Journal of Transportation Engineering, Part A: Systems

A reasonable method for estimating alighting stops can provide a data guarantee for schedule design and vehicle scheduling. However, few studies, to our best knowledge, have addressed this problem using the variable distance threshold for different types of passengers. This paper is based on available smart card data, and a two-stage algorithm is designed to obtain missing alighting stations in the entry-only urban public transportation system. This algorithm divides passengers into seven types, determines varying distance thresholds for different passengers, and includes five assumptions. Based on the first three assumptions, the corresponding subalgorithms are designed, respectively, constituting the first-stage algorithm, and the first match for all passengers’ alighting stations is carried out. The second-stage algorithm is designed on the basis of the latter two assumptions. A secondary match is conducted on data that were not successfully matched in the first stage to ensure that alighting stops for all passengers are identified. The proposed methods are verified using urban public transport data from the city of Zhuhai as an example. Research results show that the drop-off stops where approximately 74% of passengers can be determined after matching the first-stage algorithm. The distance thresholds for various passengers are different, and the distance thresholds for the same type of passengers over different days are also different. Compared with the traditional fixed distance threshold, the variable distance threshold for various passengers proposed in this paper can improve the match rate of alighting station estimate, about 9%. By appropriately extending the distance thresholds for different passengers, the match rate for passengers who work or live in the suburbs can be further improved, which is in line with the actual travel status of such passengers. The second-stage algorithm can identify alighting sites for a single or unlinked trip. The efficiency of the methodology proposed in this study can meet the needs for practical applications, which may be applied to urban public transport systems.

An algebraic method of arterial progression based on the green center line was presented in this study, which optimized the common signal cycle, phase sequence of each intersection, and offsets by seeking the ideal intersection positions with a small bias distance from the actual intersection positions. Since the bias split would determine the progression bandwidth and impact the progression effect directly, this method took the maximum difference of all bias splits as an evaluation index for scheme optimization. This method was applied to obtain an arterial progression for the 21 signalized intersections on Huanshi West Road in Nansha District, Guangzhou. Compared with the MAXBAND, MULTIBAND model, and the Synchro software, the VISSIM simulation results showed that the algebraic method achieved the best coordinated control effect for the through vehicles along the artery. The schemes obtained by the MULTIBAND model and algebraic method achieved the minimum delay and number of stops for all through vehicles on each road section.

Accurate driver emotion recognition is one of the key challenges in the construction of an intelligent vehicle safety assistant system. In this paper, we conduct a driving simulator study on driver emotion recognition. Taking the car-following scene as an example, the multimodal parameters of a driver in the five emotional states of neutral, joy, fear, sadness, and anger are obtained from the emotion induction experiment and the simulated driving experiment. Wavelet denoising and debase processing are used to reduce the influence of signal noise and the individual differences between drivers. The statistical domain and the time-frequency domain features of the electrophysiological response signals, nasal-tip temperature signals, and vehicle behavior signals are analyzed. The factor analysis method is used to extract and reduce the feature parameters, and the driver’s emotion recognition model is established based on machine learning methods such as random forest (RF), K-nearest-neighbor (KNN), and extreme gradient boosting (XGBoost). Through the verification and the comparison of different modalities and different modality combinations with different machine learning methods, the RF model, based on the feature combination of three types of modal data, has the best model recognition effect. The research results can provide a theoretical basis for driver emotion recognition of intelligent vehicles and have positive significance for promoting the development of human-computer interaction (HCI) systems of intelligent vehicles and improving road traffic safety.

The quality of lane markings is pivotal for safe operations and efficient trajectory generations of connected and autonomous vehicles (AVs). However, most studies are devoted to enhancing in-vehicle detection systems and ignore the impact of faulty lane markings. An instrumented vehicle was employed to mimic the data input of an AV and real-world trials were conducted on (1) live motorways; and (2) a controlled motorway facility. From the live motorway data, causal factors affecting computer vision lane detection and classification algorithms were examined, and an enhanced lane classification algorithm was developed to overcome the limitations posed by poor lane markings. In the controlled motorway facility, experiments to modify the physical appearance of the lane markings were conducted to further test the performance of the developed algorithm. The detection rates of the developed lane classification algorithm were compared with the lane departure warning (LDW) system already implemented in the vehicle. Findings revealed that the LDW system is accurate over 95% and 54% of the time when lanes are faded by 50% and 75% respectively. Further testing on the quality of the lane markings was carried out virtually in such a way that the experiments were replicated in a simulation environment to (1) identify lane marking conditions that can be reliably adopted for safe operations of AVs, (2) estimate the effect of adverse weather and lighting conditions on road markings detection, and (3) address localization issues for AVs. Simulation results show that poor lane markings have a significant negative impact on AV safety, especially in inclement weather and poor light conditions inducing an increase in conflicts and delays. This can be compensated for if more sophisticated sensors are employed in AVs, and the operators of road networks develop lane-based digital road maps.

Real-world transit service often deviates substantially from schedules, leading to variability in route-choice behaviors and experienced travel times that are not adequately reflected in existing accessibility analyses based on general transit feed specification (GTFS) schedule data alone. This study presents a methodology for more accurately assessing origin–destination (OD) transit reliability and accessibility by using GTFS data to generate candidate itineraries, then using automatic passenger counting and automatic vehicle location (APC-AVL) data to determine which routes are selected in real time and calculate the experienced travel time. The methodology is applied across March 2016–2020 from the perspectives of two groups in Allegheny County, Pennsylvania: residents in origins of high socioeconomic need traveling to opportunity employment hotspots, and prospective employers looking to establish a transit-accessible worksite. The worst OD pairs in practice achieved less than 15% of the accessibility captured by 95% of scheduled trips. At the 90-min threshold, the lowest observed accessibility corresponded to a realized loss of 17% of the eligible workforce due to service unreliability. These findings reveal the need to incorporate APC-AVL data to accurately model the route choice process and evaluate rider experiences.

Travelers choose transport modes on the basis of their preferences in which the features onboard vehicles affect the choice of transport modes. In economic theory, it is found that travelers seek to maximize their travel utilities. Autonomous vehicles (AVs) are to be on the market when machines will replace human drivers. AVs are to be used for different purposes, such as transporting passengers and goods. In literature, some research studied AVs in combination with other transport modes, whereas fewer studies have considered the impact of a camera onboard carsharing and privately shared autonomous vehicles (PSAVs) on traveler choices. This research was developed to predict the impact of traveler preferences on the availability of PSAV and carsharing considering certain variables, such as travel time, travel cost, waiting/walking time, and the availability of camera onboard (i.e., as an indication of comfort and privacy factors). A stated preference (SP) survey was designed in which sociodemographic and travel variables were included. A discrete choice experiment (DCE) was designed and included in the SP survey. The SP survey was distributed in Budapest, Hungary, where 313 respondents were collected. The discrete choice modeling approach is used, in which the multinomial logit (MNL) model was applied in developing a transport mode choice model. The developed model estimated the impact of different variables on transport modes as well as the willingness of travelers to use carsharing and PSAVs. The model explained the effects of trip length, income, gender, education, occupation, frequent current transport mode, and trip purpose on travelers’ selection of transport mode. In conclusion, it demonstrated that people tend to use carsharing more than PSAVs, and the availability of cameras onboard decreases the probability of using a transport mode.

For signalized intersections with mixed traffic conditions, pedestrians and bicyclists (P&Bs) affect the vehicle saturation flow rate (SFR), which further influences the cycle length. However, existing studies always consider the effect of P&Bs on vehicle SFR as a constant, which is actually time-variant according to the P&Bs crossing state. Thus, it is necessary to consider this characteristic when determining the cycle length. This study developed a cycle-length model that considers the time-variant effect of P&Bs. The vehicle SFR is first modified using the adjustment factor method. By substituting the modified vehicle SFR into Webster’s delay formula, the numerical solutions of the optimal cycle were derived by considering the minimum vehicle delay as the target. Thereafter, using linear regression of the optimal and minimum cycles, the model was simplified, and an analytical solution of the optimal cycle was derived. The average deviation caused by simplification was 1.95%. The optimality of the model was validated by simulation, with significant reduction in vehicle delay compared with the traditional highway capacity manual (HCM) method.

This study proposes differentiated fees based on parking space attributes to regulate the turnover rate and utilization of parking spaces. To understand parkers’ willingness to pay differentiated parking fees, the study employs structural equation modeling to analyze data collected from a questionnaire that identifies the significant factors impacting parking space selection. The questionnaire also captures personal socioeconomic characteristics of drivers. The study uses K-means clustering to classify drivers based on their varying willingness to pay differentiated parking fees. Based on these results, the study establishes an appropriate price fluctuation range for different parking attributes. The study proposes measures to attract different types of drivers to select differentiated parking fees options. This approach can serve as a valuable reference for parking lot managers seeking to optimize their service capacity and revenue levels. In conclusion, this study presents a comprehensive and systematic approach to understanding the willingness of drivers to pay for differentiated parking fees.

Promoting the safety of commercial trucks by identifying countermeasures that eliminate/reduce the effect of factors that increase the severity of truck-related crashes is crucial. Crash causal factors for rural interstate roads, located within the mountain plains, are inherently unique compared to urban interstate roads. This is due to the presence of challenging road geometry coupled with severe weather conditions and high truck traffic volumes. This study investigated Interstate 80 in Wyoming using decision trees, as a data mining approach, and structural equation model (SEM) as a latent factor modeling approach. SEM was employed to clarify the direct and indirect relationships between endogenous and exogenous variables while accounting for the variation and covariation within and between the constructed measurement models. Crash severity data were processed to account for factors affecting single vehicles and multivehicle trucks. The results showed that the interaction with surrounding traffic was the most significant latent variable affecting the crash severity of multivehicle truck crashes, while adverse weather conditions were the most significant latent variable affecting the crash severity of single-truck crashes. The results of this study highlighted the importance of increasing the situational awareness of commercial truck drivers with upcoming hazardous events. This could be performed by communicating information using variable message signs, the 511 application, the commercial vehicle operator portal (CVOP), or the connected vehicle (CV) technologies.

The curve geometry influences the driver’s speed choice behavior when passing through a horizontal curve. Hence, speed prediction models are essential to ensure highway safety, particularly under heterogeneous traffc conditions prevailing in India, where various classes of vehicles occupy the roadway space. Toward this, the present study aims to develop an operating speed increment and decrement model to identify the influence of the highway geometry variables on the drivers’ speed choice behavior. The National Highway NH-953 is located on hilly terrain, from which 15 curves with various geometric characteristics were selected. The speed data were collected using the radar gun for three dominant vehicle types: motorized two-wheelers (2Ws), cars, and heavy commercial vehicles (HCVs) at three curve points. After that, the two types of speed differences are calculated: the differential of 85th percentile speed ($\mathrm{\Delta }{V}_{85}$) and 85th percentile speed differential (${\mathrm{\Delta }}_{85}V$). The backward stepwise regression (BSR) technique was used on the collected speed data to develop the operating speed increment and decrement models. The results revealed that the curve geometric variables like curve radius, deflection angle, curve length, and tangent length significantly influence the drivers’ speed choice behavior passing through the horizontal curves. Besides, for all three vehicle types, speed increments and decrement models $\mathrm{\Delta }{V}_{85}$ underestimated ${\mathrm{\Delta }}_{85}V$ by 3 to $13\mathrm{km}/\mathrm{h}$.