy the QoS requirements, we develop two channel allocation strategies, i.e., the QoS fulfillment strategy and the QoS compensation strategy. The swap-blocking allocation concept is introduced to guarantee the optimality of the obtained solutions. Simulation results show that the proposed resource allocation scheme improves the system EE and service satisfaction degree.

ABSTRACTThis paper validates the vehicle type-dependent car-following heterogeneity from micro- and macro-aspects by using Next Generation Simulation (NGSIM) trajectory data. Regarding the micro-viewpoint, traffic conflict analysis techniques and three surrogate safety indexes (i.e., time headway (TH), time to collision (TTC) and safety margin (SM)) are introduced to evaluate the safety performance during ‘steady’ car following. When driving or following a larger vehicle, Z-test results show that TH, TTC and SM all become higher significantly, that is, the driver’s accepted risk tends to be lower. From the macro-viewpoint, with the aggregated traffic data, a flexible Lagrangian fundamental diagram is calibrated by a bi-level optimization procedure. These calibrated curves illustrate that during close car following process with the same speed, the driver generally keeps a higher gap distance with the leader when following or driving a larger vehicle. Therefore, the consistent findings contribute to the conclusive results about the vehicle type-dependent car-following heterogeneity.

ABSTRACTA spatiotemporal speed contour (SSC, or time-space traffic) diagram that exhibits traffic dynamics in time and space is of importance in transportation research and applications. This paper empirically investigates the feasibility of using non-rectangular parallelogram (nRP) cells to construct the SSC diagram, given complete trajectory and probe vehicle data. Compared with the traditional way of using rectangular parallelogram (RP) cells, using nRP cells further considers the direction of backward-moving wave. The trajectory dataset provided by Next Generation Simulation is employed as the data source, and travel time is taken as a metric to measure the quality of the constructed SSC diagram. Various comparisons demonstrate that using nRP cells outperforms using RP cells, in particular when the traffic is congested and the cells are relatively large. Therefore, we recommend using nRP cells to construct the SSC diagram in future, which is as simple as the traditional way but more accurate particularly in terms of estimating travel time.

Driving behavior studies based on electroencephalography (EEG) have mostly investigated the relationship between various risky driving behaviors and brain activity, while only a few studies have discussed the relationship between ordinary driving behavior (drivers’ behavior in normal situations) and brain activity. To bridge the gap, we conducted a driving simulator experiment to collect data on ordinary driving behavior, including acceleration, space headway, speed, time headway, lane deviation, and amplitude of steering wheel movements. At the same time, the amplitude, log-transformed power (LTP), and power spectral density of EEG were extracted as EEG features. The quantitative relationships between ordinary driving behavior features and EEG features were investigated, where power spectrum analysis was performed to process EEG signals and Pearson correlation analysis was utilized for statistical analysis. The results indicated that ordinary driving behavior relates to all four brain regions, especially the temporal, occipital, and frontal regions. β-LTP was found to be most relevant to ordinary driving behavior. Furthermore, acceleration, speed, and space headway may have potential correlation with EEG features (e.g., β-LTP). These findings improve our understanding of the correlation between brain activity and driving behavior, and show potential for application in transportation safety, such as advanced driver assistance systems design.

ABSTRACTIn contrast to a large number of studies in the route choice behavior with absolute bounded rationality, little concern is laid on that with relative bounded rationality which is hypothetically suggested in this paper and can also be an indispensable component of the bounded rationality. This study is conducted to investigate travelers’ day-to-day rerouting behaviors with the above two bounded rationalities. Along this line, an absolute and a relative boundedly rational swapping process (abbreviated as ABRSP and RBRSP, respectively) model are proposed in the frame of nonlinear pairwise swapping rule. It is found that both ABRSP and RBRSP show satisfactory properties. The numerical results suggest that ABRSP and RBRSP share some general similarity in macroscopic convergence, while still showing obvious distinctions in swapping process, and RBRSP cannot be approximated or substituted by ABRSP or its extensions. Therefore, the suggested relative bounded rationality, even if it is hypothetical, still deserves to be taken seriously.

Imagine in the future that autonomous vehicles coordinated and guided by signal-free autonomous intersections are able to pass through the intersections immediately after the vehicles in the conflicting direction leave. Meanwhile, with the coordination of the autonomous intersections, autonomous vehicles on any approaching lane are able to turn onto any downstream lane, expecting that driving in such a road system, autonomous vehicles can reach their destinations without any on-road lane changes, and high traffic efficiency will be achieved as well as great traffic safety. To draw the picture in detail, this paper designs a signal-free autonomous intersection with all-direction turn lanes (ADTL) under the environment of autonomous vehicles, and proposes a conflict-avoidance-based approach to coordinate all approaching vehicles in different directions. Communicating with the approaching autonomous vehicles and utilizing the approach, the autonomous ADTL intersection is able to coordinate the approaching vehicles in all directions and guide them to safely and efficiently pass through the intersection. Two simulation scenarios are conducted in a road network with an isolated intersection composed of four three-lane arms. One scenario validates the collision-free design of the system, and the other shows that the designed ADTL intersection outperforms the conventional signal controlled intersection in terms of traffic efficiency, and is potentially better than the autonomous intersection with specific-direction turn lanes. The autonomous ADTL intersection can be an important basis for designing a future autonomous urban road traffic system.

Variable message signs (VMSs) provide important traffic information to help drivers travel better on a transportation network. The effectiveness of VMSs largely depends on the numbers and locations of VMSs in a transportation network. Although a few optimization models have been proposed to find candidate roads for locating VMSs, few have been devoted to developing algorithms that can be used in a real transportation network. A large amount of traffic data, such as traffic flow data, is widely available, collected by various means. Based on those traffic data, a sorting algorithm for a VMS location problem is presented in this paper. The algorithm gave a VMS location order rather than a location set. The proposed method divided roads into categories according to multilevel attributes and preferentially selected roads of a higher class with larger flows and more information and the minimal effect of existing VMS in a certain order to locate VMS. The proposed algorithm was analyzed and verified through a practical case on the Beijing, China, urban road network.

ABSTRACTThis study proposes an anisotropic continuum model derived from a simplified Helly's model. Its admissibility of a wide cluster solution is then analyzed by the nonlinear weak solution theory. After that, it is calibrated by a monkey algorithm with dynamic adaptation (DAMA) under two cases, and it is concluded that (1) DAMA makes the performance index and model parameters converge to their optimal values in limited iterations; (2) this calibrated model well reproduces the traffic congestions originating from a rubberneck effect within US101 site, besides of those propagated from the downstream boundary; (3) this proposed model with relations of Del Castillo and Benitez, or Greedshields gets better calibration and validation results; (4) the performance index is more sensitive to density-related parameters than to speed-related parameters. In summary, this anisotropic model calibrated by DAMA can well reproduce real traffic waves, and DAMA is a promising algorithm for the higher-order model calibration.

Variable message signs (VMSs) provide important traffic information to help drivers travel better on a transportation network. The effectiveness of VMSs largely depends on the numbers and locations of VMSs in a transportation network. Although a few optimization models have been proposed to find candidate roads for locating VMSs, few have been devoted to developing algorithms that can be used in a real transportation network. A large amount of traffic data, such as traffic flow data, is widely available, collected by various means. Based on those traffic data, a sorting algorithm for a VMS location problem is presented in this paper. The algorithm gave a VMS location order rather than a location set. The proposed method divided roads into categories according to multilevel attributes and preferentially selected roads of a higher class with larger flows and more information and the minimal effect of existing VMS in a certain order to locate VMS. The proposed algorithm was analyzed and verified through a practical case on the Beijing, China, urban road network.

The purpose of this paper is to improve the intelligence and universality of the classical method for gating control in the SCOOT system. First, we introduce a method to identify spillovers, and use the occupancy threshold for spillover recognition to trigger this special control logic. Second, the interrelationship of the traffic flows among adjacent traffic links is analyzed. Accordingly, we present an influence rate model for upstream links of the bottleneck link and a share ratio model for the downstream links. With known threshold values for the influence rate and share ratio, we propose a rule and process for selecting the intersections that should be included in the sub-area of the gating control. Third, we determine total capacity adjustments for the incoming and outgoing streams of bottleneck links. Under the measures, the queue can be dissipated to a permissible length within a given period of time. After that, the apportion models for the total adjustments among different paths and links are presented. Therefore, the correlation coefficients of the traffic flows are between the bottleneck link and the other links. Next, we ascertain the capacity decrements and increments for the gated and benefiting streams. The optimization schemes are defined so as to calculate splits for the gated and benefiting intersections. Finally, we evaluate the advanced method using a VISSIM simulation. The results show that a new control method brings significant and positive effects to the bottleneck link itself and to the entire test area.

This paper proposes a day-to-day route choice model based on travellers’ behavioural characteristics. The model is able to reflect travellers’ route choice in one day and describes the diversity of route choice behaviours. First, characteristics of the day- to-day route choice and travellers’ bounded rationality are analysed. A loop architecture based on the evolution of travellers’ daily choices is presented; deliberate route choice and habitual route choice in the architecture are discussed. A model is then proposed based on the theoretical analysis. In this model, travellers’ behavioural characteristics in the day-to-day route choice are considered; i.e., forgetting previous travel times, risk attitude, habit effect, traveller's amount of attention on the route choice being made, perception of actual travel time, etc. Results of a numerical experiment show that the tendencies of the simulated travellers’ choices are reasonable and similar with that in reality, and that the model is able to reproduce travellers’ route choice with more behavioural characteristics.

A stochastic extension of Newell's three-detector method is presented. The method predicts the traffic states at an intermediate location given boundary data from downstream and upstream detectors. The method presented takes into account day-to-day variations in the arrivals, sensor detection errors, and variability in the fundamental diagram parameters. The output is the probabilistic distribution of predicted cumulative counts, which can be used to obtain confidence bounds on any traffic variable. The method is tested with empirical data.

A stochastic extension of Newell's three-detector method is presented. The method predicts the traffic states at an intermediate location given boundary data from downstream and upstream detectors. The method presented takes into account day-to-day variations in the arrivals, sensor detection errors, and variability in the fundamental diagram parameters. The output is the probabilistic distribution of predicted cumulative counts, which can be used to obtain confidence bounds on any traffic variable. The method is tested with empirical data.