Journal of Transportation Engineering, Part A: Systems

In the prediction problem of urban expressway travel time, in addition to the influence of traffic flow characteristics on travel time, the influence of various traffic environmental factors makes the change of traffic conditions with time uncertain, and the uncertainty and ambiguity in the transportation environment affect the travel-time prediction to varying degrees. This paper studied the influence of weather conditions on expressway travel-time prediction, focusing on the impacts of rain intensity and visibility. The southern section of Sanyuanli-Guangzhou Airport Expressway was selected as a case study to analyze characteristics of travel time under different weather conditions, to determine the change law of travel time and vehicle speed under different rainfall intensity and visibility, and to quantify the uncertainty and fuzziness factors through membership function and parameter weight. The mapping relationship between the influencing factors and travel time was obtained through decision rules, and a travel-time prediction model was established based on soft set theory. The experimental results showed that, compared with the Bureau of Public Roads (BPR) function model, the travel-time prediction model considering weather conditions reduces the prediction error and effectively improves the calculation accuracy.

The current geometric design guide in the US uses design values for the required lengths of acceleration lanes that are based on research studies conducted more than 50 years ago. Those design values need to be updated to reflect current drivers’ behavior patterns and vehicle mechanical characteristics. This paper presents a new method for determining the required lengths of acceleration lanes at freeway interchanges based on actual driver behavior and vehicle acceleration capabilities. Realistic acceleration profiles were established for different drivers accelerating from the design speed of the entrance ramp to that of the highway into which they are merging. The acceleration profiles were established based on field data collected using Global Positioning System (GPS) data-logging devices that recorded the positions and the instantaneous speeds of different vehicle types piloted by different drivers at $1\text{-}\mathrm{s}$ intervals. Design tables were developed for different grades to help designers select the required length of the acceleration lane based on the design speeds of the freeway and the entrance ramp. The developed design tables have the potential to provide design values for the lengths of acceleration lanes that are more realistic and representative of current vehicle and driver characteristics.

The interaction between the bike-sharing behavior and public transportation system is still not fully understood. This paper aims to explore the impacts of public transportation disruption on the bike-sharing behavior utilizing the metro shutdown case in Washington, DC, in 2016. Multisource data, including bicycle trip and geographic information, was collected before, during, and after the public transportation disruption. Through the spatial-temporal analysis, it is a recurrent pattern for bike-sharing members to take their trips on weekdays and the casual users without membership to take the trips on weekends. Moreover, it is noted that the bike-sharing trips for casual user increased 191% and the variation of trip duration is insignificant on metro shutdown day. The bike-sharing stations adjacent to the metro lines are found to experience a higher trip increase. Furthermore, the complex network approach was employed to explore the bike-sharing network properties. In network perspective, the connectivity of bike-sharing network has increased due to the shifted travelers. In addition, the modularity maximization approach was also introduced to infer the community structure of bike-sharing network. The results suggest the disruption strengthen the interaction between central clusters and outer clusters. Implications for further research, policy, and design are discussed.

The objective of this paper is to develop motorized level of service models using advanced heuristic techniques like functional networks, ordered probit analysis, and multivariate autoregressive spline for uncontrolled intersections under mixed traffic conditions. A total of 13 uncontrolled intersections from India are chosen and geometric, traffic, and roadside environmental data are collected using high-definition cameras. A novel influence area (INA) method for critical gap and follow-up time measurement is also developed. Over 7,000 effective uncontrolled intersection motorized driver responses are collected based on user satisfaction scores (1 = excellent to 6 = very bad). Eight variables with high significance ($p<0.001$) effects on perceived scores from Spearman’s correlation technique are modeled. Functional networks model showed greater efficiency as indicated by greater marginal ranking index of 11 (leading to lowest rank of first) among all the three level of service models proposed in this study. According to the sensitivity analysis, analogous to the proposed functional network model, volume to capacity ratio (highest effect), percentage of on-street parking (second highest effect), and service delay (third highest effect) showed superlative sensitivity toward level of service predictions at uncontrolled intersections compared to other parameters. The imperative outcomes of this research may help transport planners and traffic engineers to quantify operational evaluation of uncontrolled intersections and take crucial decisions toward their improvement.

With a growing number of intelligent transportation system sensors and the networkwide deployment of those across the nation’s roadway facilities, current research and practices should concentrate on more proactive safety strategies. In recent years, real-time traffic data collected from ITS sensors have been utilized to develop crash prediction models. Real-time crash prediction models can be used to identify hazardous traffic conditions that might cause a crash. This study aims to examine how employing data mining techniques that account for imbalanced data could improve the predictive capability of real-time crash prediction models. The term imbalanced data refers to a condition where the number of observations in each class is not equally distributed among the data set (noncrash cases outnumber crash cases). To decrease the within-class variation of imbalanced data, the data were split into two traffic-state data sets: free-flow speed (FFS) and congestion. Three models, including logistic regression as the baseline, random forest (RF) with random undersampling, and Adaptive Boosting (AdaBoost), were estimated with each data set. The results were compared with the models that were estimated using the complete set of data. Model comparisons indicated that all three models achieved significantly better predictive results with the congested and FFS data sets as opposed to the data set containing all crashes and that, while in some cases the results of the undersampled RF model were slightly better than those of AdaBoost, both models outperformed the logistic regression model. The results of this study demonstrated that using models to deal with imbalanced data and lowering the variation of imbalanced data could substantially improve crash prediction accuracy. The findings could help traffic agencies to practically implement and deploy crash prediction models for real-time applications and develop crash prevention strategies accordingly.

This article develops and validates an algorithm to estimate the lateral position of wheel-rail contact and the corresponding rail profile radius. The lateral contact position and contact radius are two novel rail profile performance measures that enable more refined characterization of the rail profile and its influence on rail wear and vehicle dynamics. Leveraging recent advancements in rail profile monitoring techniques, the algorithm contributes to rail maintenance research and practice by developing new measures of performance based solely on commonly available rail profile data. The algorithm developed in this article is an automated process that estimates the lateral position of wheel-rail contact and the corresponding rail profile radius along a rail segment. It uses measured rail profile data as an input and applies rigid contact theory to model contact between a linear wheel profile and the rail profile. The lateral contact position and contact radius are calculated using computer programming that provides graphical and numerical results on a profile-by-profile basis as well as summary statistics for each rail segment. This methodology produces expected results when subjected to validation tests. The validation process analyzes the rationality of algorithmic output against a series of expected results using rail profile data from selected tangent segments of a closed-loop captive-fleet North American rail transit property. The algorithm output does not significantly deviate from any of the expected results, and as such, the algorithm is considered validated.

The signal cycle plays an important role in the design of timing optimization and evaluation for a signal intersection. However, in some cities, there is a lack of unified management of intelligent transportation system, and even the intersection signal cycle has not been archived online and updated in real time, resulting in a signal cycle that is not convenient and reliable. Therefore, it is necessary to use a simple and effective method to obtain the real-time signal cycle in a large regional area. This paper proposes an innovative traffic grid model, which matches the massive floating car data with the intersections across the entire urban road network, to estimate the signal cycle length of intersections accurately. This estimation method is composed of four major parts: (1) a grid model is built to transform intersections into discrete cells, and the floating car data are mapped to the grids through a simple assignment process; (2) based on the grid model, a set of key traffic parameters (e.g., the time stamp of vehicle stops or starts and the position of stop) is derived; (3) the augmented Dickey–Fuller (ADF) test and Pettitt test are used to identify the timing scheme constant fixed signal cycle in 1 day (CFSC) or variable fixed signal cycle in various periods of 1 day (VFSC); and (4) the $k$-means clustering algorithm with the trajectory similarity measurement is used to accurately estimate the cycle of each period. Taking the intersections of Beijing as an example, the effectiveness and feasibility of this method were demonstrated. The proposed cycle estimation method can provide valuable insights for the study of traffic signal control, and management and can be extended to other cities.

Because continuous-flow intersections (CFIs) are a relatively new intersection design, there are few existing guidelines for designing signal timings for CFIs. An appropriate signal-timing plan will maximize the capacity of the intersection, reduce congestion, and improve safety. This research developed a new signal-timing strategy for CFIs that is based on traffic progression. This new CFI signal-timing strategy was evaluated by conducting traffic simulation-based experiments, and the results of the evaluation showed that it outperformed the signal-timing plan provided by a commonly used existing signal-timing optimization tool. The proposed signal-timing strategy can reduce average traffic delay by 24%, average vehicle travel time by 8.5%, and average queue length by 28.8% at the studied CFI.

Based on the fact that there are nonstrict priority left-turning maneuvers at signalized intersections with a permitted phase in some countries, this study proposes a left-turn unit-based approach to estimating permitted left-turn capacity with an exclusive lane. First, a field study is conducted to look at the capacity limitations and crossing decisions in a left-turn unit using regression analysis and classification tree techniques. With the core idea that left-turn units cross the intersection during the duration of the initial green time, or use gaps provided by opposing right-turning vehicles, the permitted left-turn capacity model is developed using the probability theory. It involves nine factors related to the phasing schedule, the intersection geometry, driving maneuvers, and the proportion of traffic streams making different movements. Finally, validation and discussion of the method compared with previous methods show the proposed model better reflects the operational traffic performance of an intersection with a permitted phase.

Heavy rainfall events are becoming more common in many areas with escalating climate change, and these events can considerably affect travel speed and road safety. It is critical to understand when and how rainfall events affect congestion in the transportation network to help improve decision making for infrastructure planning and real-time operations. This study incorporates high-resolution rainfall and wind data into a travel speed prediction model, along with other data related to weather conditions, incidents, and real-time speeds, to assess if localized rainfall data can inform travel speed prediction during light and heavy rainfall events, and how this compares with the classical method of using a single city-wide rain gauge data point. The travel speed prediction model holistically selects the most related features from a high-dimensional feature space by modeling by wind direction, correlation analysis, and the least absolute shrinkage and selection operator (LASSO) to overcome overfitting issues and is applied to two urban arterials for case studies located in Pittsburgh, Pennsylvania. The results indicate that high-resolution rainfall features in many instances are better predictors of future rainfall on the target segments, leading to overall better prediction results (in $30\text{-}\min$ lag time), when compared with models that use a single city-wide rain gauge. This has implications for other cities that are interested in improving travel speed prediction modeling and traffic modeling under increasing impacts from climate change and extreme weather.