GPS Trajectory Data Research Articles

Predicting the Impact of Disruptions to Urban Rail Transit Systems

Service disruptions of rail transit systems have become more frequent in the past decade in urban cities, due to various reasons, such as power failures, signal errors, and so on. Smart transit cards provide detailed tapping records of commuters, which enable us to infer their trajectories under both normal and disruptive circumstances. In this article, we study and predict the impact of disruptions on commuters and further evaluate the vulnerability of the rail system. Specifically, we define two metrics, stay ratio and travel delay, to quantify the impact, and we derive the predictor of each metric based on the inferred alternative route choices of commuters under disruptive circumstances. We demonstrate that the alternative route choices contribute to more similar feature distribution among different disruptions, which is crucial to tackling the main challenge of abnormal data scarcity and is beneficial for obtaining more reliable predictors for future disruptions. We evaluate our approach with a real-world transit card dataset. The result demonstrates the effectiveness of our method. Based on the predictors, we further analyse the vulnerability of the rail system. An evaluation with cross validation from taxi GPS trajectory data indicates its efficacy in discovering vulnerable rail stations as well as Origin-Destination pairs.

Cluster analysis of microscopic spatio-temporal patterns of tourists’ movement behaviors in mountainous scenic areas using open GPS-trajectory data

Understanding spatio-temporal patterns of tourist movement behaviors is vital for Destination Management Departments (DMDs) on destination planning, marketing, and resource management. This study uses open GPS-trajectory data to analyze the microscopic spatio-temporal patterns of tourists' movement behaviors in Mount Huashan in China. Two major measures, Markov chains and cluster analysis, are used to cluster tourists into groups to show their spatio-temporal movement behaviors within the study site. The Markov chain analysis unveiled three kinds of spatial patterns on microscopic tourists’ movement: “proximity transfer”, “span transfer” and “hub transfer”, while the cluster analysis further demonstrated three kinds of microscopic spatio-temporal patterns: “day-climbing”, “night-climbing”, and “full-day sightseeing”. These results provide vital implications for tourism management of the mountainous scenic areas in general and in Mount Huashan in particular.

Effect of Social Vulnerability on Taxi Trip Times during Hurricane Sandy

The increase in the availability of GPS-based movement data has enabled the exploration of mobility patterns in urban transportation networks. Understanding the relationship between social vulnerability and transportation flows from big data during natural disasters is crucial for utilities and policymakers for decision-making purposes, such as evacuation and restoration planning. In this study, we explore the geographic variation of changes in trip times of taxi trips in New York City (NYC) before and after Hurricane Sandy (2012) using GPS trajectory data in relation to the underlying socio-economic distribution of impacted populations using localized regression technique with GWR. The findings reveal how the spatial patterns of trip change times with respect to SVI, income levels and population density in NYC.

Research on Urban Road Network Extraction Based on Web Map API Hierarchical Rasterization and Improved Thinning Algorithm

Most existing research on the vector road network is based on GPS trajectory travel information extraction, and urban GPS trajectory data are large and difficult to obtain. Based on this, this study proposes a road network extraction method based on network map API and designs a vector road network based on an improved image-processing algorithm using trajectory data. Firstly, a large number of trajectory data are processed by hierarchical rasterization. The trajectory points of the regional OD matrix are obtained by using the map API interface to generate the trajectory. Then, the image expansion processing is performed on the road network raster image to complete the information loss problem. The improved Zhang–Suen refinement algorithm is used to refine the idea to obtain the road center line, and the vector road network in the study area is obtained. Finally, taking the Harbin City of Heilongjiang Province as an example, compared with the road network of the network map, it has been demonstrated that using this technology may improve the traveler experience and the sustainability of urban traffic flow while reducing the number of manual procedures required, performing online incremental rapid change detection, and updating the present road network at a cheaper cost.

Towards renewable public transport: Mining the performance of electric buses using solar-radiation as an auxiliary power source

• Developing a framework to estimate solar PV potential of urban fleet-wide electric buses. • Publicly street-view panoramas, GPS trajectory data and DEM data are used as input. • A case study of Qingdao, China with 547 bus routes, 28,661 street-view panoramas is studied. • Benefits under different driving conditions and operation schedules are evaluated. Transforming the road public transport to run on renewable energy is vital solution to achieve carbon neutral and net zero goals. This paper evaluates the potential of using solar radiation-generated electricity as an auxiliary power supplementary for the battery of electric buses, based on a developed framework that using publicly street-view panoramas, GPS trajectory data and DEM data as input parameters of solar radiation model. A case study of Qingdao, China with 547 bus routes, 28,661 street-view panoramas shows that the solar-radiation electricity generated at noon during the operation accounts for about one-fifth, one-eighth of the total electricity consumption of a bus traveling one kilometer in a sunny day and a cloudy day, respectively. Spatial variability shows significant solar-radiation power generation advantages in newly-launched areas and expressway. The solar power generated in a sunny day can make a bus half of passengers and with air conditioner off at least one extra trip in 2:1 replacement schedule, and nearly close to one extra trip in 4:3 replacement schedule. A correlated relation between the solar-radiation power generation benefit and the operation schedule of electric buses is observed, implying that the high cost of 2:1 replacement schedule for long-distance routes during summer or winter can be reduced. The proposed framework can help us evaluate and understand the feasibility of solar radiation-generated electricity energy of electric bus fleets covering the large-scale urban areas at different times, locations, and weather conditions, so as to support effective decisions at better planning of PV-integrated electric buses.

Exploring the Spatial Characteristics of Stay Areas in Walking Tours through the Lens of Volunteered GPS Trajectories: A Case Study of the Zhuhai–Macao Metropolitan Area

The stay areas in walking tours are the service and management unit of recreational walking in metropolitan areas. The rational characterization of stay areas in walking tours is of great significance for developing local tourism, constructing appropriate public facilities, optimizing the configuration of tourist elements, and improving facility efficiency. The existing research focuses mainly on functional, top-down classifications of tourism, tourist behavior patterns, and route designs, but it has left tourists’ stay areas largely unaddressed. To fill this gap, we propose a new framework for the interpretation of stay areas in walking tours based on GPS trajectory data and accompanying photos uploaded by users. Taking the Zhuhai–Macao metropolitan area as an example, we first captured the stay points and clustered them to the walking tour stay areas using DBSCAN. The characteristics of the stay areas were then collected, and a hierarchical analysis was conducted in terms of spatial features and geotagged photos. The results show that the stay areas can be grouped into six categories displaying obvious differences in spatial distribution, landscape features, and tourist activities. We also found the connections between Zhuhai City and the Macao Special Administrative Region (SAR) to be relatively weak. In conclusion, our results can contribute to tourism planning as well as the further management and allocation of recreational service facilities in the area researched.

Temporal Aggregated Analysis of GPS Trajectory Data Using Two-Fluid Model

The market for on-demand mobility services is growing worldwide. These services include, for example, ride-hailing, ride-sharing, and car-sharing. Large-scale fleets of such services collect GPS trajectory (probe vehicle) data constantly everywhere in the network. At a certain penetration rate, this data becomes representative of the entire road network. It can give valuable insights into traffic dynamics and the evolution of congestion. In this paper, we use such GPS trajectory data from Chengdu, China, to investigate the stability and recurrence of macroscopic traffic patterns. Using the two-fluid theory, we find that the two-fluid coefficients are robust on between-day variation, not only supporting the theory itself but also emphasizing that the general evolution of traffic is a robust pattern. We investigate the deviations from the model using time series analysis of the residuals of the two-fluid model. Here, we find evidence for daily and weekly seasonality in the residuals, indicating that congestion patterns are convincingly recurring. These patterns can be used for network-wide traffic state prediction. We conclude that GPS trajectory data from large on-demand mobility fleets is a promising data source for observing traffic patterns in urban road networks once the data becomes representative.

Exploring the Cascading Failure in Taxi Transportation Networks

To explore the ability of taxi transportation service capacity in unexpected conditions, based on the taxi GPS trajectory data, this paper presented a taxi transportation network and explored a cascading failure model with the non-linear function of traffic intensity as the initial load. Moreover, the cascading failure conditions for different initial loads with different parameter settings were derived by combining the complex network theory. We verified the ability of taxi transportation networks to withstand unexpected conditions and analyzed the differences and features of taxi transportation service capacity for different areas of Lanzhou city. Three sets of comparative simulation experiments were implemented. The results show that when the initial load regulation factor α<1/θ, the failure of nodes with smaller initial loads in the network is more likely to cause cascading failure phenomena. When α>1/θ, the failure of nodes with larger initial loads in the network is more likely to cause cascading failure phenomena. Additionally, when α=1/θ, there is no significant correlation between whether cascading failure phenomena occur in the network and node loads. This study can provide a prior basis for decision-making in the management of urban taxi operations under different passenger flow intensities.

CLUSTMOSA: Clustering for GPS trajectory data based on multi-objective simulated annealing to develop mobility application

Mobility analysis is the core idea of many applications such as vehicle navigation, trajectory analysis, POI recommendation, and traffic flow analysis. These applications collect huge spatio-temporal information represented as trajectories of a moving object such as a vehicle or people using Global Positioning System enabled devices. Various techniques are evolved to process, manage and extract useful information from trajectories. Among these techniques, clustering plays an important and integral role in developing various mobility applications. Popular traditional clustering techniques such as DBSCAN, K-means, OPTICS, hierarchical clustering, and DJ-clustering are used for this purpose. However, these techniques suffer from major issues such as entrapping in local optima and being less effective in varying densities. Further, these methods have low search capability in search space, work upon single criteria optimization, and are less scalable for the big dataset. To overcome these issues, a new multi-objective criterion-based evolutionary clustering termed CLUSTMOSA is proposed. It exploits the search capability of archived multi-objective simulated annealing (AMOSA) to cluster the dataset. It stabilizes the exploratory and exploitative behavior of the solution. In this paper, three clustering evaluation metrics are simultaneously exploited as objective functions of CLUSTMOSA. Also, a new segmentation method is presented using bearing measurement for trajectory data. It helps to eliminate multiple waypoints localized over the straight roads and prevents multiple cluster formations for the same segment. To investigate the performance, the proposed CLUSTMOSA, along with a new segmentation method using bearing measurement is compared with the state-of-art methods of trajectory data mining. The extensive experiments and analysis prove the superiority of our clustering model over state-of-art approaches.

Do Historic Landscape Images Predict Tourists’ Spatio-Temporal Behavior at Heritage Sites? A Case Study of West Lake in Hangzhou, China

Rich in history and culture, heritage sites often evoke stirring emotions and memories. We analyzed historical poetry using grounded theory and high-frequency word and semantic analysis to construct historic landscape images (HLIs) of the West Lake UNESCO World Heritage Site in Hangzhou, China. GPS trajectory data were used to identify hot and cold tourist spots and understand the site’s intra-attraction tourist behaviors (IATBs). Finally, we analyzed the HLI–IATB relationship. We found that the tourist distribution was uneven, and different attractions had different visitor behaviors. Our findings should inform future heritage site management—and neighboring cities—about the benefits of using HLIs to predict attraction visitors’ behaviors and leveraging those insights to optimize multiple-attraction sites proportionally. Such projections can provide new perspectives for heritage studies, landscape planning, and tourism image-making.

Location Selection of Charging Stations for Electric Taxis: A Bangkok Case

The transition from ICE to BEV taxis is one of the most important methods for reducing fossil fuel consumption and air pollution in cities such as Bangkok. To support this transition, an adequate number of charging stations to cover each area of charging demand must be established. This paper presents a data-driven process for determining suitable charging locations for BEV taxis based on their characteristic driving patterns. The location selection process employs GPS trajectory data collected from taxis and the locations of candidate sites. Suitable locations are determined based on estimated travel times and charging demands. A queueing model is used to simulate charging activities and identify an appropriate number of chargers at each station. The location selection results are validated using data from existing charging services. The validation results show that the proposed process can recommend better locations for charging stations than current practices. By using the traveling time data that take the current traffic condition into account, e.g., via Google Maps API, we can minimize the overall travel time to charging stations of the taxi fleet better than using the distance data. This process can also be applied to other cities.

Estimating fundamental diagram for multi-modal signalized urban links with limited probe data

Being one of the most classic concepts in the traffic flow theory, the fundamental diagram (FD) describes the relationship between average flow and average density of link-level traffic flow dynamics. Inductive loop detectors or closed-circuit television are commonly used for FD estimations and they are known to have cost-effective and accuracy issues. Thanks to the GPS-enabled smartphones and GPS-equipped probe vehicles, high temporal and spatial resolution traffic data are available which enable traffic condition inference over time and space continuously. Several existing studies have explored FD estimation algorithms on freeways where flow is generally uninterrupted and uni-modal, based on GPS trajectory data. These developments motivate this study, where the objective is to extend the application to multi-modal and interrupted environment, i.e., urban signalized areas. In this paper, an estimation method is developed to capture the FD of multi-modal traffic streams on signalized urban links. The proposed algorithm is empirically tested using real-world GPS datasets collected on a signalized arterial road in Shenzhen City. Promising results show that the proposed algorithm is capable to estimate the FD under such condition. Furthermore, impacts of multi-modal traffic and signal operations on the FD estimation are analyzed and discussed.

Trajectory Clustering in an Intersection by GDTW

GPS trajectory data in intersections are series data with different lengths. Dynamic time wrapping (DTW) is good to measure the similarity between series with different lengths, however, traditional DTW could not deal with the inclusive relationship well between series. We propose a unified generalized DTW algorithm (GDTW) by extending the boundary constraint and continuity constraint of DTW and using the weighted local distance to normalize the cumulative distance. Based on the density peak clustering algorithm DPCA using asymmetric GDTW to measure the similarity of two trajectories, we propose an improved DPCA algorithm (ADPC) to adopt this asymmetric similarity measurement. In experiments using the proposed method, the number of clusters is reduced.

Discovering Implicit Working Pace of Online Ride-Hailing Drivers: An Exploratory Study

Due to convenience and flexibility, online ride-hailing has become increasingly more prevalent across the world. However, many violations and road crashes involving online ride-hailing were related to the unhealthy working pace of drivers, especially inadequate rest. This paper enriches the literature by providing a first look into the latent break patterns of online ride-hailing drivers based on a one-month order record dataset. A data mining and knowledge discovery process is presented for extracting and analyzing characteristics of online ride-hailing drivers’ work and rest based on GPS trajectory data, as follows: 1) logical judging to identify non-work order-gaps; 2) dynamic topic modeling to discover latent break patterns; and 3) integrating the topic modeling results with feature analysis results of order-gaps to summarize the time-dependent characteristics of online ride-hailing drivers’ special working pace. The case study results show that the latent break patterns extracted from two cities’ online ride-hailing order records are significantly different in the strength and cycles of the topics, which is greatly related to the travel supply-demand conditions and urban characters. Furthermore, the proposed analytical framework can help mobile transportation platform companies to better understand online ride-hailing markets from the perspective of drivers and to adjust their marketing strategies in real time.

Traffic Volume Estimate Based on Low Penetration Connected Vehicle Data at Signalized Intersections: A Bayesian Deduction Approach

The emergence of connected vehicle (CV) technologies has created new traffic control opportunities, among them, is the potential to estimate volume without approach lane detection. Rather than requiring the expense and effort to install and maintain detector systems, this new “detector-free” method permits traffic volume to be estimated from CV GPS trajectory data. Unfortunately, however, CV GPS methods are limited not only to locations where CV GPS data can be recorded, but also limited to time when CV GPS data is recorded. The goal of this research was to overcome these limitations and permit volume estimation to be accomplished under any location or condition, including low-penetration CV environments. The contributions made by this work are significant in two respects. First, it creates an improved queue-based method to estimate intersection approach volumes during each signal cycle with sparse CV data. Second, the research demonstrates the application of a Bayesian deduction method to approximate volume with no CV trajectory data. To accomplish this, traffic volumes are assumed to be time-dependent Poisson distributed throughout the day, and CV data were used to estimate CV volume and further set as prior to deduce the time-dependent Poisson arrival rate. To verify and evaluate the accuracy and effectiveness of this new method under a range of potential traffic conditions, a simulation case study and a NGSIM case study were implemented. Results of both case studies resulted in estimated-to-actual arrival rate average errors as low as 4.2 percent and volume estimation errors as low as 0.9 percent.

Towards big data behavioral analysis: rethinking GPS trajectory mining approaches from geographic, semantic, and quantitative perspectives

The question regarding the actual usage of built environments is of immense importance in behavioral research. Yet traditional methods of collecting and analyzing data on movements and activities often lack needed accuracy and granularity. Thus, this article reviewed and summarized the applicability of emergent GPS trajectory mining approaches in the field of architecture from geographic, semantic, and quantitative perspectives, respectively. Accordingly, three experiments based on a case study using real GPS trajectory data from visitors to the Palace Museum in China were conducted to examine the usefulness and weakness of the aforementioned approaches. The findings revealed that although all three dimensions of the trajectory mining approaches had the potential to provide useful information for architectural and urban design, the higher the dimensionality in utilizing the data, the more effective the approach was in discovering generalizable knowledge of human behavioral pattern. Furthermore, the results suggested that to gain insights into the typological characteristics of human behaviors related to the built environments, the contribution of trajectory data alone was limited, hence, conventional field surveys and questionnaires which contain information on individual characteristics and spatial features should be used in conjunction. Future research and practical implications were outlined.

Application of Clustering Algorithms in the Location of Electric Taxi Charging Stations

The reasonable layout of charging stations is an important measure to improve the penetration rate of the electric taxi market. Based on the multi-type clustering algorithm, a widely applicable electric taxi charging stations locating method is proposed. By analyzing the massive gasoline taxi GPS trajectory data, the parking information and charging requirements of electric taxis are extracted, and the research area is divided into reasonable grids. Then, the divided grids are respectively subjected to multiple same-type clustering and multiple multi-type clustering algorithms, so as to help find out the location of the charging station, and a comparative analysis is performed. The empirical analysis shows that the positioning results of the multiple multi-type clustering algorithms are more reasonable than the multiple same-type clustering algorithms, which can effectively prolong the driving distance of electric taxis and save the travel time of drivers.

CrowdExpress: A Probabilistic Framework for On-Time Crowdsourced Package Deliveries

Most of current urban logistic systems fail to strike a nice trade-off between speed and cost. An express logistic service often implies a high delivery cost. Crowdsourced logistics is a promising solution to alleviating such contradiction. In this article, we propose a new form of crowdsourced logistics that organizes passengers and packages in a shared room, i.e., using taxis that are already transporting passengers as package hitchhikers to achieve on-time deliveries. It is well-recognized that taxi drivers are good at delivering passengers to their destinations efficiently. As a result, the proposed new urban logistics system has potentials to lower the cost and accelerate package deliveries simultaneously. Specifically, we propose a probabilistic framework containing two phases called CrowdExpress for the on-time package express service. In the first phase, we mine the historical taxi GPS trajectory data <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">offline</i> to build the package transport network. In the second phase, we develop an <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">online</i> taxi scheduling algorithm to adaptively discover the path with the maximum arriving-on-time probability “on-the-fly” upon real-time passenger-sending requests, and direct the package routing accordingly. Finally, we evaluate the system using the real-world taxi data generated by over 19,000 taxis in a month in the city of New York, US. Results show that around 9,500 packages can be successfully delivered daily on time with the success rate over 94 percent.

Turn-level network traffic bottleneck identification using vehicle trajectory data

Identifying traffic bottlenecks is a prerequisite to alleviate traffic congestion in urban networks. However, the state-of-the-art methods for bottleneck identification stay at the segment level which assume bottlenecks only result from congested road segments (CRSs). The possibility that a turning direction in a road segment could be a bottleneck has not been thoroughly investigated. In addition, most existing techniques only focus on the congestion degree of road segments, however ignoring the correlations between CRSs and their underlying congestion propagations. This study proposed a framework for turn-level bottleneck identification in large-scale road networks using vehicle trajectory data. First, filtered GPS trajectory data was used to identify CRSs and their congestion start time. Then, congestion correlation graphs and corresponding spanning trees were constructed by investigating forward and backward correlations between CRSs. These directional correlations were later modeled via a Bayesian inference approach. Last, congestion correlation cost models were built up to identify turn-level bottlenecks considering the probability of congestion correlation. Both simulation and field experiments were conducted to evaluate the performance of the proposed framework and compare with the state-of-the-art methods. Results reveal that our framework can effectively capture directional correlations between CRSs, and successfully identify turn-level bottlenecks in large-scale networks.

Joint prediction of zone-based and OD-based passenger demands with a novel generative adversarial network

Online ride-hailing plays an important role in modern urban transportation systems, and accurate short-term passenger demand prediction contributes to improving ride-hailing services. Many existing studies have made great achievements in zone-based demand prediction. However, origin–destination (OD)-based demand prediction has attracted little attention, even though it provides abundant information and is of high practical importance, e.g., it facilitates the routing and matching of ride-hailing services. Considering the importance of both types of demands, a conditional generative adversarial network (CGAN) with Wasserstein divergence (CWGAN-div) is proposed to jointly predict zone-based and OD-based demands. Residual blocks are utilized to capture internal spatiotemporal features, which benefit the training and prediction processes of the CWGAN-div model. Conditional information is also incorporated to characterize the external dependencies of the demands. Numerical experiments are performed by using GPS trajectory data from Didi Chuxing, Chengdu, China. The results show that the proposed CWGAN-div model yields good joint prediction performance and outperforms both classic models (historical average (HA) and convolutional neural network (CNN) models) and other prevailing GAN models (i.e., GAN, CGAN, Wasserstein GAN with a gradient penalty (WGAN-GP), and GAN with Wasserstein divergence (WGAN-div)). The proposed CWGAN-div model displays promise for the joint prediction of zone-based and OD-based passenger demands.