Journal of Transportation Engineering, Part A: Systems

Two-lane rural roads constitute a significant part of the roadway system in India. Geometric characteristics such as facility type, lane width, shoulder width, and horizontal and vertical alignments, are essential parameters that influence vehicle behavior and traffic flow characteristics on two-lane rural roads. Among all geometric characteristics, the magnitude of the grade most substantially impacts the operational characteristics of traffic flow on two-lane rural roads. The present study investigates the effect of grade on the operating speed and capacity of two-lane rural roads under mixed traffic conditions. Traffic video data for eight road sections with grades varying from 1% to 8% were collected under dry weather conditions. The investigation revealed a significant effect of grade on the operating speeds of different vehicle types. The operating speed decreased with an increase in the grade magnitude. The road capacity was derived by calibrating various single regime models. The Northwestern model was deemed appropriate to derive the capacity values, based on theoretical and statistical investigation. The results showed that the capacity decreases by 6.4% with every 1% increase in grade. Further, the effect of grade on passenger car units (PCU) of different vehicle types was investigated. For varying volume-to-capacity (V/C) ratios, it was observed that the PCU of heavy vehicles increases as the magnitude of the grade increases. The present study develops operating speed and capacity prediction models as an essential practical outcome. The developed models can facilitate planners and traffic engineers to estimate operating speed and capacity based on the magnitude of the grade.

Accurate incident impact prediction is one of the crucial tasks in road safety management that contributes to mitigating the resultant congestion and accelerating emergency rescue. However, the impact of traffic incidents is affected by numerous and redundant factors, and their incompleteness and scalability make the task of impact prediction particularly challenging. To address these issues, a novel multimodule dynamic prediction framework, named information enhancement cascade forest (IECF), is proposed for traffic incident impact prediction in this paper. In the IECF model, XGBoost is adopted to screen the original incident factors to reduce the information redundancy. Additionally, a novel enhancement generative adversarial network (EGAN) was custom designed to complete the missing factors by historical large-sample adversarial training. Then a multigrained permutation cascade forest (GPCF) was constructed to predict the incident impact degree through integrated learning. Furthermore, sufficient comparative experiments were conducted, and the results show the proposed model has better performance on traffic incident impact dynamic prediction than the state-of-art methods, especially in partial-factors-available situations.

Automatic camera calibration is a fundamental technology for 3D traffic parameter extraction. With the popularity of pan-tilt-zoom cameras, this technique demonstrates great potential to enhance traffic safety and efficiency, especially for highways. This paper aims to present a fully automatic calibration method for surveillance cameras in highway scenes. Our system is divided into two stages. In the first stage, a deep convolution neural network was used to estimate a pair of orthogonal vanishing points from multiple vehicles. This process transformed vanishing point detection into an estimation of vehicle direction, which was further determined by introducing the central residual mechanism. In the diamond space, the straight lines formed by these directions accumulated the final positions of the vanishing points. More importantly, we proposed a novel algorithm for estimating the lane width using vehicle trajectories in the second stage. It can be used to calculate the camera height, making the calibration fully automated. We also corrected the distorted lens using vehicle trajectories. Comprehensive experiments were conducted on the proposed data set and the BoxCars116k data set. The results indicate that the composite mechanism (i.e., classification and central residual) significantly improves the accuracy and robustness of the vanishing point estimation. Combined with automatic camera height estimation, our technology is superior to the most representative methods in calibration performance. Since it does not have any constraints on road geometry and camera placement, our approach applies to most highway surveillance systems.

Transportation agencies make substantial efforts to implement safety improvement countermeasures to mitigate safety hazards. However, region-specific safety considerations, rather than accepting more general, widespread methods, and the most ideal investment decisions to improve safety, are often unclear. This is of increased concern for horizontal curve segments, because they are locations of elevated safety risk. Yet, there exists a gap in literature on the development and use of a geospatial tool to investigate horizontal curve safety. To fill this gap, a methodological approach to create a region-specific geospatial horizontal curve safety tool was developed in this research. The tool was created using two regions as application areas to ensure the methodological approach was reproducible and transferable. Geolocated crash data, roadway infrastructure data, and curve data were spatially integrated using a GIS. Bayesian hierarchical models were estimated with crash severity data to gain region-specific safety results. A GIS tool was then derived from the applied model coefficients. The tool was applied to prioritize the most safety-critical horizontal curves and to select optimal countermeasures, especially in cases with specific infrastructure investment identification needs. The results of this research benefit regional agencies in their aim to efficiently distribute investments and identify the most appropriate countermeasures to improve roadway safety in their region.

Presently, 15 US states require that passenger vehicles undergo periodic safety inspections. Past studies estimating the effectiveness of these safety inspection and maintenance programs (I/M programs) in their stated aim of mitigating road accidents and fatalities have tended to rely on outdated data sets or to focus on specific geographic regions. Since inspection program effectiveness continues to be deliberated in legislative bodies across the country, this paper aims to present a replicable and data-driven quantification of the effects of I/M programs on road fatalities, applying the largest available data set, covering all 50 US states over a 44-year period. Numerous panel data regressions showed a negative correlation between the presence of state I/M programs and the fleet size–adjusted roadway fatality rate. Fixed-effects (FE) estimates suggest that states with I/M programs had 5.5% fewer roadway fatalities per 100,000 registered passenger vehicles—95% confidence interval (CI): 0.4% to 10.6%—nationwide based on nearly exhaustive data for fatal passenger vehicle accidents from 1980 to 2015. These results are complemented by an additional FE specification with fewer variables, but over a longer period. A two-stage least-squares specification is also presented that not only supports this finding but also implies a causal relationship between the presence of I/M programs and lower road fatality rates. The convergent results of the regressions presented in this paper provide compelling evidence that jurisdictions experience lower roadway fatality rates due to the presence of an active safety I/M program for passenger vehicles.

When metro networks are shut down due to an unconventional emergency, numerous passengers will get stranded and wait to be evacuated. A clear understanding of the stranded passengers in the network is the basis for designing an efficient evacuation strategy. Previous studies focused more on disruption management and assumed a known evacuation demand. Few studies considered the distribution of stranded passengers, especially when the metro network is shut down due to unconventional emergencies. This motivates us to develop a discrete event-based simulation model to estimate the number and distribution of stranded passengers. The proposed model takes the numbers of stranded passengers at stations and trains as the state variables and develops the state evolution rule of the metro system. The numbers and destinations of stranded passengers at trains and on stations at each time step, including the transfer and non-transfer stations, are calculated to describe the state evolution of the metro system. Chongqing Rail Transit (CRT) network is taken as an example. The result shows that the stations with the most stranded passengers are located downtown. Origin-Destinations (ODs) with more stranded passengers have a relatively shorter travel distance, while ODs with fewer stranded passengers have a relatively long travel distance. The average directional disequilibrium factor of stranded passengers of ODs in the whole network is high, especially for those ODs with many stranded passengers. These findings reveal the distribution characteristics of stranded passengers and can provide significant assistance to an emergency manager who is designing an efficient emergency evacuation plan.

In recent years, China’s rail network has been expanding to western China. Most of western China is a complicated and steep mountainous area with complex terrain and a harsh natural environment. Many engineering geological problems and unfavorable geological hazards have been encountered in railway construction in this area. Railway route selection in complex and steep mountainous (CSM) areas has become one of the fundamental issues in railway design. In view of the existing literature, this paper is aimed to summarize the characteristics and difficulties of railway route selection in CSM areas. In view of the low efficiency of railway route selection in CSM areas, a set of general models for railway route selection in CSM areas is established in the present study. A new multifactor interactive matrix dynamic weight approach is proposed in this model, and the corresponding evaluation index system is then developed. This methodology has been successfully applied in two typical CSM railway route selection cases, which proves the validity and applicability of this novel approach. This scrutiny provides a newly solid reference for the route selection method of the railway in CSM areas and gives a new direction for the evaluation and decision-making of other complex system engineering problems.

The new generation smart highways (NGSH) have emerged as irresistible trends to enhance the efficiency and safety of transportation systems. An integral component of the NGSH is the automation of the intelligent traffic management system (ITMS). This study investigates an integrated framework for the ITMS that incorporates the fine-grained microscopic simulation and deep learning technologies based on real-time traffic data. The framework commences by performing dynamic corrections based on the license plate, vehicle speed, location, and other information provided by the real-time bayonet data in order to simulate the realistic traffic flow along the highway. A deep learning model based on long short-term memory (LSTM) is then applied to predict the short-term traffic volume on major highway segments. Based on prediction results, a collaborative management method is constructed that combines variable speed limits and ramp metering. The case study on the Shanghai–Hangzhou–Ningbo Highway in China suggests the real-time simulation model can control the average error of the traffic volume on the main segments by 4.58%. The LSTM-based model can accurately predict the short-term traffic volume with a relative error of 85% below 15% in both offline and online modes. Consequently, the proposed collaborative framework improves the average speed and traffic volume of controlled sections by 3.62% and 4.35%, respectively, demonstrating its effectiveness in improving the operation and management of the smart highways.

Accurate traffic estimation on urban networks is a prerequisite for efficient traffic detection, congestion warning, and transportation schedule. The current estimation methods can be roughly divided into model-driven and the data-driven methods. The estimation accuracy of the model-driven methods cannot satisfy certain applications. Meanwhile, the data-driven methods have the disadvantages of poor generalization ability and weak interpretability. To overcome these challenges, this paper proposes a framework named the physics-informed spatiotemporal graph convolution neural network (PSTGCN) based on physics-informed deep learning theories. The PSTGCN uses a spatiotemporal graph convolution neural network combined with traffic flow models to estimate the traffic state. The proposed model not only considers the temporal and spatial dependence of traffic flow but also abides by the internal law of traffic flow. Furthermore, the estimation objects of the proposed model are multiple variables that comprehensively represent the traffic state. Experiments on real-world traffic data reveal the error of the PSTGCN is reduced by 38.39% compared to the baselines. Also, the PSTGCN can achieve a similar prediction effect as the baselines by using half of the global spatial information. These results demonstrate that the PSTGCN outperforms the state-of-the-art models in urban traffic estimation and is robust under variable road conditions and data scales.

The dwell time (DT) is an essential element in the compilation of urban rail transit train diagrams. To improve the reliability of urban rail transit operations, it is necessary to identify the major factors that influence the actual DT and formulate effective measures. Via data mining, this study proposes a random forest (RF) based identification model to diagnose the major problems that affect the DT at different stations in different periods and put forward corresponding targeted measures. Considering the Beidajie (BDJ) station of Xi’an Metro in the upstream direction of Line 2, this model sensitively identifies major factors leading to the variation in DT during the morning peak, consistent with the detailed analysis. Then, the targeted improvement measures are proposed for the BDJ station. The obtained application results indicate that the established influencing factor identification model can effectively dig out the factors causing a discrepancy between actual and scheduled DT, which can provide auxiliary decision making in the compilation of train diagrams and daily operation organization.

This study focuses on vehicle and pedestrian traffic management in freeway tunnels during emergencies such as fire propagation caused by traffic accidents. Such accidents can result in traffic breakdowns in a short time and a limited space, leading to severe health and safety issues. This study establishes a parsimonious bimodal tunnel evacuation management strategy that can provide real-time evacuation instructions by combining real-world historical traffic data and a cell transmission model (CTM). To exploit the evacuation capability of road tunnels with both vehicle cross and passenger passages, an equivalent cellular traffic network of a road tunnel and its escape facilities is first developed for bimodal evacuation. Simultaneously considering efficiency and social fairness in an emergency evacuation, a risk evaluation function is initially established and embedded in a system optimal objective function. The rationale of the proposed compact system models and the mechanism of effective traffic management measure generation is validated based on a real freeway tunnel emergency happened in Melba and Mullum-Mullum tunnels in Melbourne, Australia. The results suggest that the proposed CTM-based model has twinned the bimodal traffic characteristics and the flexibility to balance the evacuation efficiency and social fairness consideration in road tunnels under emergency conditions.

This study aims to assess the safety effect of the paved shoulder width of a 175-km six-lane rural highway in India. First, the applicability of the matched case-control (C-C) method in the case of highway segments with multiple crashes was assessed. Second, the safety effectiveness of the paved shoulder and other risk factors were evaluated using the matched C-C method. Three scenarios with different case definitions were developed using the matched C-C method and ordinal logistic regression. The study used 100-m segment-specific speed, volume, geometrics, roadside characteristics, and fatal crash data between 2016 and 2018. Developed scenario results suggested that the matched C-C method can assess crash risks on segments with multiple crashes by defining cases separately or considering the number of crashes as an ordinal variable. Results also suggested that the probability of observing crashes increases as paved shoulder width increases. However, the probability of observing multiple fatal crashes decreases for wide paved shoulder width increases. The safety effectiveness of paved shoulder width assessment showed that a wide paved shoulder had a higher risk of a fatal crash than a paved shoulder width of 1.0 to 1.5 m. The segments with unpaved shoulders were unsafe compared to segments without them. These results can be used to prepare guidelines by road-owning and construction agencies to adopt safe paved shoulder widths for six-lane and multilane highways in India. Results may also support the development of a new highway design policy concerning preferable safe paved shoulder width for multilane greenfield highways in India.

Robotaxis have provided services to the public in some cities by different operators. There would exist connected and autonomous vehicles (CAVs) belonging to multiple firms on urban roads in the future. Each firm takes charge of its CAVs by giving the plan of origin-destination (OD) routes and intersection trajectories. It is essential to investigate firms’ interactions with each other and estimate their influence on the traffic system. The competition for limited road resources among CAV firms needs to be considered during their joint planning of intersection trajectories and OD routes. First, potential conflicts in the autonomous intersection are analyzed in detail. Next, a simulation framework is proposed to model the competition process of CAV firms. Third, two safety management strategies are designed for resolving conflicts among CAVs from different firms at intersections. Finally, some simulation experiments are accomplished in a road network with two firms. The results reveal that CAV firms’ right to vehicle route planning should not be abridged. The competition ensures the fairness of using road resources and results in only a 15% system efficiency loss in a specific strategy.

Roundabouts are a superior solution to traffic operation issues at intersections with complex geometry, such as skewed and offset approaches where there are property constraints or where realigning approaches is costly. Although roundabouts with offset, skewed approaches are unavoidable, the current design standards are typically established based on standard roundabouts having radial approach alignments and perpendicular approaches. This work aimed to (1) examine the geometric characteristics of roundabouts with offset or skewed approaches or both, (2) establish the relationships of the design parameters with the approach offset and angle, (3) develop geometric optimization models, and (4) recommend geometric design guidelines for inscribed circle radius, entry curve radius, and exit curve radius. The approach or intersection angle ranges are also recommended for single-lane and two-lane roundabouts under urban and rural conditions. Design graphs and tables of the minimum required radii were developed. They not only help highway designers and engineers to select the appropriate parameters and reduce trial-and-error efforts during design, but also offer a guide to property needs during the planning stage.

Ice cover on pavement may reduce the road adhesion coefficient and increase the crash risks, which might result in more traffic crashes. The primary factor utilized to assess whether the wet pavement is icy or not is the pavement temperature. Therefore, forecasting pavement temperature is an effective method to judge road conditions and improve traffic safety. This paper proposes a combination model based on the extreme gradient boosting (XGBoost) model and long- and short-term time-series network (LSTNet) model to predict pavement temperature. Pavement temperature and meteorological data were collected for the cities along the Shandong part of the Beijing-Taipei Expressway (G3). In this study, nine meteorological variables were used. Subsequently, after correlation analysis, five variables, including air temperature, dew point temperature, relative humidity, evaporation, and potential evaporation, were selected for prediction. The method proposed in this study comprises the following steps. First, the XGBoost and the LSTNet models are respectively formulated based on the time-varying characteristics of pavement temperatures. Then, using the preset weight of the variable, the XGBoost model is used for preliminary prediction to add features. Finally, the experimental analysis is performed on the Qihe data set after the two models have been integrated using the inverse variance method. As revealed by the experimental results, the mean absolute error (MAE) and root-mean-square error (RMSE) of the proposed XGBoost-LSTNet model are 0.8235 and 1.2412, respectively. Compared with the long short-term memory (LSTM) model, random forest (RF) model, XGBoost model, and LSTNet model, the XGBoost-LSTNet model proposed in this paper has higher accuracy. The study’s findings can successfully increase wintertime expressway traffic safety and serve as a guide for managing maintenance and preventing icing-related accidents.

With the development of traffic detection facilities technology, it is currently possible to obtain high-resolution traffic flow data. Due to the particular driving characteristics of vehicles on freeways, once traffic crashes occur, they are generally with serious consequences, and hence traffic safety issues on freeways have always been popular topics. In order to better realize the change from static analysis after the crash to dynamic analysis before the crash toward freeway safety, as well as explore the relationship between dynamic traffic flow characteristics and real-time traffic crash risk under different temporal conditions, this research constructed a real-time traffic crash risk prediction model considering the temporal effect difference. First, traffic crash information and the matched big data of high-resolution traffic flow located on the section of milepost 100–130 of Interstate 5 (I-5) in Washington State, were extracted. In terms of temporal dimension, the research object was divided into weekdays and weekends, and the traffic state was divided into unsaturated and saturated. The random forest (RF) algorithm was introduced to identify the traffic flow variables of crash precursors, and support vector machine (SVM) was applied to build the traffic crash risk prediction model under the condition of temporal difference. A confusion matrix, receiver operating characteristic (ROC) curve, and area under curve (AUC) values were used to evaluate the accuracy of the model performance. Furthermore, the prediction performance of the proposed model was tested via constructing the risk model without consideration of temporal effect and traffic state difference. Finally, the rationality of variable screening was verified by inputting the data set without variable screening into the constructed model. The results showed that the occurrence mechanism of dynamic traffic crashes under different temporal effect conditions varies; the AUC values of the constructed prediction model were all between 0.7 and 0.9, indicating that the recommended model has good prediction accuracy. In conclusion, the real-time freeway traffic crash risk prediction model considering the temporal effect difference has certain advantages compared with the conventional model, and its performance is better than the prediction model without screening of important traffic flow variables. This approach can provide theoretical guidance for dynamic traffic safety management toward freeway under temporal difference conditions.

This study analyzes the effect of the restrictions in traffic movement enforced in order to combat the spread of coronavirus on air quality and travel time reliability under heterogeneous and laneless traffic conditions. A comparative analysis was conducted to examine quantity of pollutants, average travel time distributions (TTD), and their associated travel time reliability (TTR) metrics during the COVID-19 pandemic, postpandemic, and during partial restrictions. Pollutants data (${\mathrm{PM}}_{2.5}$, ${\mathrm{NO}}_2$, and ${\mathrm{NO}}_{\mathrm{X}}$) and travel time data for selected locations from Chennai City in India were collected for a sample period of one week using Wi-Fi sensors and state-run air quality monitoring stations. It was observed that the average quantity of ${\mathrm{PM}}_{2.5}$, ${\mathrm{NO}}_2$, and ${\mathrm{NO}}_{\mathrm{X}}$ were increased by 433.1%, 681.4%, and 99.2%, respectively, during the postlockdown period. Correlation analysis also indicated that all considered air pollutants are moderately correlated to Wi-Fi hits, albeit to varied degrees. From the analysis, it was also found that average TTD mean and interquartile range values were increased by 47.2% and 105.2%. In addition, the buffer time index, planning time index, travel index, and capacity buffer index associated with these TTD metrics were increased by 148.1%, 63.7%, 42.8%, and 202.9%, respectively, soon after relaxing travel restrictions.

Abstract

Low-speed urban midblock sections of old cities of developing countries that have multiple access point, illegal median cuts or breaks, or restricted and broken shoulders witness good number of severe and fatal crashes, comparable to crashes at intersections. Limited studies have focused on analyzing factors influencing fatal and severe crashes for these types of locations. Traditionally, crash severity patterns are analyzed by a two-step pattern mining approach. In the first step the heterogeneous crash data set is partitioned into more homogeneous groups, and important crash severity patterns are mined using the association algorithm for each group. However, large number of association rules are obtained, which are overlapping, and filtering nonoverlapping rules is a challenge. In this study a three-step crash pattern mining approach is proposed, where after obtaining the association rules, important nonoverlapping rules are filtered using the $K$-mode clustering algorithm. This provides a clear set of important nonoverlapping crash severity patterns. Important crash severity patterns for crashes occurring at midblock sections of low-speed urban roads are analyzed using the proposed algorithm. It was observed that heavy vehicle or vulnerable road user involvement play an important role in deciding severity outcomes. Important influence of illegal median cut, shoulder, and marking on crash severity outcomes could be observed.

Abstract

Researchers have proposed modified and staged pattern mining to increase efficiency of obtaining significant patterns. However, many overlapping patterns generated by these methods make filtering of significant patterns difficult. This work proposes a three-stage pattern mining approach that can mine few important nonoverlapping significant patterns sufficient for understanding the process. The proposed approach is illustrated with data for urban road network to understand crash occurrence and severity outcome patterns. Various major insights could be drawn that provide valuable input to safety experts and practitioners for safety improvement of such roads. It was observed that shoulder condition and pavement distress condition play important role in deciding safety. Proper maintenance of shoulder and pavements is essential for safety improvement. Also, illegal median breaks at midblock segments, providing illegal crossing points for vulnerable road users, often witnessed major or fatal crashes. Thus, crosswalks need to be provided at regular intervals. The presence of proper road markings also reduced crash severity. Controlled intersections proved to be safer and when the controls were relaxed during night time, the intersections were observed to witness severe crashes. Also, the proposed three-stage pattern mining method may be used in any field to understand process patterns.