Journal of Transportation Engineering, Part A: Systems

Public transit is a critical component of transportation planning. One important task is to optimize the public transit service for people by improving current service or having new investments. Under such circumstances, the main objective of this study is to develop optimization models for improving transit equity and accessibility for people by integrating the transit gap index (TGI), a performance metric using the General Transit Feed Specification (GTFS) data. TGI was developed by considering demographic features, as well as spatial and temporal transit service characteristics. Two models were built with the intention to optimize the transit equity by mitigating the transit deficiency (i.e., TGI). Two different conditions were considered while developing both models: (1) maximizing the level of transit services under the constraint of a limited budget that will be invested in the new constructions of public transit stops; and (2) minimizing the total cost for constructing new public transit stops with constraints on certain improvements of transit services for a certain amount of block groups. Finally, a case study in the City of Charlotte, North Carolina, was conducted, and comprehensive numerical results and analyses based on the proposed models are provided.

The lane-changing behavior of different types of vehicles at the merging section of an urban road is assessed by using a new lane-changing model proposed in the current work. The lane-changing model known as MOBIL (minimizing overall braking induced by lane changes) is modified and combined with the intelligent driver model (IDM) to implement the lane-changing rules for different vehicle classes by applying the politeness and vehicle-type factors. It is noted that the heterogeneity of surrounding vehicles during lane-changing, which is common in developing countries, is not considered in most of the existing lane-changing models. The preceding problem is addressed by incorporating the vehicle-type dependent factor in the proposed lane-changing model. This new model considers the effect of motorcycle movement on the lane-changing decision and evaluates the aggressiveness of motorcyclists during lane-changing. The merging maneuver data from video recording is utilized to calibrate and validate the models. The lane changing rate is the highest for motorcycles and the least for trucks (at a given politeness factor). Motorcycles exhibit lower lane changing durations for politeness factors $p=0$ and $p=1$, showing integrated movements due to their pushy or erratic maneuverability.

The statistical models supporting the Highway Safety Manual quantify associations between aggregate traffic measures, such as average daily traffic volume or posted speed limit, and crash frequencies accumulated over several years. For some time though, it has been recognized that crash risk can vary as traffic conditions vary due to special events or within-day changes in traffic. Additionally, the Highway Safety Manual’s predictive tools are essentially statistical summaries of conditions present during the recent past, and transferring this knowledge to environments containing automated vehicles is likely to be problematic. This paper illustrates how both issues can be addressed by supplementing standard statistical modeling with models describing crash mechanisms. In particular, Brill’s random walk model of how traffic shockwaves generate rear-end crashes is combined with a traffic flow model based on a fundamental diagram in order to quantify the relation between traffic density and rear-end crash risk. Approximating Brill’s random walk with a finite Markov chain leads to a computationally tractable model, and the model’s predicted relationship is consistent with empirical findings. Transferring the model to a hypothetical environment with automated vehicles is then illustrated.

Aiming to analyze the waiting decision behavior of commuters for bus transits at a bus station of departure during morning peak hours, this paper builds a bus-waiting decision model based on the prospect theory, with consideration of commuters’ total travel costs comprehensively. Taking the total costs of taxi carpooling for individuals as the reference point for travel costs, the travel prospect value models of commuters choosing $K_1$ bus and $K_2$ bus were established, respectively. The results of the numerical example show that the best choice of commuters is closely related to the bus departure interval and the in-vehicle travel time. This paper constructs a two-variable dynamic selection model to explore the decision-making behavior of commuters, with bus departure intervals and in-vehicle travel times as core variables. The research result can help the public transportation management department adjust departure intervals, as well as improve the commuters’ travel satisfaction and the utilization efficiency of the public transportation system.

Driver’s perception error plays a significant role in determining the dynamical properties of a platoon of vehicles driving on a straight road. Therefore, we attempt to understand the role of driver’s perception errors for vehicle motion state information in affecting traffic flow oscillations in this study. A real vehicle test experiment indicates that the drivers have different levels of perception errors for velocity and headway of the preceding vehicle in the car-following process. Based on this, an extended full velocity difference (EFVD) model is used to investigate the effects from the driver’s perception errors of the preceding vehicle’s velocity and headway changes on a road without lane-changing. We incorporated the perception errors of velocity and headway into the EFVD model, and constructed a new EFVD model considering the driver’s perception error (EFVDDPE). Moreover, the confidence levels of perception errors for velocity and headway are introduced. Results from numerical experiments illustrate that the increase of the interval of the confidence levels is not conducive to smooth traffic flow in those cases. In order to eliminate the impact of stochastic fluctuation of the confidence levels on traffic flow stability and car-following safety, a delayed feedback control is proposed. Numerical simulations are also carried out to verify its effectiveness in improving the stability of the stochastic EFVDDPE model. These results are useful in setting an effective control method for adaptive cruise control (ACC) or intelligence driver systems to stabilize traffic flow in the platoon when perception errors exist in the vehicle-mounted sensors.

Developing efficient operational strategies to improve service quality of bus transit, such as reducing travel time, can stimulate ridership. A mathematical model is formulated to optimize integrated bus service which maximizes ridership considering demand elasticity with respect to travel time and fare. The proposed integrated service, consisting of local (e.g., all-stop) and express (e.g., stop-skipping) services, is optimized using a genetic algorithm (GA) subject to minimum service frequency and fleet size constraints. A numerical analysis is conducted under various operation scenarios based on a real-world bus route in Chengdu, China. The results suggest that the optimized integrated service may increase the ridership. The sensitivity analysis is conducted, and the impacts of model parameters on decision variables to the ridership are explored.

Lane-based driving (lane-keeping) problems generally prevail in undeveloped and developing countries, where drivers’ lane-keeping behaviors are fairly weak for reasons such as poor surface conditions, undisciplined driving, low-quality lane line visibility, inappropriate number of lanes and lane width, and nonexistence of raised lane lines with vibration. Drivers’ undisciplined lane-based driving behaviors generally cause an increase in dangerous lateral interactions between vehicles. Thus, problems such as irregular lane utilization, difficulties in traffic management, arbitrary lane changing, and illegal overtaking can be observed on divided multilane urban roads. It is necessary to investigate lane-keeping violations on urban roads to find causal parameters and thereafter formulate suitable solutions. This studied aimed to model lane-based driving behavior by considering the parameters of road geometry (e.g., lane width and number of lanes), vehicle type (passenger car, minibus, light commercial vehicle, midibus, bus, truck, lorry) and traffic flow (volume, heavy vehicle ratio, speed, vehicles in adjacent lanes, and roadside parking) on major urban divided roads in three Turkish cities. For this purpose, lateral position data were collected. A probit regression (PR) analysis was conducted to model the relationship between lane-based driving behavior and the examined parameters. It was found that lane width and lane volume have a positive effect, and heavy vehicle ratio (%), vehicle width, speed, and number of lanes have a negative effect as covariates of the proposed model. Numerical tests also showed the developed PR model to be statistically significant and to have a good ability to fit the observed and predicted data for the examined parameters.

The coronavirus disease of 2019 (COVID-19) pandemic has affected every aspect of peoples’ lives, including their mobility. In this study, the impact of closures related to the pandemic on traffic patterns is assessed for the state of North Carolina and the Commonwealth of Virginia. The results of the initial assessments indicated that the average reduction in traffic volume was 27% across North Carolina and Virginia. Findings suggest that a slight increase in traffic at the beginning of the stay-at-home orders was followed by at most a 40% reduction after closures. The changes in traffic volumes in terms of road and vehicle classification are most substantial in Interstate routes and for passenger cars and buses. The average recovery rate of traffic volume after the occurrence of the maximum reductions was between 2.3% and 3.4% per week.