Transportation Letters-The International Journal of Transportation Research最新文献

The integration of shared e-scooters with public transit is a promising solution for urban mobility's first/last-mile challenge. This study explores spatiotemporal factors influencing this integration, using 35-day e-scooter trip data from Chicago. Employing a random-effect negative binomial approach, we modeled the frequency of e-scooter trips to access/egress to/from bus stops and train stations. Results indicate that weather conditions, design features like intersection density, and multimodal network density significantly influence usage. The transit system characteristics such as service frequency have a positive effect on the integration of e-scooters and trains while a similar effect for bus and e-scooter integration was not significant. Furthermore, safety-related variables such as accident and crime rates as well as demographic characteristics were also revealed to be significant factors in our study. These findings offer vital insights to urban planners and policymakers for infrastructure, safety enhancements, and interventions to encourage efficient e-scooter-public transit integration.

This study proposes a multilayered deep neural network (MLDNN) and a congestion index (CI) based on traffic density factor to forecast traffic congestion directly. Data were collected in Delhi city from a selected location using video cameras during peak hours of weekdays from Monday to Sunday to test the proposed model. Collected data were categorized in a matrix format in the intervals of five-minutes. The input matrix was divided into a number of intervals to train, validate, and test the MLDNN and baseline models, including support vector regression, multi-layer perceptron neural network, gated recurrent unit (GRU) neural network, long short-term memory (LSTM) neural network, convolutional neural network (CNN), CNN-GRU neural network, and CNN-LSTM neural network. Results of the study show that the MLDNN and proposed CI can be applied to predict traffic congestion successfully in heterogeneous traffic.

The COVID-19 pandemic, characterized by travel restrictions and reduced traffic volumes, heightened the risk of severe single-vehicle crashes on Florida's freeways. This study utilized random parameter multinomial logit models, accounting for heterogeneity in means and variances, to analyze driver injury severities in 2020 and compare variations in the magnitude of factors contributed to these injuries across different freeway systems. The estimated models identified 31 statistically significant variables across Florida's major freeways (I-4, I-10, I-75, and I-95). Among these variables, only two—normal driving and restraint usage—were statistically significant across all four freeway systems. Moreover, the model results indicated that factors contributing to severe driver injuries were most prominent on I-95 compared to the other freeways in 2020. These findings improve our understanding of freeway safety measures during a pandemic and provide valuable insights to enhance traffic management strategies for state highway agencies, prioritizing operational safety in potential future pandemics.

This study aims to understand how attitudinal factors influence individuals’ choices for further developing carsharing. To address the gap, a household survey was carried out to obtain the individual’s attitudes toward carsharing. The Transtheoretical Model is used to divide the adoption process of carsharing into five stages, including precontemplation, contemplation, preparation, action, and maintenance. Two Integrated Choice and Latent Variable Models with or without the attitudinal features are evaluated for the effects of latent attitudes on the above five stages. Additionally, the attitudinal factors at each stage are analyzed and compared, and the effects of the flexible and hard measures on the adoption process are discussed. The results indicate that the proposed attitudinal factors well reflect individuals’ adoption stage of carsharing and show different effects on the five adoption stages. The policy implications for facilitating carsharing require coordinating flexible and hard measures in different cognitive stages.

Modeling spatial search processes such as residential location search are challenging, particularly, due to the need to deal with a large dataset and wide array of factors. This introduces a multi-dimensionality challenge to location search modeling. With the motivation to accommodate multi-dimensionality, this paper develops a machine learning–based Gaussian mixture model (GMM) for location search. This study accommodates the effects of several factors including accessibility, land use, dwelling, transportation infrastructure, and neighborhood attributes on location search decisions. The spatial unit of analysis is dwelling-level. This study conceptualizes that households’ search for location based on their reason to move. The pool of alternatives for each household is generated based on probability estimates of GMM. The location choice model considering the reason-based GMM outperforms the model without considering relocation reasons in GMM and random sampling-based model in-terms of predictive performance. The search model has been implemented in an integrated urban model.

The rebalancing of bikes and demand prediction at the station level plays a fundamental role in the regular operation and maintenance of bike-sharing systems (BSSs). In this paper, a novel model which incorporates discrete wavelet transform (DWT), autoregressive integrated moving average (ARIMA), and long-short term memory neural network (LSTM NN), is proposed for BSS station-level check-in/out demand prediction. This study adopts the wavelet analysis method to denoise the raw BSS demand series firstly. Then, DWT is developed to decompose the denoised sequence into three high-frequency components (i.e. details) and one low-frequency component (i.e. approximation). ARIMA and LSTM are employed to forecast the detailed components and one approximation component, respectively. The predicted results of each model are reconstructed into the final outputs by DWT. An experiment on a real-world trip dataset showed that the proposed approach consistently outperforms the standard ARIMA model and LSTM model.

The global COVID-19 pandemic produced several changes in nearly every aspect of our lives. Ride-sharing platforms such as Uber and Lyft must adapt their strategies and aims to stay afloat. The analysis in this study is based on 216,120 tweets in the U.S. between January 1, 2019, and December 30, 2021, about Uber. It includes four separate analyses: Popularity and Usage Analytics, Sentimental Analytics, Voice Analytics, and Topic Mining Analytics. The result shows that usage and popularity of Uber on Twitter negatively affect Covid and death cases. In contrast, vaccination helps mitigate the shock of Covid. Additionally, ’ ‘Uber’s policy and business model was beneficial in improving its positive image during the pandemic; On the early breakout of Covid in the U.S. Uber had a jump on the positive sentiment, mainly because Uber provided safer service than public transportation.

In mixed traffic condition, varying vehicle dimensions and lack of lane discipline lead to parallel movement of vehicles in the same lane or between lanes. This results in a condition where the longitudinal response of vehicles gets affected by adjacent vehicles and their configurations. Correspondingly, the adjacent vehicle configurations are significantly influenced by the lane position of subject vehicle. To accommodate this scenario, the study formulates longitudinal acceleration models from trajectory data considering the influence of subject vehicle’s lane position along with adjacent vehicle characteristics. The response under different cases of subject vehicle’s lane position and adjacent vehicle configurations are evaluated. The statistical analysis indicates that the following behavior varies based on lane position of subject vehicle and adjacent vehicle attributes. It is also found that disregarding these attributes can produce significantly erroneous acceleration estimates. These features can improve existing following behavior models and can enhance the realism of the microscopic modeling scheme for mixed traffic conditions.

To alleviate the performance degeneration of the cooperative adaptive cruising control (CACC) of connected and automated vehicles (CAVs) with low market penetration rates (MPRs), it is necessary to resolve the performance degeneration of CAVs operating at weak mixed-platoon intensities for any-MPR CAVs. A car following model (PMCC) for CAVs is proposed, accounting for human reaction time, communication delay, traffic information memory, and vehicle-to-vehicle (V2V) communication. Linear stability analysis was conducted to validate the stability of mixed platoons at weak platoon intensities for different CAV following models. Numerical simulations were conducted to simulate weak-intensity mixed platoons for different CAV following models operating at high and low equilibrium velocities. The results suggest that PMCC exhibits wider stability equilibrium intervals under different theoretical time headway scenariosand performs well at dissipating the speed fluctuation propagation and minimizing the speed fluctuation amplitude.

When a work zone is located at an intersection, it greatly reduces the capacity and change the traffic flow characteristics. However, the impacts of work zones have not attracted much attention, and the signal control method of the intersection with a work zone has not been investigated yet. This paper focuses on a specific type of work zone, which is located within the area of an intersection. The saturation flow rate model, traffic wave model, traffic delay model, and emptying time model are proposed to capture the traffic flow characteristics of the intersection with an island work zone, and a signal control model is further put forward. The real data is collected to calibrate, validate, and evaluate the proposed models. The results indicate that the proposed signal control model can reduce about 15.6% of queue length and 17.2% of traffic delay for the intersection with an island work zone.