Search

Managing traffic flow in high-occupancy toll (HOT) lanes is a tough balancing act and current tolling schemes often lead to either under- or over-utilization of HOT lane capacity. The inherent linear/nonlinear relationship between flow and tolls in HOT lanes suggest that recent advances in machine learning and the use of a data-driven model may help set toll rates for optimal flow and lane use. In this research project, a data-driven model was developed, using long short-term memory (LSTM) neural networks to capture the underlying flow-toll pattern on both HOT and general-purpose lanes. Then, a dynamic control strategy, using linear quadratic regulator (LQR) feedback controller was implemented to fully utilize the HOT lane capacity while maintaining congestion-free conditions. A case study of the I-580 freeway in Alameda County, California was carried out. The control system was evaluated in terms of vehicle hours traveled and person hours traveled for solo drivers and carpoolers. Results show that the tolling strategy helps to mitigate congestion in HOT and general-purpose lanes, benefiting every traveler on I-580.

Traffic simulations are valuable tools for urban mobility planning and operation, particularly in large cities. Simulation-based microscopic models have enabled traffic engineers to understand local transit and transport behaviors more deeply and manage urban mobility. However, for the simulations to be effective, the transport network and user behavior parameters must be calibrated to mirror real scenarios. In general, calibration is performed manually by traffic engineers who use their knowledge and experience to adjust the parameters of the simulator. Unfortunately, there is still no systematic and automatic process for calibrating traffic simulation networks, although some methods have been proposed in the literature. This study proposes a methodology that facilitates the calibration process, where an artificial neural network (ANN) is trained to learn the behavior of the transport network of interest. The ANN used is the Multi-Layer Perceptron (MLP), trained with back-propagation methods. Based on this learning procedure, the neural network can select the optimized values of the simulation parameters that best mimic the traffic conditions of interest. Experiments considered two microscopic models of traffic and two psychophysical models (Wiedemann 74 and Wiedemann 99). The microscopic traffic models are located in the metropolitan region of São Paulo, Brazil. Moreover, we tested the different configurations of the MLP (layers and numbers of neurons) as well as several variations of the backpropagation training method: Stochastic Gradient Descent (SGD), Adam, Adagrad, Adadelta, Adamax, and Nadam. The results of the experiments show that the proposed methodology is accurate and efficient, leading to calibration with a correlation coefficient greater than 0.8, when the calibrated parameters generate more visible effects on the road network, such as travel times, vehicle counts, and average speeds. For the psychophysical parameters, in the most simplified model (W74), the correlation coefficient was greater than 0.7. The advantage of using ANN for the automatic calibration of simulation parameters is that it allows traffic engineers to carry out comprehensive studies on a large number of future scenarios, such as at different times of the day, as well as on different days of the week and months of the year. [ABSTRACT FROM AUTHOR]

Community-based carpooling has the potential to alleviate traffic congestion and reduce the transportation carbon footprint. Once technology, communication, demographic, and economic barriers are overcome, community-based carpooling can be fully exploited. One of the major barriers to implementation is the difficulty of optimizing carpool formation in large systems. This study utilizes two different methods to solve the carpooling optimization problem: 1) bipartite algorithm and 2) integer linear programming. The bipartite method determines the maximum number of carpooling pairs given acceptable reroute costs and travel delays. The linear programming method defines the most optimal performance that minimizes the most vehicular travel mileage. These two methods are carefully compared to evaluate the carpooling potentials among single-occupancy vehicles based on the output of activity-based model’s (ARC ABM) home-to-work single-occupancy vehicle (SOV) trips that can be paired together towards designated regional employment centers. The experiment showed that under strict assumptions, an upper bound of around 13.6% of such trips could carpool together. The results are promising in terms of higher-than-anticipated carpool match rates and the predicted decrease in total vehicle mileage. Moreover, the framework is flexible enough with the potential to act as a simulation testbed, to optimize vehicular operations, and to match potential carpool partners in real-time.View the NCST Project Webpage

Volume-Delay Function (VDF) is a key input in several tasks including traffic assignment and trip distribution estimation. However, there is no clear framework to identify the VDF formula and parameters. Furthermore, the VDF performance in the oversaturated condition is uncontrolled. In this paper, most VDF-related aspects are investigated, including the key attributes, traffic model and VDF formulas, oversaturated condition assumptions, and optimization techniques. In each aspect, different methods are compared for best performance, while a new procedure is introduced, if necessary. The best traffic model and VDF formulas are nominated to satisfy a set of logic and statistical conditions, while a new optimization technique is presented. Furthermore, a new realistic assumption to control the VDF performance is discussed. This framework is applied on 32 worldwide roads of different categories and characteristics, which indicates accurate and realistic results. The results encourage for further investigation of the VDF performance in future work.

We study a 1D scalar conservation law whose non-local flux has a single spatial discontinuity. This model is intended to describe traffic flow on a road with rough conditions. We approximate the problem through an upwind-type numerical scheme and provide compactness estimates for the sequence of approximate solutions. Then, we prove the existence and the uniqueness of entropy weak solutions. Numerical simulations corroborate the theoretical results and the limit model as the kernel support tends to zero is numerically investigated. [ABSTRACT FROM AUTHOR]

Parking has long been an urban planning challenge. Providing parking in city centers is land-intensive and expensive. Moreover, drivers searching for scarce parking can increase congestion, vehicle miles traveled (VMT), and greenhouse gas (GHG) emissions. Use of automated vehicles to drop off and pick up travelers could reduce the demand for parking, which could reduce VMT and associated emissions and allow urban spaces currently used for parking to be converted to more beneficial uses. However, automated vehicles could also have negative consequences. They could generate empty vehicle travel and more cross-traffic movements due to drop-offs and pick-ups which could increase congestion, VMT, and GHG emissions.Researchers at the University of California, Davis modeled the travel effects of changes in drop-off and pick-up activity and parking supply that might be triggered by widespread automated vehicle use in San Francisco’s city center. A primary goal of this research was to determine an optimal level of automated vehicle adoption that minimizes negative consequences. The researchers also modeled methods to control these negative consequences, including expanding drop-off and pick-up zones and imposing auto pricing policies to curb demand. This policy brief summarizes the findings from that research and provides policy implications.View the NCST Project Webpage

The potential for automated vehicles (AVs) to reduce parking in city centers has generated much excitement among urban planners. AVs could drop-off (DO) and pick-up (PU) passengers in areas where parking costs are high: personal AVs could return home or park in less expensive locations, and shared AVs could serve other passengers. Reduced on-street and off-street parking present numerous opportunities for redevelopment that could improve the livability of cities, for example, more street and sidewalk space for pedestrian and bicycle travel. However, reduced demand for parking would be accompanied by increased demand for curbside DO/PU space with related movements to enter and exit the flow of traffic. This change could be particularly challenging for traffic flows in downtown urban areas during peak hours, where high volumes of DOs and PUs are likely to occur. Only limited research examines the travel effects of a shift from parking to DO/PU travel and the impact of changes in parking supply. This study uses a microscopic road traffic model with local travel activity data to simulate personal AV parking scenarios in San Francisco's downtown central business district. These scenarios vary (1) the demand for DO and PU travel versus parking, (2) the supply of on-street and off-street parking, and (3) the total demand for parking and DO/PU travel due to an increase in the cost of travel to the central business district.View the NCST Project Webpage

Nowadays, the effectiveness of any smart transportation management or control strategy would heavily depend on reliable traffic data collected by sensors. Two problems regarding sensor data quality have received attention: first, the problem of identifying malfunctioning sensors; second, reconstruction of traffic flow. Most existing studies concerned about identifying completely malfunctioning sensors whose data should be discarded. This project focuses on the problem of error detection and data recovery of partially malfunctioning sensors that could provide valuable information. By integrating a sensor measurement error model and a transportation network model, the authors propose a Generalized Method of Moments (GMM) based estimation approach to determine the parameters of systematic and random errors of traffic sensors in a road network. The proposed method allows flexible data aggregation that ameliorates identification and accuracy. The estimates regarding both systematic and random errors are utilized to conduct hypothesis test on sensor health and to estimate true traffic flows with observed counts. The results of three network examples with different scales demonstrate the applicability of the proposed method in a large variety of scenarios. This research improves fundamental knowledge on transportation data analytics as well as the effective management of data and information infrastructure in transportation practice.View the NCST Project Webpage

Traffic during peak hours is getting worse over time and the duration of the peak is increasing in most metropolitan areas as more drivers try to use limited roadway capacity. Bottlenecks caused by traffic incidents or road construction limit roadway capacity even further and can cause traffic “shock waves.” When an incident causes a highway lane to close unexpectedly, vehicles are forced to change lanes close to the incident and at low speeds. These forced lane changes interfere with traffic flow in open lanes and decrease the overall flow of the roadway. Heavy-duty trucks can exacerbate congestion because they are larger and slower than passenger vehicles. Advanced technologies may help to improve traffic flow in these situations. Variable speed limits can change based on road, traffic, and weather conditions. Speed limits can be reduced in real time when congestion is imminent to smooth traffic flow and handle more traffic volume at a slower, but not stop-and-go, speed. Lane change control systems provide lane change recommendations well upstream of blocked lanes, spreading lane changes over a greater distance and minimizing bottlenecks that disrupt traffic flow.This policy brief summarizes findings from researchers at the University of Southern California who simulated traffic patterns along a section of Interstate 710 near the Ports of Long Beach/Los Angeles, a congested area that gets substantial truck traffic. They simulated the use of variable speed limit and lane change control systems to evaluate the potential traffic impacts of these systems.This brief is based on research from two NCST projects: Eco-Friendly Intelligent Transportation System Technology for Freight Vehicles, and Reducing Truck Emissions and Improving Truck Fuel Economy via ITS Technologies.

Numerical simulations have shown that the network fundamental diagram (NFD) of a signalized network is significantly affected by the green ratio. An analytical approximation of the NFD has been derived from the link transmission model. However, the consistency between these approaches has not been established, and the impacts of other factors are still unrevealed. This research evalutes the impacts of start-up and clearance behaviors in a signalized network from a network fundamental diagram approach. Microscopic simulations based on Newell’s car-following model are used for testing the bounded acceleration (start-up) and aggressiveness (clearance) effects on the shape of the NFD in a signalized ring road. This new approach is shown to be consistent with theoretical results from the link transmission model, when the acceleration is unbounded and vehicles have the most aggressive clearance behaviors. This consistency validates both approaches; but the link transmission model cannot be easily extended to incorporate more realistic start-up or clearance behaviors. With the new approach, this project demonstrates that both bounded acceleration and different aggressiveness lead to distinct network capacities and fundamental diagrams. In particular, they lead to start-up and clearance lost times of several seconds; and these lost times are additive. Therefore, the important role that these behaviors play in the NFD shape is studied to reach a better understanding of how the NFD responds to changes. This will help with designing better start-up and clearance behaviors for connected and autonomous vehicles.

Abstract A system to obtain vehicular energy‐use and emissions in transient traffic is proposed. It operates at a microscopic 1‐s time scale, capturing critical stop‐and‐go dynamics in a controlled and repeatable computational environment. This allowed the definition of a sensitivity function linking energy‐use to traffic flow, providing an additional comparison parameter. To obtain the dynamics data, crucial for transient traffic modeling implementation and validation, a video‐tracking technique is used to overcome velocimeter and GPS data shortcomings. This approach allowed to observe and reproduce a traffic scene and agreed with a M/D/n queuing traffic model. Once the system contains precise dynamics, it can be implemented for traffic scenarios such as intersections, school areas, or future road projects. Any simulated vehicular technology can be incorporated into the analysis. A case study was implemented using 1000 electric (EV) and combustion engine vehicles (CV) interacting with trucks, buses, and motorcycles, while monitoring individual vehicular systems. These included the powertrain and regenerative braking to obtain energy‐use and emissions. EVs proved to be not only more efficient, but less sensitive to traffic buildup than CVs. The techniques presented here, lend themselves to technological design and evaluation and allows testing real‐life vehicular systems for sustainable traffic and energy planning.

The Los Angeles/Long Beach area is important for freight as it involves the twin ports and warehouses and freight hubs. The way freight is consolidated and distributed affects what is going on within the terminals and roadway and rail networks. The complexity and dynamics of the multimodal transportation networks in Los Angeles/Long Beach region that are also shared by passengers, together with the unpredictability of the effect of incidents, disruptions and demand, in temporal and special coordinates makes the local impact analysis of freight transportation a very challenging task despite recent advances in information technologies.Under this project, the researchers developed a set of traffic simulation models for the Los Angeles/Long Beach region that allowed them to evaluate the impact of new traffic flow control systems, vehicle routing, policy interventions such as land use changes and other ITS technologies on the efficiency of the transportation system and on the environment. The developed simulation models include: macroscopic simulation model for studying and evaluating large traffic networks, and microscopic simulation model for smaller networks. The macroscopic model focusses on flows and covers a much larger area as it is computationally much more efficient than the microscopic one. The microscopic model models the motion of each truck and vehicle, traffic lights, stop signs, speed limits, traffic rules etc. and resembles the real situation as close as possible.The developed simulation models have been used to evaluate different systems and application scenarios, including freight priority traffic signal control, multimodal freight routing and the impact analysis of the spatial pattern changes of warehousing and distribution.View the NCST Project Webpage

Since heavy-duty truck operations can significantly affect traffic congestion, especially on road grade, the creation of exclusive lanes for trucks has been viewed as a potential alternative to reduce congestion delay, fuel consumption, and emissions. However, few studies have rigorously evaluated the effectiveness of truck-only lanes in achieving these benefits. This study demonstrates a model framework that combines a microscopic traffic simulation with emissions and microscale dispersion models to quantify the potential impacts of truck-only lanes on fuel consumption, emissions, and near-road pollutant concentrations. As a case study, the framework was used to evaluate a proposed $2 billion project to construct 40-miles of truck-only lanes on Interstate 75 (I-75) between Atlanta and Macon, Georgia (USA).The findings of this study suggest that truck-only lanes could significantly improve the traffic flow, and reduce energy, emissions, and pollutant concentrations. The research team expects that the extensive simulation results of this study help to understand the performance of truck-only lanes on a large-scale network with a heavy mixture of truck and general purpose lane traffic. The methodology and framework developed in this study can be effectively and efficiently applied to a wide variety of scenarios to evaluate the environmental impacts of other transportation projects under various conditions.View the NCST Project Webpage

Traffic simulations are valuable tools for urban mobility planning and operation, particularly in large cities. Simulation-based microscopic models have enabled traffic engineers to understand local transit and transport behaviors more deeply and manage urban mobility. However, for the simulations to be effective, the transport network and user behavior parameters must be calibrated to mirror real scenarios. In general, calibration is performed manually by traffic engineers who use their knowledge and experience to adjust the parameters of the simulator. Unfortunately, there is still no systematic and automatic process for calibrating traffic simulation networks, although some methods have been proposed in the literature. This study proposes a methodology that facilitates the calibration process, where an artificial neural network (ANN) is trained to learn the behavior of the transport network of interest. The ANN used is the Multi-Layer Perceptron (MLP), trained with back-propagation methods. Based on this learning procedure, the neural network can select the optimized values of the simulation parameters that best mimic the traffic conditions of interest. Experiments considered two microscopic models of traffic and two psychophysical models (Wiedemann 74 and Wiedemann 99). The microscopic traffic models are located in the metropolitan region of São Paulo, Brazil. Moreover, we tested the different configurations of the MLP (layers and numbers of neurons) as well as several variations of the backpropagation training method: Stochastic Gradient Descent (SGD), Adam, Adagrad, Adadelta, Adamax, and Nadam. The results of the experiments show that the proposed methodology is accurate and efficient, leading to calibration with a correlation coefficient greater than 0.8, when the calibrated parameters generate more visible effects on the road network, such as travel times, vehicle counts, and average speeds. For the psychophysical parameters, in the most simplified model (W74), the correlation coefficient was greater than 0.7. The advantage of using ANN for the automatic calibration of simulation parameters is that it allows traffic engineers to carry out comprehensive studies on a large number of future scenarios, such as at different times of the day, as well as on different days of the week and months of the year.

Optimal cordon-metering rates are obtained using Macroscopic Fundamental Diagrams in combination with flow conservation laws. A model-predictive control algorithm is also used so that time-varying metering rates are generated based on their forecasted impacts. Our scalable algorithm can do this for an arbitrary number of cordoned neighborhoods within a city. Unlike its predecessors, the proposed model accounts for the constraining effects that cordon queues impose on a neighborhood's circulating traffic. It does so at every time step by approximating a neighborhood's street space occupied by cordon queues, and re-scaling the MFD downward to describe the state of circulating traffic that results. The model is also unique in that it differentiates between saturated and under-saturated cordon-metering operations. Computer simulations show that these enhancements can substantially improve the predictions of both, the trip completion rates in a neighborhood and the rates that vehicles cross metered cordons. Optimal metering policies generated as a result are similarly shown to do a better job in reducing the Vehicle Hours Traveled in a city. The VHT reductions stemming from the proposed model and from its predecessors differed by as much as 18%.

Traffic congestion at arterial intersections and freeway bottlenecks degrades the air quality and threatens the public health. Conventionally, air pollutants are monitored by sparsely distributed Quality Assurance Air Monitoring Sites. Sparse mobile crowd-sourced data, such as cellular network and Global Positioning System (GPS) data, contain large amount of traffic information, but have low sampling rate and penetration rate due to the cost limit on data transmission and archiving. The sparse mobile data provide a supplement or alternative approach to evaluate the environmental impact of traffic congestion. This research establishes a framework for traffic-related air pollution evaluation using sparse mobile data and traffic volume data from California Performance Measurement System (PeMS) and Los Angeles Department of Transportation (LADOT). The proposed framework integrates traffic state model, emission model and dispersion model. An effective tool is developed to evaluate the environmental impact of traffic congestion for both arterials and freeways in an accurate, timely and economic way. The proposed methods have good performance in estimating monthly peak hour fine particulate matter (PM 2.5) concentration, with error of 2 ug/m3 from the measurement from monitor sites. The estimated spatial distribution of annual PM 2.5 concentration also matches well with the concentration map from California Communities Environmental Health Screening Tool (CalEnviroScreen), but with higher resolution. The proposed system will help transportation operators and public health officials alleviate the risk of air pollution, and can serve as a platform for the development of other potential applications.View the NCST Project Webpage

This report is one of two NCST Research Report documents produced as part of a project to advance the technical modeling tools for resiliency and adaptation planning, especially those used for criticality rankings. The official final technical report, Climate Adaptation and Resiliency Planning: Agency Roles and Workforce Development Needs, summarizes a climate adaptation framework and describes current planning practices and workforce needs of Departments of Transportation and other planning agencies. This additional technical documentation report looks specifically at the network data challenges of objectively assessing asset criticality, one step in the larger adaptation planning framework and a prerequisite for efficient allocation of limited adaptation resources. Specifically, this report explores the modeling resolution (in terms of the completeness of the road network and the spatial disaggregation of origin and destination matrices) needed to produce accurate criticality ratings. Original modeling work using a well-establish criticality measure, the Network Robustness Index (NRI), on both a small hypothetical network and the road network for Chittenden County, Vermont, demonstrated a need for higher resolution networks for criticality modeling. Since this part of the work was published in the Transportation Research Record it is only summarized here. A conceptual discussion of methods explored for creating networks for larger real-world areas that are sufficiently complete for criticality assessment is also included based on exploratory work using the travel demand model for the Greater Sacramento California area.View the NCST Project Webpage

A system to obtain vehicular energy‐use and emissions in transient traffic is proposed. It operates at a microscopic 1‐s time scale, capturing critical stop‐and‐go dynamics in a controlled and repeatable computational environment. This allowed the definition of a sensitivity function linking energy‐use to traffic flow, providing an additional comparison parameter. To obtain the dynamics data, crucial for transient traffic modeling implementation and validation, a video‐tracking technique is used to overcome velocimeter and GPS data shortcomings. This approach allowed to observe and reproduce a traffic scene and agreed with a M/D/n queuing traffic model. Once the system contains precise dynamics, it can be implemented for traffic scenarios such as intersections, school areas, or future road projects. Any simulated vehicular technology can be incorporated into the analysis. A case study was implemented using 1000 electric (EV) and combustion engine vehicles (CV) interacting with trucks, buses, and motorcycles, while monitoring individual vehicular systems. These included the powertrain and regenerative braking to obtain energy‐use and emissions. EVs proved to be not only more efficient, but less sensitive to traffic buildup than CVs. The techniques presented here, lend themselves to technological design and evaluation and allows testing real‐life vehicular systems for sustainable traffic and energy planning. [ABSTRACT FROM AUTHOR]

We focus on entropy admissible solutions of scalar conservation laws in one space dimension and establish new regularity results with respect to time. First, we assume that the flux function f is strictly convex and show that, for every x ∈ ℝ, the total variation of the composite function f ο u(∙, x) is controlled by the total variation of the initial datum. Next, we assume that f is monotone and, under no convexity assumption, we show that, for every x, the total variation of the left and the right trace u(∙,x±) is controlled by the total variation of the initial datum. We also exhibit a counterexample showing that in the first result the total variation bound does not extend to the function u, or equivalently that in the second result we cannot drop the monotonicity assumption. We then discuss applications to a source-destination model for traffic flows on road networks. We introduce a new approach, based on the analysis of transport equations with irregular coefficients, and, under the assumption that the network only contains so-called T-junctions, we establish existence and uniqueness results for merely bounded data in the class of solutions where the traffic is not congested. Our assumptions on the network and the traffic congestion are basically necessary to obtain well-posedness in view of a counterexample due to Bressan and Yu. We also establish stability and propagation of BV regularity, and this is again interesting in view of recent counterexamples. [ABSTRACT FROM AUTHOR]

In this paper, we derive second order hydrodynamic traffic models from kinetic-controlled equations for driver-assist vehicles. At the vehicle level we take into account two main control strategies synthesising the action of adaptive cruise controls and cooperative adaptive cruise controls. The resulting macroscopic dynamics fulfil the anisotropy condition introduced in the celebrated Aw–Rascle–Zhang model. Unlike other models based on heuristic arguments, our approach unveils the main physical aspects behind frequently used hydrodynamic traffic models and justifies the structure of the resulting macroscopic equations incorporating driver-assist vehicles. Numerical insights show that the presence of driver-assist vehicles produces an aggregate homogenisation of the mean flow speed, which may also be steered towards a suitable desired speed in such a way that optimal flows and traffic stabilisation are reached. [ABSTRACT FROM AUTHOR]

A signalized intersection regulated by pre-timed control in an urban traffic network is considered. A method for computing the mean delay of vehicles based on the theory of queues with server vacations is adapted from the existing literature and an algorithm for numerical implementation is proposed. Arrivals are assumed to occur according to a batch Markovian arrival process (BMAP). The proposed algorithm is an alternative to the previous methods that includes a numerical improvement in the computation of the queue invariant probabilities. In the case of isolated intersections we compare the new algorithm with other traffic models, including simulation model, Webster model and the Highway Capacity Manual model. The mean delay of vehicles given by the new algorithm is a short distance away from the mean delay obtained in the simulations. The approximation for large flows given by the proposed algorithm is as good as the Highway Capacity Manual model and with no need to tune unknown parameters. Our results are of interest for traffic engineers, since other approaches have been subject to criticism when flows near capacity are involved. [ABSTRACT FROM AUTHOR]

A dynamical system is studied. This system belongs to the class of contour networks introduced by A.P. Buslaev. The system is a version of the system called an open chain of contours. Continuous and discrete versions of the system were considered earlier. The system contains N contours. There is one adjacent contour for the utmost left and the utmost right contour, and there are two adjacent contours for any other contour. There is a common point of any two adjacent contours. This point is called a node. There is a cluster in each contour. For the continuous version, the cluster is a segment, moving with constant velocity if there is no delay. For the discrete version, the cluster is a group of adjacent particles. Delays are due to that two clusters may not move through the same node simultaneously. The main system characteristic, studied earlier, is the average velocity of clusters. In this paper, the continious and discrete version of the open chain of contours are considered. The following version of the system is also considered. Each cluster is a segment, and the cluster is shifted onto the distance α at any discrete moment if no delay occurs. We have obtained the limit distribution of the system states. We have also obtained the limit state distribution (invariant measure) for the open chain of contours with continuous state space and continuos time and for the open chain of contours with discrete state space and discrete time. [ABSTRACT FROM AUTHOR]

Traffic safety at intersecting streams is studied quantitatively using methods from Statistical Mechanics on the basis of simple microscopic traffic flow models. In order to determine a relationship between traffic flow and the number of crashes, the modeling focus is on the building block of any road network, namely the crossing of two streams. In this paper, it is shown that the number of crossing conflicts is proportional to the product of the two traffic flows from which a simple model is developed. This model substantiates known empirical findings. Since real crash data are obtained by an involved process from such building blocks, there is a difference between the theoretical and empirical results. This process is modeled here as well and narrows the mismatch between theory and observation. • A physical model for the conflicts at two crossing streams of vehicles is presented. • For small flows, a simple analytical model is in line with the simulation results. • These results seem to differ from empirical observations. • An additional statistical model narrows the mismatch between simulation and empirics. [ABSTRACT FROM AUTHOR]

This paper develops Lárraga and Alvarez-Icaza (LAI) cellular automata model which is based on safe human reaction. Vehicles are moving at two single-lane streets with periodic boundary condition and pass from intersection point via yielding mechanism avoiding collision. The model characteristics and fundamental diagrams have been obtained by Monte Carlo simulations. Our results suggest that LAI model with yielding mechanism is able to reproduce the realistic acceleration and deceleration capabilities, which are desired parameters. Moreover, the plateau region in total current illustrates the hardness of the yielding mechanism and limited speed strategy, which is applied near the intersection point. [ABSTRACT FROM AUTHOR]

Accurate modelling of local population movement patterns is a core, contemporary concern for urban policymakers, affecting both the short-term deployment of public transport resources and the longer-term planning of transport infrastructure. Yet, while macro-level population movement models (such as the gravity and radiation models) are well developed, micro-level alternatives are in much shorter supply, with most macro-models known to perform poorly at smaller geographical scales. In this paper, we take a first step to remedy this deficit, by leveraging two novel datasets to analyse where and why macro-level models of human mobility break down. We show how freely available data from OpenStreetMap concerning land use composition of different areas around the county of Oxfordshire in the UK can be used to diagnose mobility models and understand the types of trips they over- and underestimate when compared with empirical volumes derived from aggregated, anonymous smartphone location data. We argue for new modelling strategies that move beyond rough heuristics such as distance and population towards a detailed, granular understanding of the opportunities presented in different regions.

We consider a one-dimensional traffic model with a slow-to-start rule. The initial position of the cars in R is a Poisson process of parameter λ . Cars have speed 0 or 1 and travel in the same direction. At time zero the speed of all cars is 0; each car waits a mean-one exponential time to switch speed from 0 to 1 and stops when it collides with a stopped car. When the car is no longer blocked, it waits a new exponential time to assume speed one, and so on. We study the saturated regime λ > 1 and the critical regime λ = 1 , showing that in both regimes all cars collide infinitely often and each car has asymptotic mean velocity 1 / λ . In the saturated regime the moving cars form a point process whose intensity tends to 1. The remaining cars condensate in a set of points whose intensity tends to zero as 1 / t . We study the scaling limit of the traffic jam evolution in terms of a collection of coalescing Brownian motions. [ABSTRACT FROM AUTHOR]

Non-commercial Land Ports of Entry (LPOEs) are unique transportation facilities controlling the ingress and egress of passenger vehicles from Mexico to the United States and vice versa. The calibration of microscopic traffic simulation models of non-commercial LPOEs requires a deep understanding of operational processes and driving behavior at these facilities. This paper provides a methodology to guide modelers in calibrating microscopic traffic simulation models of non-commercial LPOEs. The methodology establishes a common framework for modeling operational processes and traffic operations. Moreover, the methodology includes the value of four state variables that characterizes operations of passenger vehicles at non-commercial LPOEs. These variables are speed, acceleration, deceleration, and headway. The authors evaluated this methodology using the Otay Mesa non-commercial LPOE as a case study. Results showed that this methodology could be potentially used to model non-commercial LPOEs along the US–Mexico border and other border regions worldwide. [ABSTRACT FROM AUTHOR]

Long queues at signals cause fuel-consuming stop-and-go traffic. In this work we use a complete microscopic spatiotemporal measurement of congested city traffic at a signal to i) calibrate a both longitudinal and latitudinal driving model and then to ii) examine how changes in single vehicle's driving behaviour could improve the situation. The model calibration is realized using a genetic algorithm. In this way, a realistic heterogeneous traffic scenario that has similar properties as empirical traffic could be simulated. We then show that already changing the behaviour of a single vehicle per traffic light cycle can significantly reduce the number of vehicles waiting in queues. [ABSTRACT FROM AUTHOR]

We study the derivation of second order macroscopic traffic models from kinetic descriptions. In particular, we recover the celebrated Aw–Rascle model as the hydrodynamic limit of an Enskog-type kinetic equation out of a precise characterisation of the microscopic binary interactions among the vehicles. Unlike other derivations available in the literature, our approach unveils the multiscale physics behind the Aw–Rascle model. This further allows us to generalise it to a new class of second order macroscopic models complying with the Aw–Rascle consistency condition, namely the fact that no wave should travel faster than the mean traffic flow. [ABSTRACT FROM AUTHOR]

We investigate the problem of controller design after linearization around the steady-state solutions of second-order traffic models. At steady-state the traffic variables follow space-dependent autonomous ODEs. The steady-state solutions can be used to account for the heterogeneity of the road in space, which can be caused by different environmental factors. Proceeding by analogy with open-channel flow studies, where such states are relevant for operating a network with efficient control strategies, we analyse and compare steady-state solutions for different second-order traffic models. We show that the steady-state density profile remains in three different regions. The behaviour of the considered models is discussed, leading to proposed corrections of the models. We select one corrected model, which we use for the design and implementation of a boundary controller to illustrate the effectiveness of our approach. [ABSTRACT FROM AUTHOR]

Consider designing a transportation network on n vertices in the plane, with traffic demand uniform over all source-destination pairs. Suppose the cost of a link of length and capacity c scales as for fixed 0

Volume-Delay Function (VDF) is a key input in several tasks including traffic assignment and trip distribution estimation. However, there is no clear framework to identify the VDF formula and parameters. Furthermore, the VDF performance in the oversaturated condition is uncontrolled. In this paper, most VDF-related aspects are investigated, including the key attributes, traffic model and VDF formulas, oversaturated condition assumptions, and optimization techniques. In each aspect, different methods are compared for best performance, while a new procedure is introduced, if necessary. The best traffic model and VDF formulas are nominated to satisfy a set of logic and statistical conditions, while a new optimization technique is presented. Furthermore, a new realistic assumption to control the VDF performance is discussed. This framework is applied on 32 worldwide roads of different categories and characteristics, which indicates accurate and realistic results. The results encourage for further investigation of the VDF performance in future work. [ABSTRACT FROM AUTHOR]