Estimating traffic volumes for signalized intersections using connected vehicle data

Recently connected vehicle (CV) technology has received significant attention thanks to active pilot deployments supported by the US Department of Transportation (USDOT). At signalized intersections, CVs may serve as mobile sensors, providing opportunities of reducing dependencies on conventional vehicle detectors for signal operation. However, most of the existing studies mainly focus on scenarios that penetration rates of CVs reach certain level, e.g., 25%, which may not be feasible in the near future. How to utilize data from a small number of CVs to improve traffic signal operation remains an open question. In this work, we develop an approach to estimate traffic volume, a key input to many signal optimization algorithms, using GPS trajectory data from CV or navigation devices under low market penetration rates. To estimate traffic volumes, we model vehicle arrivals at signalized intersections as a time-dependent Poisson process, which can account for signal coordination. The estimation problem is formulated as a maximum likelihood problem given multiple observed trajectories from CVs approaching to the intersection. An expectation maximization (EM) procedure is derived to solve the estimation problem. Two case studies were conducted to validate our estimation algorithm. One uses the CV data from the Safety Pilot Model Deployment (SPMD) project, in which around 2800 CVs were deployed in the City of Ann Arbor, MI. The other uses vehicle trajectory data from users of a commercial navigation service in China. Mean absolute percentage error (MAPE) of the estimation is found to be 9–12%, based on benchmark data manually collected and data from loop detectors. Considering the existing scale of CV deployments, the proposed approach could be of significant help to traffic management agencies for evaluating and operating traffic signals, paving the way of using CVs for detector-free signal operation in the future.     

Experimental study and modeling of car-following behavior under high speed situation

To investigate the car-following behavior under high speed driving conditions, we performed a set of 11-car-platoon experiments on Hefei airport highway. The formation and growth of oscillations have been analyzed and compared with that in low speed situations. It was found that there is considerable heterogeneity for the same driver over different runs of the experiment. This intra-driver heterogeneity was quantitatively depicted by a new index and incorporated in an enhanced two-dimensional intelligent driver model. Using both the new high-speed and the previous low-speed experimental data, the new and three existing models were calibrated. Simulation results show that the enhanced model outperforms the three existing car-following models that do not take into account this intra-driver heterogeneity in reproducing the essential features of the traffic in the experiments.     

Theory of safety surrogates using vehicle trajectories in macroscopic and microscopic settings: Application to dynamic message signs controlled traffic at work zones

This paper presents a new mathematical framework for obtaining quantitative safety measure using macroscopic as well as microscopic traffic data. The safety surrogate obtained from the macroscopic data is in terms of analysis performed on vehicle trajectories obtained from the macroscopic data. This method of obtaining safety measure can be used for many different types of applications. The safety surrogate for the traffic dynamics are developed in terms of a new concept of Negative Speed Differentials (NSD) that involve a convolution of vehicle speed function obtained from vehicle trajectories and then performing the integration of the square of the output for its negative values. The framework is applicable to microscopic traffic dynamics as well where we can use car following models for microscopic dynamics or the LWR model for macroscopic dynamics. This paper then presents the use of this new safety surrogate on the development of a feedback control law for controlling traffic in work zones using Dynamic Message Signs. A hybrid dynamics model is used to represent the switching dynamics due to changing DMS messages. A feedback control design for choosing those messages is presented as well as a simple simulation example to show its application.     

The symmetric intersection design and traffic control optimization

It is well recognized that the left-turning movement reduces the intersection capacity significantly, because exclusive left turn phases are needed to discharge left turn vehicles only. This paper proposes the concept of Left-Hand Traffic (LHT) arterial, on where vehicles follow left-hand traffic rules as in England and India. The unconventional intersection where a LHT arterial intersects with a Right-Hand Traffic (RHT) arterial is named as symmetric intersection. It is only need three basic signal phases to separate all conflicts at symmetric intersection, while it at least need four signal phases at a conventional intersection. So, compared with the conventional intersection, the symmetric intersection can provide longer green time for the left-turning and the through movement, which can increase the capacity significantly. Through-movement waiting areas (TWAs) can be set at the symmetric intersection effectively, which can increase the capacity and short the cycle length furthermore. And the symmetric intersection is Channelized to improve the safety of TWAs. The Binary-Mixed-Integer-Linear-Programming (BMILP) model is employed to formulate the capacity maximization problem and signal cycle length minimization problem of the symmetric intersection. The BMILP model can be solved by standard branch-and-bound algorithms efficiently and outputs the lane allocation, signal timing decisions, and other decisions. Experiments analysis shows that the symmetric intersection with TWAs can increase the capacity and short the signal cycle length.     

Mental models of driving and speed: biases,choices and reality

This paper provides a review of research performed by Svenson with colleagues and others work on mental models and their practical implications. Mental models describe how people perceive and think about the world including covariances and relationships between different variables, such as driving speed and time. Research on mental models has detected the time-saving bias [Svenson, O. (1970). A functional measurement approach to intuitive estimation as exemplified by estimated time savings. Journal of Experimental Psychology, 86, 204–210]. It means that drivers relatively overestimate the time that can be saved by increasing speed from an already high speed, for example, 90–130?km/h, and underestimate the time that can be saved by increasing speed from a low speed, for example, 30–45?km/h. In congruence with this finding, mean speed judgments and perceptions of mean speeds are also biased and higher speeds given too much weight and low speeds too little weight in comparison with objective reality. Replacing or adding a new speedometer in the car showing min per km eliminated or weakened the time-saving bias. Information about braking distances at different speeds did not improve overoptimistic judgments of braking capacity, but information about collision speed with an object suddenly appearing on the road did improve judgments of braking capacity. This is relevant to drivers, politicians and traffic regulators.     

A two-dimensional simulation model for modelling turning vehicles at mixed-flow intersections

The turning behavior is one of the most challenging driving maneuvers under non-protected phase at mixed-flow intersections. Currently, one-dimensional simulation models focus on car-following and gap-acceptance behaviors in pre-defined lanes with few lane-changing behaviors, and they cannot model the lateral and longitudinal behaviors simultaneously, which has limitation in representing the realistic turning behavior. This paper proposes a three-layered “plan-decision-action” (PDA) framework to obtain acceleration and angular velocity in the turning process. The plan layer firstly calculates the two-dimensional optimal path and dynamically adjusts the trajectories according to interacting objects. The decision layer then uses the decision tree method to select a suitable behavior in three alternatives: car-following, turning and yielding. Finally, in the action layer, a set of corresponding operational models specify the decided behavior into control parameters. The proposed model is tested by reproducing 210 trajectories of left-turn vehicles at a two-phase mixed-flow intersection in Shanghai. As a result, the simulation reproduces the variation of trajectories, while the coverage rate of the trajectories is 88.8%. Meanwhile, both the travel time and post-encroachment time of simulation and empirical turning vehicles are similar and do not show statistically significant difference.     

Congestion and pollution consequences of driving-to-school trips: A case study in Beijing

Parents compete for high-quality education for their children by enrolling them in good schools. However, in a Chinese mega-city like Beijing, three factors jointly lead to the spatial separation between schools and homes: the centralized public goods provision mechanism, the historical dependency in school location, and the constrained supply of housing in downtown. Without an adequate number of school buses, this spatial separation of schools and homes triggers the numerous long-distance driving-to-school trips by private vehicle during workday morning rush hours in Beijing. We use the start and end dates of “school holiday” as exogenous repeated shocks to the aggregate traffic congestion, and employ the two-stage least squares (2SLS) regression approach to examine the congestion and pollution consequences of such driving-to-school trips in Beijing. We find that, all else being equal, workdays during school holidays have a traffic congestion index 20% lower than that of non-school-holiday workdays. Such a sharp reduction in congestion leads to a significant decrease in PM₁₀ concentration. Policymakers should lower such “extra” congestion and environmental costs via optimizing the spatial balance between school supply and demand.     

Liner shipping network design with emission control areas: A genetic algorithm-based approach

To curb emissions, containerized shipping lines face the traditional trade-off between cost and emissions (CO₂ and SO_x) reduction. This paper considers this element in the context of liner service design and proposes a mixed integer linear programming (MILP) model based on a multi-commodity pickup and delivery arc-flow formulation. The objective is to maximize the profit by selecting the ports to be visited, the sequence of port visit, the cargo flows between ports, as well as the number/operating speeds of vessels on each arc of the selected route. The problem also considers that Emission Control Areas (ECAs) exist in the liner network and accounts for the vessel carrying capacity. In addition to using the MILP solver of CPLEX, we develop in the paper a specific genetic algorithm (GA) based heuristic and show that it gives the possibility to reach an optimal solution when solving large size instances.     

A hybrid deep learning based traffic flow prediction method and its understanding

Deep neural networks (DNNs) have recently demonstrated the capability to predict traffic flow with big data. While existing DNN models can provide better performance than shallow models, it is still an open issue of making full use of spatial-temporal characteristics of the traffic flow to improve their performance. In addition, our understanding of them on traffic data remains limited. This paper proposes a DNN based traffic flow prediction model (DNN-BTF) to improve the prediction accuracy. The DNN-BTF model makes full use of weekly/daily periodicity and spatial-temporal characteristics of traffic flow. Inspired by recent work in machine learning, an attention based model was introduced that automatically learns to determine the importance of past traffic flow. The convolutional neural network was also used to mine the spatial features and the recurrent neural network to mine the temporal features of traffic flow. We also showed through visualization how DNN-BTF model understands traffic flow data and presents a challenge to conventional thinking about neural networks in the transportation field that neural networks is purely a “black-box” model. Data from open-access database PeMS was used to validate the proposed DNN-BTF model on a long-term horizon prediction task. Experimental results demonstrated that our method outperforms the state-of-the-art approaches.     

Modeling free-flow speed according to different water depths—From the viewpoint of dynamic hydraulic pressure

In this paper, we propose a method of modeling free flow speed from the viewpoint of hydroplaning. First, the lift forces for different water depths were estimated using Bernoulli’s equation. Compared with the result of the experimental test performed by the Japan Automobile Research Institute, the hydrodynamic pressure coefficient was determined to be 0.03 (tf s²/m⁴). The validation of the predicted lift force is found in another published paper. A very good match is found between the computed values by the proposed numerical model and the data in other published papers. Then, the loss of contact force is considered to evaluate the hydroplaning performance of a tire. To simulate the hydroplaning speed, a tire-sliding model was utilized to obtain the traction and friction forces between the road surface and the tire. The observation data obtained in Japan in 2009 is compared with the physically computed hydroplaning speed, yielding the conclusion that the traction force at the measured desired speed is, on average, 23.4% of the traction force at hydroplaning speed. The analytical model offers a useful tool to quantitatively show that the free flow speed changes as the water depth increase.