Travel Mode Choice Model Research Articles

Understanding Influencing Factors of Travel Mode Choice in Urban-Suburban Travel: A Case Study in Shanghai

After the rapid expansion of the subway system over the past two decades, some cities are preparing to build more suburban railways. The emergence of suburban railways is bound to change the choice of suburban passenger transportation. This paper studies the factors that affect the choice of travel mode at the construction stage of suburban railways, aiming to design a more rational suburban railway network and urban public transport service system. Taking Shanghai as an example, this study first surveyed revealed preference (RP) and stated preference (SP) among urban-suburban travelers. Then, we used discrete choice models (DCM) and machine learning algorithms to build a travel mode choice model based on data collection and analysis. Furthermore, the importance of each factor was analyzed, and the effects were predicted under several traffic demand management schemes. Finally, this study proposed some strategies for increasing the share of public transport. On the one hand, it is suggested that Shanghai should continue to develop suburban railways and maintain low pricing of public transport services. Considering the construction and operation costs, the government needs to provide certain subsidies to stabilize prices. On the other hand, as passengers are very sensitive to the “last mile” trips in their suburban railway travel, transport planners should strengthen the connection from and to the suburban railway stations by developing services such as shared bikes and shuttle buses. In addition, the results indicated that some traffic demand management measures can also contribute to a larger share of public transport.

The Potential of Carbon Emissions Reductions of Public Bikes

The reduction of carbon emissions has become a heated background topic in the context of climate change. This paper estimates the potential for carbon reduction from the use of public bikes, on the basis of a travel mode choice model and a carbon emission calculation model. A probability model for the travel mode choice is built to predict travel demands of different modes, and is based on the Logit-based stochastic user equilibrium model. According to this, the generalized travel cost of choosing to walk increases with distance, but the cost of choosing a taxi decreases with distance. When the trip distance is 1.4 km, the walk cost equals to that of the taxi, while if the trip distance is smaller than 1.4 km, the probability of the walk is larger than of a taxi, and vice versa. The case of Ningbo is analyzed. Based on the monthly travel data, the travel characteristics of the public bikes are first analyzed; these indicate that the medium travel distance is 1.44 km, and that the number of trips less than 1.6 km accounts for 70% of all trips. This reveals that the public bike trips are mainly short-distance and in workday rush hour. The related carbon emission reductions of Ningbo on average are 1.97 kg/person and 1.98 kg/km2, and the reductions are positively linearly related to the average hourly total turnover rate, which means the turnover rate is a great parameter to reflect the capability of carbon emission reductions.

The Impact of E-Commerce and Ride Hailing on Emissions from Shopping-Related Transport: A Case Study of the Shopping Habits of University Students from Ningbo University

To clarify the impact of new transport services on consumers’ shopping behaviors and shopping-related transport emissions, a back-propagation neural network shopping channel choice model is established to estimate the number of times that consumers engage in online and offline shopping. A brick-and-mortar store choice model and travel mode choice model are developed, and a method to measure the quality of life of consumers is established to evaluate the impact of new transport services on shopping behaviors and the corresponding shopping transport emissions. The findings reveal that a new passenger transport service increases the number of times that consumers shop in brick-and-mortar stores and correspondingly shopping transport emissions; a new commodity transport service reduces the number of times that consumers shop in brick-and-mortar stores and in turn shopping transport emissions. In a scenario with both new commodity services and new passenger transport services, although online shopping is convenient, consumers are still willing to pay travel expenses for offline shopping; in a scenario with the new commodity transport service but without the new passenger transport service, the emissions from shopping-related transport are the lowest.

Modeling the Behavior in Choosing the Travel Mode for Long-Distance Travel Using Supervised Machine Learning Algorithms

The long-distance travel (LDT) mode choice modeling is important for transportation planners. This study investigated alternative mode choice behavior for the LDT between the intercity buses and trains. A questionnaire survey, consisting of important mode choice attributes, was conducted on various groups of people in Bangladesh. Numerous travel mode choice contributing features (e.g., travel time, travel costs, origin-destination, comfort, safety, travel time reliability, ticket availability and schedule flexibility) were considered and the LDT mode choice models were developed using various machine learning algorithms typically applied for classification problems. With 95.31 % accuracy and 0.95 F1-score, Random Forest model was the best performing model for the dataset. According to the findings of this study, the intercity bus is preferred over the intercity train for LDT in Bangladesh.

Environmental impact analysis of on-demand urban air mobility: A case study of the Tampa Bay Area

Urban Air Mobility (UAM) uses electric vertical take-off and landing (eVTOL) vehicles to provide an alternative way of transporting passengers in urban areas. This study proposes a method of comparing air pollutant emissions from using ground transportation versus switching to multimodal UAM, including greenhouse gases and other air pollutants (NO x , SO 2 , PM 2.5 ). A study by Wu and Zhang (2021) addressed planning questions regarding the optimal placement of vertiports and estimated diverted demand from ground transportation modes by combining network design and travel mode choice models. The proposed method is applied to the case study of Tampa Bay region. A sensitivity analysis was conducted to further evaluate the impacts. The study reveals that on-demand UAM (considering vertiport access and egress modes) generates more greenhouse gases and other air pollutants (except NO x ) in the case study region compared to ground transportation modes. The emissions depend on the structure of power production. A lower UAM service price and more vertiports will further increase emissions.

Mode choice analysis of school trips using random forest technique

Mode choice analysis of school trips becomes important due to the fact that these trips contribute to the second largest share of peak hour traffic. This scenario is more relevant in India, which has almost 265 million students enrolled in different accredited urban and rural schools of India, from Class I to XII as per the UDISE report of 2019-20. Thus, it becomes necessary to understand what mode of transport will be mostly used for school trips in order to design an efficient transportation system. Modal attributes and socio-economic characteristics are mostly considered as explana-tory variables in travel mode choice models. Multinomial Logit (MNL) model is one of the classic models used in the development of mode choice models. These logistic regression models predict outcomes based on a set of independent variables. With the recent advances in machine learning, transportation problems are getting a wide arena of methods and solutions. Among them the method of ensemble learning is finding a prominent place in contemporary modelling. This study explores the potential of using ensembles of random decision trees in mode choice analysis by Random Forest Technique with a comparative analysis on conventional method. It was observed that Random Forest method outperforms MNL method in predicting the mode choice preference of students. The high accuracy of machine learning models is mainly due to its ability to consider complex nonlinear relationship between socio-economic attributes and travel mode choice. These models can learn and identify pattern characteristics extracted from sample data and form adaptive structures through computational process thereby offering insights into the relationships between variables that random utility models cannot recognize. This study considered activity -travel information, personal data and household characteristics of students as attributes for model development and observed that the age of the student and distance of school from home plays a significant role in deciding the mode choice of school trips.

Public Preferences Towards Car Sharing Service: The Case of Djibouti

Transportation demand management is a successful complement to urban infrastructure. The emergence of shared mobility strategies such as car sharing offers sustainable mobility in urban areas. Car sharing has launched in different cities worldwide to mitigate severe transportation problems such as traffic congestion, air pollution, and traffic safety. Therefore, this study aims to investigate the intentions and preferences of travelers toward car sharing services in Djibouti, Africa. The data was collected through an online stated preference (SP) survey. The SP survey included the awareness of car sharing services, attributes related to transport modes, and demographic characteristics. A total of 600 respondents were received. In this study, we employed the multinomial logit (MNL) model to travel mode choice modeling and compared the results with the AdaBoost algorithm. The MNL model results showed that generic attributes such as travel time, travel cost, maintenance charges, and membership fees were found significant. In addition, several demographic characteristics like gender, education, and income were also found significant. The modeling and prediction performances of the MNL model and AdaBoost algorithm were compared using multi-class predictive errors. According to the goodness-of-fit results, the AdaBoost algorithm achieved overall higher prediction accuracy than the MNL model. This study could be helpful to transport planners and policymakers for the implementation of car-sharing services in urban areas.

Travel mode choice in the North-eastern Indian City of Kohima: lessons from empirical study

ABSTRACT One factor that is keeping the urban agglomeration of Kohima at halt is its almost immobile mobility. The same is accentuating air pollution in the city. For effective congestion mitigation, feasible inputs from the respondents should be incorporated in Supply Side Management as well as Demand Side Management to bring about consumer-relevant and effective improvements in urban mobility. This study is therefore envisaged to provide preliminary and original evidence towards addressing that gap. A Logit Travel Mode Choice Model is developed based on Urban Transport Sample Survey. The result suggests that “cost and time” considerations are important for the city commuters; thus, congestion-easing alternative mode that is cheaper and takes lesser time should be provided, and then followed by pricing policy. Regarding trip-related locational factors, it is negative suggesting that the probability of taking buses increases with an increase in the availability of bus stops within reach.

Travel Mode Choice Modeling: Predictive Efficacy between Machine Learning Models and Discrete Choice Model

Background: A complex travel behaviour among users is intertwined with many factors. Traditionally, the exploration in travel mode choice modeling has been dominated by the Discrete Choice model, nonetheless, owing to the advancement in computational techniques, machine learning has gained traction in understanding travel behavior. Aim: This study aims at predicting users’ travel model choice by means of machine learning models against a conventional Discrete Choice Model, i.e., Binary Logistic Regression. Objective: To investigate the comparison between machine learning models, namely Neural Network, Random Forest, Decision Tree, and Support Vector Machine against the Discrete Choice Model (Binary Logistic Regression) in the prediction of travel mode choice amongst Kuantan City. Methodology: The dataset was collected in Kuantan City, Malaysia, through the Revealed/Stated Preferences (RP/SP) Survey. The data collected was split into a ratio of 80:20 for training and testing before evaluating them between the aforesaid models. The hyperparameters of the models were set to default. The performance of the models is evaluated based on classification accuracy. Results: It was shown in the present study that the Neural Network Model is able to attain a higher prediction accuracy as compared to Binary Logistic Regression (Discrete Choice Model) in classifying mode choice of Kuantan users either to choose public transport or private vehicles as daily transportation. Feature importance technique is crucial for identifying the significant features in modelling travel mode choice. It is demonstrated that the Neural Network Model can yield exceptional classification of mode choice up to 73.4% and 72.4% of training and testing data, respectively, by considering the features identified via the feature importance technique, suggesting the viability of the proposed technique in supporting an informed decision. Conclusion: The findings highlight the strengths and limitations of the Machine Learning Technique as well as the Discrete Choice Model in modeling travel mode choice. It was shown that Machine Learning models have the capability to provide better prediction that could assist the urban transportation planning among policymakers. Meanwhile, it could be also demonstrated that the Discrete Choice Model (Binary Logistic Regression) is helpful in getting a better understanding in expressing the inference relationship between variables for improvising the future transportation system.

The psychological decision-making process model of non-commuting travel mode choice under parking constraints

In order to understand the psychological factors that influence car owners’ travel mode selection under parking constraint during non-commuting, a corresponding psychological decision-making process model is established with reference to the stage model of self-regulated behavioral change (SSBC). Partial least squares structural equation model (PLS-SEM) is applied to verify and analyze the model with empirical data. According to the purpose of non-commuting and the degree of dependence on private cars, comparative analyses are made in eight cases. The research results show that SSBC is applicable in the study of non-commuting behavior under parking constraints. The choice of non-commuting mode is influenced by psychological factors such as awareness of consequences, responsibility attribution, personal norms, social norms, perceived behavior control, behavior attitudes, parking supply strategy perception, action plan and coping plan. The psychological decision-making process can be divided into four stages, and the implementation intention, behavioral intention and goal intention are important psychological variables to promote the phase transition. Various psychological factors have different degrees of influence on the behavior decisions of people with different degrees of dependence on private cars under different non-commuting travel purposes (business, medical, shopping and entertainment). According to the analysis results, the paper puts forward some suggestions for parking supply optimization and travel behavior guidance.

Behavior2vector: Embedding Users’ Personalized Travel Behavior to Vector

We investigate how to effectively and efficiently embed users’ personalized travel behaviors to vectors in this paper. Based on an example scenario of travel mode choice in intelligent transportation system, three data structures representing users’ travel behaviors are defined, namely heterogeneous graph of users’ travel behaviors, user travel behavior <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula> -partite graph, and personalized user travel behavior sentence set. This paper systematically analyzes the principle of existing methods and provides intuitions for the problem of learning travel behavior representation in intelligent transportation system. Then we propose the Behavior2vector, which is an improved method tailored for embedding users’ personalized travel behaviors to vectors. In our experiments, we design a travel mode choice model based on machine learning, which uses both hand-crafted basic features and embedded vector features. We further quantify the impact of various factors on travel mode choice and use travel big data to test the hypothesis of traffic assignment models, e.g., travelers always choose the path with the shortest path. In addition, we also compared with the existing graph embedding methods and essentially discussed their advantages and disadvantages.

Mode Choice Modelling Between Bus and Train Under the New Normal Condition

The passenger movements were limited by the government policies that made new system decisions, namely large-scale social distancing policies. However, over time several regions in Indonesia have begun to end large-scale social distancing called new normal. The new normal condition has undoubtedly changed the pattern of mode choice of the passenger. Little attention has been paid to the travel mode choice under the new normal condition. Therefore, this study aims to understand the travel mode choice model of train and bus, especially in the new normal era. The primary data was collected using the stated preference online-based survey. This study performed a Binomial-Logit-Difference model. From the modelling result, 89% of the passenger will choose the bus if the train's travel fare is IDR 160,000 higher. The probability value will be equal when the ticket fare of the bus is IDR 25,000 higher than train’s travel cost. It indicates that people choose the bus mode because of the travel cost factor. Directions for the future study are presented

Analysis of Travel Mode Choice in Seoul Using an Interpretable Machine Learning Approach

Understanding choice behavior regarding travel mode is essential in forecasting travel demand. Machine learning (ML) approaches have been proposed to model mode choice behavior, and their usefulness for predicting performance has been reported. However, due to the black-box nature of ML, it is difficult to determine a suitable explanation for the relationship between the input and output variables. This paper proposes an interpretable ML approach to improve the interpretability (i.e., the degree of understanding the cause of decisions) of ML concerning travel mode choice modeling. This approach applied to national household travel survey data in Seoul. First, extreme gradient boosting (XGB) was applied to travel mode choice modeling, and the XGB outperformed the other ML models. Variable importance, variable interaction, and accumulated local effects (ALE) were measured to interpret the prediction of the best-performing XGB. The results of variable importance and interaction indicated that the correlated trip- and tour-related variables significantly influence predicting travel mode choice by the main and cross effects between them. Age and number of trips on tour were also shown to be an important variable in choosing travel mode. ALE measured the main effect of variables that have a nonlinear relation to choice probability, which cannot be observed in the conventional multinomial logit model. This information can provide interesting behavioral insights on urban mobility.

Optimization Scheme of Tradable Credits and Bus Departure Quantity for Travelers’ Travel Mode Choice Guidance

To analyze the influence of tradable credits and bus departure quantity on travelers' travel mode choice, this study investigated car travel and bus travel as research objects and established a two-mode day-to-day travel mode choice model based on tradable credits and bus departure quantity. To improve the guiding effect of tradable credits and bus departure quantity, an optimization scheme of tradable credits and bus departure quantity was developed with the goal of minimizing the system total travel time of car travel and the system total comprehensive cost of bus travel. Taking a test transportation network as an example, the influence of no tradable credits scheme, tradable credits scheme, and tradable credits and bus departure quantity scheme on the travelers’ travel mode choice behavior was analyzed. The results showed that the tradable credits and bus departure quantity scheme could reduce the saturation of road traffic and improve bus service quality.

Intermodal Connection of High-Speed Rail with Interregional Bus Services in Japan

An intermodal transportation service consisting of high-speed rail (HSR) and an interregional bus service is one policy option for rural areas where interregional travel demand is too low to justify the construction of HSR. This study reviews current interregional bus services connecting with interregional rail, particularly HSR, in Japan, and analyzes the market potential for improving intermodal transportation by integrating HSR with an interregional bus service. It reviews the current interregional transportation network and related travel demand, including for air, rail, and bus. It also analyzes the connectivity of rail+bus intermodal transportation. The analysis showed poor connectivity of HSR and interregional bus services in Japan. Next, an interregional travel mode choice model is estimated with a nested-logit model using data from the Interregional Travel Survey 2010. Then, origin–destination pairs constituting the potential travel demand of the rail+bus option are identified using simple market analysis. This revealed that origin–destination pairs connecting prefectural cities along the Tohoku Shinkansen (HSR) with Kofu City could gain modal shift from other travel modes to rail+bus through improvement in the connection or introduction of a new interregional bus service connected with HSR. Expected changes in modal shares for rail+bus are estimated through a case study where a connection at the HSR station is hypothetically improved by a newly introduced interregional bus service. This suggests that improvements in connectivity at the HSR station could encourage the intermodal transportation service of rail+bus, even for areas not connected with the HSR network.

Nested Logit Joint Model of Travel Mode and Travel Time Choice for Urban Commuting Trips in Xi'an, China

Abstract Commuter trips are an important part of urban travel, and studying the influencing factors, changing rules, and choice behaviors of urban commuting trips is of great significance for optim...

Quantifying Travel Physical Energy Expenditure for Urban Travellers – A Case Study of Beijing

Travel physical energy expenditure for travellers has impact on travel mode choice behaviour. However, quantitative study on travel physical energy expenditure is rare. In this paper, the concept of travel physical energy expenditure coefficient has been presented. A case study has been carried out of young travellers in Beijing to get the value of physical energy expenditure per unit time under three transport modes, walking, car and public transportation. A series of experiments have been designed and conducted, which consider influence factors including age, gender, travel mode, riding posture, luggage level and crowded level. By analysing the travel data of money, travel time and physical energy expenditure, we determined that the value of travel physical energy expenditure coefficient δ is 0.058 RMB/KJ, which means that travellers can pay 0.058 RMB to reduce 1 KJ physical energy expenditure. Next, a travel mode choice model has been proposed using a multinomial logit model (MNL), considering economic cost, time cost and physical energy cost. Finally, the case study based on OD from Xizhimen to Tiantongyuan in Beijing was conducted. It is verified that it will be in better agreement with the actual travel behaviour when we take the physical energy expenditure for different types of travellers into account.

Development and Application of a Travel Mode Choice Model and Policy Implications for Home-to-Work Commuters toward Reduction of Car Trips in Metro Manila

Development and Application of a Travel Mode Choice Model and Policy Implications for Home-to-Work Commuters toward Reduction of Car Trips in Metro Manila

Do air passengers behave differently to other regional travellers?: A travel mode choice model investigation

This research aims to investigate travel mode choices and behaviours of air passengers and community respondents in regional Western Australia. Multinomial logit and Nested logit models were used for the mode choice analysis based on Stated-Preference survey data. The results indicate that travel cost, journey time, service frequency and seat comfort played important roles in affecting travellers' mode choices. For business trips, air passengers are willing to pay more to reduce journey time and increase seat comfort and service frequency compared to community respondents. While for non-business trips, these differences were much smaller. The findings will provide some insights in understanding people's mode choice behaviours and therefore guide policy makers and airlines in developing policies and practice.

The effectiveness of parking policies to reduce parking demand pressure and car use

Evaluating the effectiveness of parking policies to relieve parking demand pressure in central areas and to reduce car use requires an investigation of traveler responses to different parking attributes, including the money and time costs associated with parking. Existing parking studies on this topic are inadequate in two ways. First, few studies have modeled parking choice and mode choice simultaneously, thus ignoring the interaction between these two choice realms. Second, existing studies of travel choice behavior have largely focused on the money cost of parking while giving less attention to non-price-related variables such as parking search time and egress time from parking lot to destination. To address these issues, this paper calibrates a joint model of travel mode and parking location choice, using revealed-preference survey data on commuters to the University of Michigan, Ann Arbor, a large university campus. Key policy variables examined include parking cost, parking search time, and egress time. A comparison of elasticity estimates suggested that travelers were very sensitive to changes in egress time, even more so than parking cost, but they were less sensitive to changes in search time. Travelers responded to parking policies primarily by shifting parking locations rather than switching travel mode. Finally, our policy simulation results imply some synergistic effects between policy measures; that is, when pricing and policy measures that reduce search and egress time are combined, they shape parking demand more than the sum of their individual effects if implemented in isolation.