Monash University, Institute of Transport Studies: World Transit Research (WTR)
Not a member yet
11112 research outputs found
Sort by
Origin–destination matrices from smartphone apps for bus networks
No full textThe knowledge of passenger flows between each origin–destination (OD) pair is a main requirement in public transport for service planning, design, operation, and monitoring, and is represented by OD matrices. Although they can be determined by traditional approaches (e.g., surveys, ride-check counts, and/or smartcard-based methods), the availability of new technologies and the proliferation of portable devices triggers an emerging interest in building OD matrices from the apps of bus operators. This research proposes the first framework for the estimation of OD matrices on transit networks by processing smartphone app call detail records (SACDRs). The framework is experimentally tested on a sample of 30 workdays of an Italian bus operator. The results are represented by easy-to-read control dashboards based on maps, which help quantify and visualise the OD matrices in the metropolitan area of Cagliari (Italy). The experimentation shows that the framework can properly estimate the number of trips for both origin and destination w.r.t. OD matrices built from household surveys: the mean absolute error is on average lower than five movements for 90% of the origins and 85% of the destinations
Improving multi-modal transportation recommendation systems through contrastive De-biased heterogenous graph neural networks
No full textConventional uni-modal transportation recommendation systems focused on single modes of transportation are limited in providing satisfactory solutions since passengers often undertake journeys involving multiple modes. Multi-modal transportation recommendation systems are becoming increasingly popular within navigation applications. However, these systems face challenges from biased raw data, data sparsity and long-tail distribution, as well as complexities in representing large-scale graph structures, which collectively hinder their optimal performance. This study introduces a novel approach for enhancing multi-modal transportation recommendation systems: the Contrastive De-biased Heterogeneous Graph Neural Network (CDHGNN). By integrating contrastive learning, the model generates augmented samples to mitigate bias and overcome the data-skewing problem. The heterogeneous graph neural network adaptively captures temporal and spatial patterns among users and locations, as well as spatial adjacency and attribute relations, leading to enhanced representations of nodes, and consequently, improved model performance. The proposed method was evaluated using real-world data from over 300,000 users’ records in Beijing over two months in 2018. The extensive experiments demonstrate that the approach outperforms four contemporary state-of-the-art methods. The results underscore the potential of the CDHGNN in large-scale city-level problems in practical applications, revealing a promising advancement for multi-modal transportation recommendation systems
Origin–destination matrix estimation for public transport: A multi-modal weighted graph approach
No full textEstimating the large-scale Origin–Destination (OD) matrices for multi-modal public transport (PT) in different cities can vary largely based on the network itself, what modes exist, and what traffic data is available. In this study, to overcome the issue of traffic data unavailability and effectively estimate the demand matrix, we employ several data sets like the total boarding and alighting, smart card as well as the General Transit Feed Specification (GTFS) in order to capture the PT dynamic patronage patterns.
First, we propose a new method to model the dynamic large-scale stop-by-stop OD matrix for PT networks by developing a new enhancement of the Gravity Model via graph theory and Shannon’s entropy. Second, we introduce a method entitled “Entropy-weighted Ensemble Cost Features” that incorporates diverse sources of costs extracted from traffic states and the topological information in the network, scaled appropriately. Last, we compare the efficiency of a single travel cost versus various combinations of travel costs when using traditional methods like the Traverse Searching and the Hyman’s method, alongside our proposed “Entropy-weighted” method; we demonstrate the advantages of using topological features as travel costs and prove that our method, coupled with multi-modal PT OD matrix modelling, is superior to traditional methods in improving estimation accuracy, as evidenced by lower MAE, MAPE and RMSE, and reducing computing time
Integrated optimization of train timetabling and rolling stock circulation problem with flexible short-turning and energy-saving strategies
No full textIn daily operations, passenger demand for metro lines traversing city centers often exhibits pronounced tidal characteristics, particularly during morning and evening peak hours. Given the unbalanced spatial and temporal distribution of passenger demand in a bi-directional metro line, this paper investigates an integrated optimization method for train timetabling and rolling stock circulation plans with flexible short-turning and energy-saving strategies. In particular, this approach simultaneously considers constraints such as limited train capacity, turnaround operations, the finite number of available trains, and regenerative energy utilization. Firstly, by introducing decisions involving service frequency, service headway, train route selection, rolling stock circulation plan, and the overlap time indicator, a nonlinear integer programming (NLIP) model is formulated to minimize the weighted sum of passenger waiting time and energy costs, accounting for both passenger and operator perspectives. Subsequently, the model is reformulated into a quadratically constrained quadratic programming (QCQP) model which can be solved directly by commercial solvers. To address large-scale real-world experiments, an adaptive large neighborhood search (ALNS) algorithm is developed. Finally, numerical experiments are conducted on a simplified metro line and Fuzhou Metro Line 1. The results demonstrate that, compared to the full-length strategy, the proposed method reduces total passenger waiting time and energy costs by approximately 8.7% and 5.7%, respectively. Moreover, the methods could support decision-makers with different passenger and operator preferences
Understanding the timing of urban morning commuting trips on mass transit railway systems
No full textThe disparity between rapid urbanization and limited service supplies has raised significant societal concerns, such as overcrowding, caused by a surfeit of individuals traveling at the same time. However, our understanding of how people decide the timing of their trips remains incomplete. Here we use anonymized smart card transaction data from mass transit railway (MTR) systems across three cities to study how commuters schedule travel time to arrive at their workplaces on time. We find two metrics—defined to scale commuters’ time scheduling preferences by investigating relationships among MTR station entry, exit time and work start time—can well indicate arrival penalty risks (early arrival, late arrival, and no penalty), and is common among varying work start times across different cities. Additionally, we explore the varying attractiveness of origin–destination (OD) station pairs to commuters with a rank-flow approach and we develop a realistic determinant to measure the penalty risks with the time reserved for the last-mile trip. Our findings verify theoretical bottleneck models, aid in the understanding of distribution of commuting demand and land uses, and support policy making, such as flexible working-hour policies for peak demand managements
Integrated robust optimization of maintenance windows and train timetables using ADMM-driven and nested simulation heuristic algorithm
No full textThis research paper focuses on the optimization of train timetables and maintenance windows, both of which significantly impact service quality and cost-effectiveness. Uncertainties in both elements can disrupt established transportation plans, causing train delays and maintenance cancellations. Accordingly, we highlight the necessity of augmenting the robustness of these schedules. In this study, we explored an integrated robust optimization of maintenance windows and train timetables using a distributionally robust optimization (DRO) model. The DRO model was established with two types of binary variables and a cross-resolution consistency constraint was introduced to couple them. We innovatively employed a multi-commodity network flow framework to reconstruct the DRO model and designed an alternating direction method of multipliers (ADMM)-based decomposition mechanism. This mechanism was applied to dualize the cross-resolution consistency and track capacity constraints. To handle the problem, we developed a heuristic algorithm driven by ADMM, along with a nested simulation. The algorithm\u27s effectiveness is demonstrated through numerical experiments
Rail accessibility and life satisfaction: Considering residential location and travel ability
No full textThe accessibility of urban rail systems in residential areas significantly impacts residents’ overall life satisfaction. Although many studies have examined the overall impact of rail accessibility on life satisfaction in various developed countries, few have focused on spatial heterogeneity and travel ability between rail accessibility and life satisfaction in developing countries. This study uses data from a 2017 activity-travel survey in Shanghai to examine the relationship between rail accessibility and residents’ life satisfaction in central urban and suburban areas with different travel abilities. It finds that both residential location and travel ability influence the impact of rail accessibility on life satisfaction. There are differences in the thresholds for rail accessibility among residents with different residential locations and private car ownership. These findings suggest that rail transit is a powerful tool for promoting the well-being in developing countries, and proper planning of rail station siting and land use is needed
A data-driven framework for natural feature profile of public transport ridership: Insights from Suzhou and Lianyungang, China
No full textUrban public transport systems, characterised by their complexity, generate vast data sets that pose challenges to traditional analytical methods. To address this issue, our research introduces an innovative natural feature profile framework, leveraging a comprehensive, data-driven approach that incorporates big data, data mining, machine learning, and correlation analysis. This approach provides detailed insights essential for transport planning and policy development. The framework\u27s core is its three-layered structure: the data layer, the feature layer, and the application layer, complemented by a unique four-level feature tagging system. This system investigates correlations, significance, and sensitivities amongst feature tags. It facilitates the extraction of natural feature profiles from voluminous data sets, rendering the framework highly applicable in practical scenarios. The implementation of this framework in Suzhou and Lianyungang demonstrated its adaptability and effectiveness. The findings underscored distinct city-specific transport patterns, highlighting the necessity for customised transport strategies. Furthermore, our framework excels at capturing spatial–temporal dynamics, offering essential insights grounded in evidence. Overall, this paper introduces a methodical, adaptable, and data-oriented framework, signalling a promising future for the development of intelligent and sustainable urban public transport systems
Offline planning and online operation of zonal-based flexible transit service under demand uncertainties and dynamic cancellations
No full textThis paper introduces a comprehensive framework for planning and operating a zonal-based flexible transit (FT) service, a public transit mode designed to accommodate uncertain demand patterns. The framework addresses both offline planning based on stochastic demand distributions and cancellations, as well as online routing considering real-time orders and cancellation behaviour. Offline interzonal route planning is formulated as a two-stage recourse problem, while the intrazonal routing problem is modelled using a Markov decision process (MDP) that incorporates online information. To solve the problem, a service reliability-based decomposition method is employed to divide the problem into three mixed-integer subproblems. The first subproblem focuses on designing interzonal routes up to a specific demand level, taking into account a designated cancellation probability as determined by reliability measures. An insertion heuristic is developed for this subproblem to improve the solution efficiency. The second subproblem allocates passengers from certain categories to vehicles based on the passenger volume designated by reliability measures. Lastly, the third subproblem refines the vehicle intrazonal route according to the passenger assignment from the previous subproblem. The reliability measures are optimised iteratively until no further improvements are observed in consecutive iterations. The proposed FT service’s performance is evaluated using numerical simulations based on real New York City (NYC) taxi demand data, illustrating the effectiveness of the integrated planning and operational approach in accommodating uncertainties in on-demand transit systems
How does the built environment affect intermodal transit demand across different spatiotemporal contexts?
No full textBus and metro are the two primary modes of public transportation in many megacities worldwide. Understanding their cooperation is crucial for the integration of the public transportation system. Despite extensive research on public transportation demand, studies focusing on bus-metro cooperation remain limited. Intermodal transit demand directly reflects the level of cooperation between the two modes in travel behavior. In this study, intermodal transit demand is extracted from smart card data in Beijing, China. The extreme gradient boosting algorithm is employed to investigate the determinants of intermodal transit demand considering spatiotemporal variation. The SHapley Additive exPlanations method further interprets these models. Findings reveal that (1) the relative spatial relationship between bus and metro service facilities significantly influences their cooperation; however, these influences gradually weaken as urban space expands from the core to the peripheral area; (2) in peripheral area, the characteristics of the bus network hold the highest average importance ranking; (3) extensive nonlinear relationships and threshold effects exist between the built environment and intermodal transit demand, with the magnitude, pattern, and direction of these impacts varying significantly across different spatiotemporal contexts; and (4) changes in the spatial layout of transportation service supplies impact their competition and cooperation, such as adequate bus service supplies potentially reducing the cooperation between bus and metro to some extent. These findings will assist planners and public transit operators in developing regulations that encourage cooperation between bus and metro, thereby increasing the attraction and competitiveness of the public transit system