Search

In the past decade, the number of cars in China has significantly raised, but the traffic jam spree problem has brought great inconvenience to people's travel. Accurate and efficient traffic flow prediction, as the core of Intelligent Traffic System (ITS), can effectively solve the problems of traffic travel and management. The existing short-term traffic flow prediction researches mainly use the shallow model method, so they cannot fully reflect the traffic flow characteristics. Therefore, this paper proposed a short-term traffic flow prediction method based on one-dimensional convolution neural network and long short-term memory (1DCNN-LSTM). The spatial information in traffic data is obtained by 1DCNN, and then the time information in traffic data is obtained by LSTM. After that, the space-time features of the traffic flow are used as regression predictions, which are input into the Fully-Connected Layer. In the end, the corresponding prediction results of the current input are calculated. In the past, most of the researches are based on survey data or virtual data, lacking authenticity. In this paper, real data will be used for research. The data are provided by OpenITS open data platform. Finally, the proposed method is compared with other road forecasting models. The results show that the structure of 1DCNN-LSTM can further improve the prediction accuracy. [ABSTRACT FROM AUTHOR]

In actual transportation systems, the response time of drivers to the stimulus from the preceding vehicle varies at different speeds. The traditional car-following model cannot capture the heterogeneity of drivers for the response time is regarded as a constant. This paper designed a sigmoid function to describe the relation between the response time of drivers and the current speed. Then, a new hydrodynamic lattice model is developed by introducing the proposed nonlinear function. The model is analyzed by using Fourier series, and the linear stability criterion is derived. Numerical experiments are conducted in a ring road. Simulation results show that the evolution of density waves occurring in actual traffic is reproduced well. Finally, the model is calibrated and verified with the real data, and the simulation results are quantitatively consistent with the detector data. [ABSTRACT FROM AUTHOR]

Traffic status recognition and classification is an important prerequisite for traffic management and control. Based on the idea of weight optimal, a weighted fuzzy c-means clustering method for improving the accuracy of traffic classification is proposed in this study to ease traffic congestion. First, since there are many indexes that affect the traffic flow state classification, three commonly used indexes namely, volume, speed and occupancy are chosen as the main parameters for the traffic flow state classification in this paper. Second, in order to quantitatively analyze the influence degree of different traffic flow parameters on traffic flow state division, based on the principle of weight optimization, the objective function of weight optimization is established. Then the weight of each attribute index is obtained by using the branch and bound algorithm. Finally, since the traditional fuzzy c-means clustering method will not consider the influence of different traffic flow parameter weights on the traffic flow state classification results, the classification effect needs to be further improved. A fuzzy weighted c-means classification method which uses weighted Euclidean distance instead of Euclidean distance is proposed to classify the traffic flow states. Based on the same traffic flow data sample on the same road section, the traffic state classification results with different methods show that it is helpful to improve the traffic flow state classification accuracy by weighting the clustering index. Because the influence of different parameters on the traffic flow state classification is considered in the process of clustering, it is more conducive to improve the classification accuracy. Moreover, it can provide more accurate classification information for traffic control and decision making. [ABSTRACT FROM AUTHOR]

The difference between the optimal current difference and the actual current difference will be used as the correction item. The dynamic multiple current information about the front lattice will be considered. A modified lattice traffic hydrodynamics model is established by considering the downstream traffic conditions in the two-lane system. Through the stability analysis, it is found that the downstream traffic condition can be added as a correction term to increase the stability of the system. The area of the stable region on the phase diagram is enlarged by the derived stability. The mKdV equation, which can describe density wave, is derived by nonlinear analysis. Finally, the phase diagram of stability condition in linear analysis and the kink wave diagram of mKdV equation in nonlinear analysis are obtained by numerical simulation, which verifies the theoretical derivation of this paper. The results show that in the two-lane traffic flow expansion model, considering the downstream traffic conditions can effectively suppress traffic jams and make the traffic flow stable. [ABSTRACT FROM AUTHOR]

Traffic congestion is now nearly ubiquitous in many urban areas. The improvement of road infrastructure is an effective way to ease traffic congestion, especially the key road links. So, it is a fundamental and important step to identify the key link for improving transportation efficiency. However, most approaches in the current literature use simulated data and need many assumption conditions. The result shows the low comprehensibility and the bad exactitude. This paper provides a new identification method of key links for urban road traffic network (URTN) based on temporal-spatial distribution of traffic congestion. The method involves identifying congestion state, computing time distribution of congestion state and determining key road link. By the cluster analysis of the history field data of URTN, the threshold to determine the traffic congestion of each link can be obtained. Then the time-interval of the traffic congestion can be computed by median filtering. At last, the time-interval coverage is defined and used to determine the target road link whether it is a key road link or not. The method is validated by a real-world case (Beijing road traffic network, BRTN). The result shows the feasibility and accuracy. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new car-following model considering driver's timid and aggressive characteristics on a gradient highway. Based on the control theory, the linear stability analysis of the model was conducted. It shows that the stability of traffic flow on the gradient highway varies with the drivers' characteristics and the slope. Adopting nonlinear stability analysis, the Burgers equation and modified Korteweg–de Vries (mKdV) equation are derived to describe the triangular shock waves and kink–antikink waves, respectively. The theoretical and numerical results show that aggressive drivers tend to stabilize traffic flow but timid drivers tend to destabilize traffic flow on a gradient highway both on an uphill situation and on a downhill situation. Moreover, the slope of the road also plays an important role in traffic jamming transition. [ABSTRACT FROM AUTHOR]

In order to reveal the impact of preceding vehicle’s velocity on traffic flow, an extended car-following model considering preceding vehicle’s velocity feedback control is proposed in this paper. The linear stability criterion of the new model is derived through control theory method and it shows that the feedback control signal impacts the stability of traffic flow. Numerical simulation results is in good agreement with the theoretical analysis, which prove that a smaller negative feedback control of the preceding vehicle’s velocity can enhance the stability of traffic flow, while a smaller positive feedback control of the preceding vehicle’s velocity can exacerbate traffic congestion. Moreover, the reaction coefficients of straight and curved road conditions also play an important role in the stability of traffic flow. [ABSTRACT FROM AUTHOR]

Traffic congestion is now nearly ubiquitous in many urban areas and frequently occurs during rush hour periods. Rush hour avoidance is an effective way to ease traffic congestion. It is significant to calculate the rush hour for alleviating traffic congestion. This paper provides a method to calculate the fuzzy peak hour of the urban traffic network considering the flow, speed and occupancy. The process of calculation is based on betweenness centrality of network theory, optimal separation method, time period weighting, probability-possibility transformations and trapezoidal approximations of fuzzy numbers. The fuzzy peak hour of the urban road traffic network (URTN) is a trapezoidal fuzzy number [m1, m2, m3, m4]. It helps us (i) to confirm a more detailed traffic condition at each moment, (ii) to distinguish the five traffic states of the traffic network in one day, (iii) to analyze the characteristic of appearance and disappearance processes of the each traffic state and (iv) to find out the time pattern of residents travel in one city. [ABSTRACT FROM AUTHOR]

Connected vehicles are expected to become commercially available by the next decade, while traffic interruption is not uncommon in the real traffic environment. In this paper, we propose a feedback control method for lattice hydrodynamic model considering the traffic interruption probability effect. The stability criterion of the new model is explored through linear stability analysis of transfer function. When the stability conditions are not satisfied, a delay feedback controller is used to control the discharging flow to suppress traffic congestion. The impact of gain coefficient and delay time on the performance is discussed. We verify the effectiveness of the devised delay feedback controller by simulations. Results show that the traffic interruption probability effect has a considerable impact on the stability of traffic flow, while the controller is effective in suppressing traffic congestion. [ABSTRACT FROM AUTHOR]

With the increase of vehicles, traffic congestion has become a prominent problem. Environmental pollution, energy waste and even traffic accidents are easily caused by congestion. Nowadays, with the help of V2V and V2I communications, real-time interaction of vehicle information can be achieved. Vehicles can easily get information about other vehicles. In reality, a driver looks at the following car (i.e. upstream traffic condition) as well as the preceding car while driving. However, the effect of upstream condition is rarely analyzed. In the paper, a term describing the upstream traffic condition is added in Nagatani's model. The effect of the upstream traffic condition to linear stability is obtained by applying the perturbation method. Nonlinear stability equations, such as the mKdV equation, are also derived. Theoretical results of linear stability and nonlinear stability are shown intuitively through numerical simulations. It can be found that traffic is more stable when the upstream traffic condition is considered. [ABSTRACT FROM AUTHOR]

Relieving complete congestion in a traffic system is an important problem. We propose a strategy to realize this, in which the packets on nodes shared by many shortest paths are dropped preferentially. A simple scale-free network is chosen to demonstrate the importance of the degree heterogeneity to the congestion problem, though this network structure cannot mimic a real traffic network. Two traffic models are simulated: in one of which, all the nodes are identical, and in the other, the delivering capacity and storing ability for each node are both proportional to its degree. Both models can give a phase transition between free-flow and congested states, while the latter model has significant strong transportation performance (a larger critical value of the packet generation rate). The strategy of preferentially dropping packets on nodes shared by many shortest paths, as proposed in this paper, can realize remarkably better transportation performance measured by the fraction of congested nodes and the average arrival rate compared with the random packet dropping strategy in the literature. [ABSTRACT FROM AUTHOR]

High quality bus service is considered as an efficient way to mitigate traffic congestion in big cities. Global positioning system (GPS) data provide sufficient sources to evaluate the performance of bus vehicles that both passengers and operator concern about. This paper aims to propose a framework to assess the operational performance of bus routes based on the GPS trajectory data collected from Jinan, China. Several important indicators of bus operation including travel time of routes, section running time, dwell time and bus bunching have been studied. The results show that the travel time of routes follow right skewed distributions. Moreover, section running time between two consecutive stations varies in different time period and it is larger in evening peak hours. Additionally, the dwell time has been discussed and the results show that there is no big variation in most stations except some stations, which provides a help to identify the key stations. Furthermore, we propose an approach to detect the bunching points. The results indicate the bunching points are easy to occur in the peak hours and the congested road section. [ABSTRACT FROM AUTHOR]

Traffic congestion has attracted considerable attention in modern society. Many researchers have proposed tremendous traffic flow models. These models can be divided into microscopic model, mesoscopic model and macroscopic model. With the progress of technology, it is possible to get the information of the preceding vehicles, i.e. the downstream traffic condition. Drivers can adjust their driving behaviors according to the downstream condition. This mechanism has been considered in microscopic model, and it is found that this mechanism can not only stabilize the traffic, but also bring about some new phenomena, such as synchronized traffic pattern. In the macroscopic model, what will the traffic flow be after considering the mechanism? In this paper, a term describing the downstream traffic condition is added in Nagatani's single-lane lattice model. Through theoretical study and numerical simulation, it can be found that traffic is more stable when the downstream traffic condition is considered. Besides, the modified model can reproduce the soliton wave, kink wave properly. Moreover, the amplitude and frequency of density wave with different influence forms of the front lattices on the current lattice are almost the same as found to reflect the different influence forms do not have significant effect on the stability. [ABSTRACT FROM AUTHOR]

The traffic during rush hour in metropolises is always in heavy congestion, so the wagon trucks of logistic transportation are mostly forbidden during rush hour. In fact, quite a few logistic transportations are fulfilled by converted passenger vehicles. This paper aims to model the logistic companies' behaviors and see through the effect of their cooperation modes on the companies and on the whole traffic conditions. The logistic companies decide their transportation pricing, warehouse using and route choosing separately or as a group (cooperation) to compete with others. It is also assumed that the heavy freight vehicles, light freight vehicles and the auto vehicles share the same network. Three models based on variational inequalities are proposed to describe the companies' behaviors when there is no cooperation, simple cooperation and deep cooperation, respectively, and the modified diagonalization method is used to solve these models. The total transportation demand is elastic, which depends on the users' choice of logistics company and further depends on all the companies' prices, warehouse and route choices. The goodness of different coalitions to the companies and the whole traffic system is well investigated in the numerical examples. [ABSTRACT FROM AUTHOR]

The existing car-following models of connected vehicles commonly lack experimental data as evidence. In this paper, a Gray correlation analysis is conducted to explore the change in driving behavior with safety messages. The data mining analysis shows that the dominant factor of car-following behavior is headway with no safety message, whereas the velocity difference between the leading and following vehicle becomes the dominant factor when warning messages are received. According to this result, an extended car-following model considering the impact of safety messages (IOSM) is proposed based on the full velocity difference (FVD) model. The stability criterion of this new model is then obtained through a linear stability analysis. Finally, numerical simulations are performed to verify the theoretical analysis results. Both analytical and simulation results show that traffic congestion can be suppressed by safety messages. However, the IOSM model is slightly less stable than the FVD model if the average headway in traffic flow is approximately 14–20 m. [ABSTRACT FROM AUTHOR]

Considering the effect of headway changes with memory, an improved continuum model of traffic flow is proposed in this paper. By means of linear stability theory, the new model's linear stability with the effect of headway changes with memory is obtained. Through nonlinear analysis, the KdV-Burgers equation is derived to describe the propagating behavior of traffic density wave near the neutral stability line. Numerical simulation is carried out to study the improved traffic flow model, which explores how the headway changes with memory affected each car's velocity, density and energy consumption. Numerical results show that when considering the effects of headway changes with memory, the traffic jams can be suppressed efficiently. Furthermore, research results demonstrate that the effect of headway changes with memory can avoid the disadvantage of historical information, which will improve the stability of traffic flow and minimize car energy consumption. [ABSTRACT FROM AUTHOR]

The traffic delay at intersections is crucial for the performance of urban traffic system. This paper analytically investigated the minimum traffic delay at a three-phase T-type intersection. We firstly demonstrate that the minimum traffic delay must be achieved on the surface of the 3D space constituted by the three constraints. Next, we prove that the minimum traffic delay must be achieved on the three borderlines of the surface. Finally, we show that the minimum delay is achieved either on one specific borderline or at the vertex of the surface. In the former case, extra green time is needed for the stream with largest demand, while no extra green time should be assigned to any stream in the latter case. [ABSTRACT FROM AUTHOR]

Various car-following models (CMs) have been developed to capture the complex characteristics of microscopic traffic flow, among which the coupled map CM can better reveal and reflect various phenomena of practical traffic flow. Capacity change at bottleneck contributes to high-density traffic flow upstream the bottleneck and contains very complex dynamic behavior. In this paper, we analyze the effect of speed bottleneck on the spatial-temporal evolution characteristics of traffic flow, and propose a method to reduce traffic congestion with the feedback control signal based on CM. Simulation results highlight the potential of using the feedback signal to control the stop-and-go wave and furthermore to alleviate the traffic congestion effectively. [ABSTRACT FROM AUTHOR]

A modified lattice hydrodynamic model is proposed by considering the driver's sensory memory and the average optimal velocity effect field. The stability conditions of the novel model are further analyzed theoretically through the linear analysis. The nonlinear modified Korteweg–de Vries (mKdV) equation near the critical point is obtained, which can describe the jamming transition of traffic flow properly. Numerical simulations for the novel model are carried out and the results validate that the traffic jam can be suppressed efficiently by considering the average optimal velocity effect field and driver's sensory memory. Besides, the energy consumption simulation is devised to investigate the stability of the traffic system. Eventually, PMES data is adopted to calibrate and evaluate the parameters of the proposed model, which proves that it precisely reflects the evolution of traffic flow. All the simulation results verify the feasibility and validity of this model. [ABSTRACT FROM AUTHOR]

The network topology and the routing strategy are major factors to affect the traffic dynamics of the network. In this work, we aim to design an optimal time-varying network structure and an efficient route is allocated to each user in the network. The network topology is designed by considering addition, removal and rewiring of links. At each time instant, a new node connects with an existing node based on the degree and correlation with its neighbor. Traffic congestion is handled by rewiring of some congested links along with the removal of the anti-preferential and correlated links. Centrality plays an important role to find the most important node in the network. The more a node is central, the more it can be used for the shortest route of the user pairs and it can be congested due to a large number of data coming from its neighborhood. Therefore, routes of the users are selected such that the sum of the centrality of the nodes appearing in the user's route is minimum. Thereafter, we analyze the network structure by using various network properties such as the clustering coefficient, centrality, average shortest path, rich club coefficient, average packet travel time and order parameter. [ABSTRACT FROM AUTHOR]

Robustness of infrastructure networks is essential for our modern society. Cascading failures are the cause of most large-scale network outages. We study the cascading failure of networks due to overload, using the betweenness centrality of an edge as the measure of its initial load. Taking into account the congestion effect of a slightly overloading edge, we define two capacities (the basic capacity and the removal capacity) of every edge and give three possible states (the free state, the congestion state, and the removal state) of every edge according to its current load. We propose a new method to dynamically adjust two capacities of the slightly overloading edge and study the dynamical features of cascading propagation induced by removing the edge with the highest load in two artificial networks, two traffic networks, and two power grids. We mainly focus on the relationship between the capacity parameters and two robust metrics. By simulations, we find two interesting and counterintuitive results, i.e. enhancing the basic capacity of every edge may weaken the network robustness, and fixing the basic capacity of every edge, simply improving the removal capacity of every edge sometimes makes the whole network more invulnerable. These findings show that investing more maintenance resources to alleviate flow congestion is not always better to avoid the cascading propagation, which is similar to Braess's paradox in traffic networks. [ABSTRACT FROM AUTHOR]