Transportation最新文献

Longitudinal studies have become increasingly popular for investigating changes in behaviour, but present additional challenges around participant recruitment, retention, engagement with survey tasks, additional burden and ultimately data quality. Personal technologies, particularly smartphones, have become integral to tackling these challenges but come with their own caveats around user acceptance and engagement. The current paper investigates these issues in the context of a longitudinal study of interventions designed to encourage use of public transport and increase associated physical activity in Tasmania, Australia. The study comprised multiple waves of data collection over a seven-month period in which travel data were collected using a smartphone app supplemented with user experience surveys. Attrition is lower for older participants, those engaging with the app more, and those responding to the research/environmental/health messaging of the survey as well as the potential for financial gain. App usage is lower among older participants while app engagement is stronger for males, those recording less travel and those indicating environmental reasons as a motivator for completing the study. Experiences with the app are mixed, participants report positive sentiments about the ease of use, hedonic motivation, and help in recalling travel; however, concerns are raised over the accuracy of trip recording, the associated burden of correcting trips, and reductions in smartphone battery-life. Despite the unplanned coincidence with the COVID-19 restrictions, outcomes provide important guidance around recruitment, retention and post-hoc analysis of results from longitudinal studies.

In this study, a new approach will be discussed in which routing is done by predicting future traffic and the learning algorithm is optimized during navigation. Traffic has a complex structure that is constantly changing. Especially for long-term travel, it is not an optimum approach to suggest a route only by considering the traffic situation at the time the navigation request is made. For this reason, the proposed algorithm recommends a route by taking into account future saturation conditions on the vehicle’s route. Singapore was chosen as the study area. The tests were carried out in a simulation environment. The four selected algorithms were tested spatially and temporally. Especially in long-term travels, the superior success of the proposed method compared to other selected methods has been demonstrated.

Walking and cycling are widely encouraged to improve safety, promote health and avoid externalities generated by other transport modes, such as air and noise pollution, and greenhouse gas emissions. Many practitioners and policymakers turn to well-established active mobility cultures, such as the Netherlands, to identify best planning practices. However, walking and cycling rates remain low, and arguments are made that besides built environment characteristics, cultural contexts and social norms are also important in encouraging walking and cycling. While travel behaviour is found to be significantly influenced by socialisation factors (e.g. cultural and social norms), the processes of influence are mediated through an intermediate step of past behaviour. In order to understand the role of socialisation factors in changes towards active travel behaviour a whole view of an individual’s life is therefore needed. This study addresses this research gap by investigating the role of long-term socialisation factors and built environment characteristics in the active travel behaviour of high-income immigrants (e.g. expats) living in the Netherlands, using a qualitative, biographical approach. The findings demonstrate that walking and cycling behaviour can significantly change due to the presence of facilitating factors in the built environment, supportive social networks and the normalisation of walking and cycling as modes of transport. People who have grown up and lived in places with little tradition of walking and cycling, can change their travel behaviour if the environment, both physical and social, makes walking and cycling a viable and attractive option to travel instead of using motorised transportation.

The Covid pandemic from 2020 has affected transport systems worldwide. The British case is examined, drawing on extensive publicly-available data to describe not only impacts on ridership, but also changes in service output, public expenditure, and some indicators of productivity, with particular emphasis on the rail system, and local buses within England outside London. Expectations that the peak would’flatten out’—resulting from the pandemic and working from home—are not supported in the bus case and only partially in the case of rail. Following very large increases in public expenditure to enable continuation of services, that in the bus industry has returned to a broadly pre-pandemic level, while that for rail remains substantially higher. Whilst the pre-Covid cost structures result in a higher degree of short-run escapability for bus, it is also the case that bus has proven to be more flexible in the medium-term, notably in returning to the level of bus-kilometres per member of staff found pre-Covid. Implications for future policy are discussed.

For choice problems in transportation, machine learning and deep learning are alternative methods to traditional choice models. While several works explored the potential of this technology for modelling mode choice, lower attention is given to route choice, especially in public transport. In this work, we propose a deep learning model designed specifically for route choice in public transport. The model can estimate a nonlinear utility function, allowing complex interactions among the variables; it can easily include non-alternative specific variables, such as weather or socio-demographic information. Moreover, compared to the traditional choice models, it numerically outperforms the Path Size Logit Model in prediction performance, and does not require pre-specification of the model by an experienced human modeler. These properties are particularly useful for route choice analyses, to capture possible heterogeneities or complex behavior, which are difficult to model a priori. We evaluated the interpretability of the model observing the marginal rates of substitution and applying Accumulated Local Effects, showing meaningful effects of the variables on the probability to choose an alternative. We tested the proposed model on a large-scale dataset based on GPS tracking. We considered both synthetic choices, to demonstrate the model properties, and real choices, to evaluate the model in practice. The results showed moderately better performance of the deep learning model compared to the Path Size Logit, confirming the possibility of using it for modeling and predicting route choice.

Multimodality has been recognised as a sustainable way of travel, triggering transport policies to seek solutions that facilitate multimodality. However, although emerging mobility services and transport options in the new urban mobility landscape unlock new possibilities for multimodality, little is known about their roles in different forms of multimodal travel. Therefore, this study investigated the forms of multimodality and their relationship with individual travel contexts considering new trends in the urban mobility sector. In the identification of modality styles, a broader and more detailed set of transport modes was considered; and in the definition of individual travel contexts, a series of factors related to the availability and accessibility of transport options and mobility services were considered. Using latent class analysis, this study identified five modality styles including three forms of multimodality that have not been found in previous research. Distinct forms of public transport (bus, tram, metro, and train) were found to be used in conjunction with other transport modes in different ways, leading to different forms of multimodality. Mopeds and motorcycles, rarely considered in previous research, were found to be the primary travel mode for a small group of people. In addition, weighted multinomial logit regression was used to assess the association between individual travel contexts and modality styles. The results indicate that new mobility services, such as (e-)bike-sharing, have the potential to promote more sustainable forms of multimodality that combine active modes with public transport.

This paper reports on a travel survey conducted in Austria in 2019/2020. The aim was to generate 1250 stated preference (SP) interviews using four types of SP experiments, which were based on revealed tours of respondents (tour-based SP-off-RP). The data were to be used as input for a new national tour-based transport model. The core element is a combined time period and mode choice experiment with several innovative new features, which aim to provide a smooth one-stop shop for both stages (RP and SP) and to depict scenarios that are as realistic as possible and achieve sufficient trade-off. The method defined and implemented for the survey is extensively documented, including all steps of survey preparation, the logic behind and development of the time period and mode choice experiment, adaptive measures in survey design and method, and survey conduct. In addition, the paper measures the response rate, describes the data by means of its key features, discusses its representativeness, draws some conclusions on the lessons learned and quality of the data obtained, and provides an outlook on the usage and availability of the data.

Utilitarian cycling speed is a crucial input for applications such as infrastructure design, mode and route choice models, traffic microsimulation, safety evaluations, and health impact assessments. However, current methods fail to distinguish between average speed and cruising speed, the latter of which is more behaviourally indicative. This study aims to identify cruising speed from GPS data and investigate how it varies with contextual and personal factors. We evaluate six algorithms to extract cruising events from cycling GPS travel data: three time series clustering methods to identify steady-state events, in combination with two labeling methods to identify which events represent cruising. The best-performing algorithm uses Toeplitz Inverse Covariance-Based Clustering and identifies cruising events based on a decision tree heuristic. The average cruising speed of 21.53 km/hr is significantly higher than the overall average speed of 19.95 km/hr. Cruising speeds are higher for commute trips, longer trips, e-cyclists, ‘Dedicated’ cyclists, and men. Regarding route factors, cruising speeds are higher in locations with lower grade, more greenery, on-street cycling facilities, high motor vehicle volume, no traffic controls, and lower relative crash risk. Distinguishing cruising events within cycling trajectory data is necessary to avoid underestimating the behavioural sensitivity of cyclists to factors such as road grade, facility type, relative crash risk, trip purpose, gender, and bicycle motorization.

We innovatively shift the research focus from traditional metro systems to the broader spectrum of urban rail transit systems to study the relationship between rail transit development and urban pollution. Previous studies have predominantly concentrated on metro systems, often overlooking the various forms of rail transit such as light rail, trams, and commuter trains, each with distinct environmental impacts. By broadening the scope to include these diverse modes, our research offers a comprehensive analysis of how urban rail transit systems contribute to pollution reduction in Chinese cities. Utilizing panel data from 2011 to 2023, we investigate the effects of rail transit development on air quality, focusing on two primary mechanisms: replacing taxi usage and lowering per capita traffic energy consumption. Our empirical findings, derived from a Difference-in-Differences approach, reveal that the expansion of urban rail transit significantly reduces urban pollution levels. Additionally, we identify variations in effectiveness across different city sizes and regions, with larger cities and eastern regions experiencing more pronounced benefits. These insights underscore the importance of tailoring urban rail policies to local contexts. The study concludes with policy recommendations aimed at maximizing the environmental benefits of urban rail transit systems.

Conventional methods of rail transit passenger flow forecasting usually use general rail transit data for analysis, such as the spatial structure of the network, the distribution of stations, historical passenger flow, etc. However, these methods tend to focus on forecasting regular passenger flow and are insufficient under special events. With the widespread of social media, special events are often disclosed in advance on social media. The attitudes of citizens towards them become an important factor affecting their travel willingness and mode. Existing models usually ignore people’s sentiment, where people’s sentiment tendencies can influence travel destination choices. Particularly during special events, sentiments expressed on social media can trigger short-term sudden changes in passenger flow, which cannot be effectively achieved using traditional automatic fare collection data alone. Therefore, this paper proposes a deep learning-based forecasting model: passenger flow forecasting model for urban rail transit based on multimodal fusion under special events (PFFM-SE), aimed at improving the accuracy of short-term passenger flow forecasting by incorporating social media sentiment data under special events. PFFM-SE includes a travel sentiment analysis, a point-of-interest association, and an outbound passenger flow forecasting. By integrating long short-term memory networks, variational auto encoders, multi-head cross-attention mechanisms, and convolutional neural networks, this model achieves enhanced forecasting of passenger flows augmented with social media sentiment. The experiments used real-world special events social media sentiment and AFC datasets from two cities in China. The results demonstrate that PFFM-SE outperforms various existing advanced models in passenger flow forecasting under special events.