Taxi Trajectory Research Articles

Spatial heterogeneity and migration characteristics of traffic congestion—A quantitative identification method based on taxi trajectory data

It is of great reference significance to exploring spatial dependence of urban traffic activities and researching internal causes of regional traffic state changes for road network optimization and residents’ travel behavior analysis. Based on trajectory data of taxis in Ningbo city of China, this study calculates average driving speed of taxis in different blocks during characteristic period and generates the global Moran’s I and the LISA clustering diagram. On this basis, the spatial clustering characteristics of congestion on working days and non-working days are analyzed. Furthermore, in order to further characterize the changes of congestion from the perspective of spatial migration, a method of measuring geometric displacement is adopted to describe spatio-temporal migration trend of traffic states, four indicators designed to identify urban frequently congested areas, including migration direction, angle, distance, and low-value area. The results show that the high-clustering area are located urban fringe and the low-clustering area are located at geometric center of major urban areas. Spatial–temporal migration law of low-value areas in city-center is obvious. Difference between trend is compared with non-working days, the offset and azimuth of low-value area in downtown on working days are even bigger. The accurate capture of the characteristics of congestion space migration at the urban scale will help to formulate more targeted congestion management strategies.

Smart City Taxi Trajectory Coverage and Capacity Evaluation Model for Vehicular Sensor Networks

In a smart city, a large number of smart sensors are operating and creating a large amount of data for a large number of applications. Collecting data from these sensors poses some challenges, such as the connectivity of the sensors to the data center through the communication network, which in turn requires expensive infrastructure. The delay-tolerant networks are of interest to connect smart sensors at a large scale with their data centers through the smart vehicles (e.g., transport fleets or taxi cabs) due to a number of virtues such as data offloading, operations, and communication on asymmetric links. In this article, we analyze the coverage and capacity of vehicular sensor networks for data dissemination between smart sensors and their data centers using delay-tolerant networks. Therein, we observed the temporal and spatial movement of vehicles in a very large coverage area (25 × 25 km2) in Beijing. Our algorithm sorts the entire city into different rectangular grids of various sizes and calculates the possible chances of contact between smart sensors and taxis. We further calculate the vehicle density, coverage, and capacity of each grid through a real-time taxi trajectory. In our proposed study, numerical and spatial mining show that even with a relatively small subset of vehicles (100 to 400) in a smart city, the potential for data dissemination is as high as several petabytes. Our proposed network can use different cell sizes and various wireless technologies to achieve significant network area coverage. When the cell size is greater than 500 m2, we observe a coverage rate of 90% every day. Our findings prove that the proposed network model is suitable for those systems that can tolerate delays and have large data dissemination networks since the performance is insensitive to the delay with high data offloading capacity.

A Semantics-Based Trajectory Segmentation Simplification Method

With the development of mobile positioning technology, a large amount of mobile trajectory data has been generated. Therefore, to store, process and mine trajectory data in a better way, trajectory data simplification is imperative. Current trajectory data simplification methods are either based on spatiotemporal features or semantic features, such as road network structure, but they do not consider semantic features of a trajectory stop. To overcome this limitation, this study presents a trajectory segmentation simplification method based on stop features. The proposed method first extracts the stop feature of a trajectory, then divides the trajectory into move segments and stop segments based on the stop features, and finally simplifies the obtained segments. The proposed method is verified by experiments on personal trajectory data and taxi trajectory data. Compared with the classic spatiotemporal simplification method, the proposed method has higher spatiotemporal and semantic accuracy under different simplification scales. The proposed method is especially suitable for trajectory data with more stop features.

A Temporal Directed Graph Convolution Network for Traffic Forecasting Using Taxi Trajectory Data

Traffic forecasting plays a vital role in intelligent transportation systems and is of great significance for traffic management. The main issue of traffic forecasting is how to model spatial and temporal dependence. Current state-of-the-art methods tend to apply deep learning models; these methods are unexplainable and ignore the a priori characteristics of traffic flow. To address these issues, a temporal directed graph convolution network (T-DGCN) is proposed. A directed graph is first constructed to model the movement characteristics of vehicles, and based on this, a directed graph convolution operator is used to capture spatial dependence. For temporal dependence, we couple a keyframe sequence and transformer to learn the tendencies and periodicities of traffic flow. Using a real-world dataset, we confirm the superior performance of the T-DGCN through comparative experiments. Moreover, a detailed discussion is presented to provide the path of reasoning from the data to the model design to the conclusions.

Comprehensive Analysis Based on Massive Taxi Trajectory Data

With the increasing demand for taxis, a large number of taxi trajectory data have been used for in-depth analysis. However, most early studies only used limited data or simulations to analyze the route choice behavior or traffic congestion problems. To extract more useful information from taxi trajectory data, this paper analyzes massive raw dates collected in Chengdu from 4 different perspectives. To begin with, passenger demand is analyzed by counting the number of orders in each period. K-means clustering is used for regional planning and provides appropriate boarding points for passengers. Furthermore, this paper focus on quantity duration of order duration, which presents an approximate Poisson distribution. Finally, the ratio of driving distance to the linear distance between the start and end points is defined as the overtravel ratio. Through the comprehensive analysis of the taxi trajectory data, it can provide assistance to the travel demand of passengers and the scheduling of taxi drivers, and maximize resource efficiency. Moreover, the data collected in Chengdu are very representative and large, so the corresponding suggestions can also provide hints for public transport problems.

Tempo‐spatial variability of urban leisure functional zones: An analysis based on geo‐big data

AbstractNew insights into urban functional zones with a specific focus on leisure is of great significance for human‐oriented urban development. The spatiotemporal pattern of urban leisure functional zones (ULFZ) reflects to what extent human psychological needs are satisfied. This article aims to discern ULFZ and reveal their changing regularity based on the taxi trajectory data and points of interests (POIs) by the DBSCAN algorithm, latent dirichlet allocations (LDA) and spatial analytical techniques. Results are concluded as follows: 1) the spatial distribution of ULFZs are dynamic and imbalanced at various periods; 2) among urban leisure subfunctions, cultural function is severely weaker than entertaining, sportive, and tourist function; and 3) the changes in ULFZ are kind of consistent at multiple spatial scales. This research benefits the leisure‐oriented urban planning.

XStar: a software system for handling taxi trajectory big data

Advances in positioning and communicating technologies make it possible to collect large volumes of taxi trajectory data, quickly providing a complete picture of the ground traffic systems and thus being applied to different fields. However, there are still challenges for data users to handle such big data. In view of this, we have developed a software system named XStar to deal with trajectory big data. Its core is a scalable index and storage structure. Based on it, raw data can be saved in a more compact scheme and accessed more efficiently. A real taxi trajectory dataset is employed to demonstrate its performance. In general, XStar facilitates processing and analyzing trajectory data affordably and straightforwardly. Since its release in Jan. 2019, it has received downloads of over 4000 by May 2021. More analytical functions are being developed.

An Online Map Matching Algorithm Based on Second-Order Hidden Markov Model

Map matching is a key preprocess of trajectory data which recently have become a major data source for various transport applications and location-based services. In this paper, an online map matching algorithm based on the second-order hidden Markov model (HMM) is proposed for processing trajectory data in complex urban road networks such as parallel road segments and various road intersections. Several factors such as driver’s travel preference, network topology, road level, and vehicle heading are well considered. An extended Viterbi algorithm and a self-adaptive sliding window mechanism are adopted to solve the map matching problem efficiently. To demonstrate the effectiveness of the proposed algorithm, a case study is carried out using a massive taxi trajectory dataset in Nanjing, China. Case study results show that the accuracy of the proposed algorithm outperforms the baseline algorithm built on the first-order HMM in various testing experiments.

Urban Hotspot Area Detection Using Nearest-Neighborhood-Related Quality Clustering on Taxi Trajectory Data

Urban hotspot area detection is an important issue that needs to be explored for urban planning and traffic management. It is of great significance to mine hotspots from taxi trajectory data, which reflect residents’ travel characteristics and the operational status of urban traffic. The existing clustering methods mainly concentrate on the number of objects contained in an area within a specified size, neglecting the impact of the local density and the tightness between objects. Hence, a novel algorithm is proposed for detecting urban hotspots from taxi trajectory data based on nearest neighborhood-related quality clustering techniques. The proposed spatial clustering algorithm not only considers the maximum clustering in a limited range but also considers the relationship between each cluster center and its nearest neighborhood, effectively addressing the clustering issue of unevenly distributed datasets. As a result, the proposed algorithm obtains high-quality clustering results. The visual representation and simulated experimental results on a real-life cab trajectory dataset show that the proposed algorithm is suitable for inferring urban hotspot areas, and that it obtains better accuracy than traditional density-based methods.

A Two-Phase Anomaly Detection Model for Secure Intelligent Transportation Ride-Hailing Trajectories

This paper addresses the taxi fraud problem and introduces a new solution to identify trajectory outliers. The approach as presented allows to identify both individual and group outliers and is based on a two phase-based algorithm. The first phase determines the individual trajectory outliers by computing the distance of each point in each trajectory, whereas the second identifies the group trajectory outliers by exploring the individual trajectory outliers using both feature selection and sliding windows strategies. A parallel version of the algorithm is also proposed using a sliding window-based GPU approach to boost the runtime performance. Extensive experiments have been carried out to thoroughly demonstrate the usefulness of our methodology on both synthetic and real trajectory databases. The results show that the GPU approach enables reaching a speed-up of 341 over the sequential algorithm on large synthetic databases. The efficiency of the proposed method to detect both individual and group trajectory outliers on a real-world taxi trajectory database is also demonstrated in comparison with baseline trajectory outlier and group detection algorithms. The results are very promising and show superiority of the proposed method both in reducing computational time and enhancing the quality of returned outliers. Finally, we prime our methodology and results for future refinement using deep learning methodologies.

Spatial and Temporal Characteristics of Urban Tourism Travel by Taxi—A Case Study of Shenzhen

Tourism networks are an important research part of tourism geography. Despite the significance of transportation in shaping tourism networks, current studies have mainly focused on the “daily behavior” of urban travel at the expense of tourism travel, which has been regarded as an “exceptional behavior”. To fill this gap, this study proposes a framework for exploring the spatial and temporal characteristics of urban tourism travel by taxi. We chose Shenzhen, a densely populated mega-city in China with abundant tourism resources, as a case study. First, we extracted tourist trips from taxi trajectories and used kernel density estimation to analyze the spatial aggregation characteristics of tourist trip origins. Second, we investigated the spatial dependence of tourist trips using local spatial autocorrelation analysis (Getis-Ord Gi*). Third, we explored the correlations between the tourist trip origins and urban geographic contextual factors (e.g., catering services and transportation facilities) using a geographically weighted regression model. The results show the following: (1) the trends between the coverage of tourist travel networks and the volume of tourist trips are similar; (2) the spatial interaction intensity of urban tourism has grouping and hierarchical characteristics; and (3) the spatial distribution of tourist trips by taxi is uneven and influenced by the distribution of urban morphology, tourism resources, and the preferences of taxi pick-up passengers. Our proposed framework and revealed spatial and temporal patterns have implications for urban tourism traffic planning, tourism product development, and tourist flow control in tourist attractions.

Optimizing matching time intervals for ride-hailing services using reinforcement learning

Matching trip requests and available drivers efficiently is considered a central operational problem for ride-hailing services. A widely adopted matching strategy is to accumulate a batch of potential passenger-driver matches and solve bipartite matching problems repeatedly. The efficiency of matching can be improved substantially if the matching is delayed by adaptively adjusting the matching time interval. The optimal delayed matching is subject to the trade-off between the delay penalty and the reduced wait cost and is dependent on the system’s supply and demand states. Searching for the optimal delayed matching policy is challenging, as the current policy is compounded with past actions. To this end, we tailor a family of reinforcement learning-based methods to overcome the curse of dimensionality and sparse reward issues. In addition, this work provides a solution to spatial partitioning balance between the state representation error and the optimality gap of asynchronous matching. Lastly, we examine the proposed methods with real-world taxi trajectory data and garner managerial insights into the general delayed matching policies. The focus of this work is single-ride service due to limited access to shared ride data, while the general framework can be extended to the setting with a ride-pooling component.

Trip destination prediction based on a deep integration network by fusing multiple features from taxi trajectories

Trip destination prediction plays an important role in exploring urban travel patterns. Accurate prediction can improve the efficiency of traffic management and the quality of location-based services. Here, a deep learning structure that contains three components: travel information extraction, classification learning mechanism, and output module is proposed. Three types of information (the partial trajectory of on-going trips, historical trajectories, and related external information) are extracted in the first component. Then, the classification learning mechanism chooses different methods (i.e. Long Short-Term Memory network and Embedding technology) according to the characteristic of variables. Finally, an output layer that integrates the prior information about destinations is constructed. Two open-source trajectory datasets are used to validate the effectiveness of the proposed model. Results show that the proposed model outperforms benchmark models using only part of the information or using all of the information but ignore the classification learning mechanism. The performance of the proposed model under different call types and travel durations is further explored. The result of this study will help understand travel behaviour in urban cities.

Discovering spatiotemporal characteristics of passenger travel with mobile trajectory big data

Mobile trajectory big data, such as global positioning system (GPS) trajectories of taxicabs, have enormous social and economic values in intelligent transportation systems, and efficient mining and in-depth analysis of these data can provide beneficial decision-making for traffic operation and especially for passenger travel. To this end, in this paper, we discover the temporal, spatial, and spatiotemporal characteristics of passenger travel by analyzing the travel duration and the travel distance with large-scale taxi trajectory data, which can not only effectively reveal the resident travel patterns but also dynamically perceive the traffic conditions. Specifically, we propose an information framework to preprocess the taxi GPS trajectory data that can thoroughly identify passenger travel trajectories. Moreover, we develop a novel distribution model to explore the rules of citizen travel accurately. Finally, we put forward a parallel clustering approach on Spark, which can discover the spatiotemporal characteristics of passenger travel in a fine-grained manner, to obtain the characteristic changes of inhabitant travel at different periods of the day. In particular, the experimental results from an empirical study show that the travel duration and the travel distance follow a log-normal distribution, and the tail of the travel distance is highly fitting to a three-parameter gamma distribution and the long-distance travel of the day is mainly to the airport between 5:00 and 6:00 AM.

Filtering Link Outliers in Vehicle Trajectories by Spatial Reasoning

Vehicle trajectories derived from Global Navigation Satellite Systems (GNSS) are used in various traffic applications based on trajectory quality analysis for the development of successful traffic models. A trajectory consists of points and links that are connected, where both the points and links are subject to positioning errors in the GNSS. Existing trajectory filters focus on point outliers, but neglect link outliers on tracks caused by a long sampling interval. In this study, four categories of link outliers are defined, i.e., radial, drift, clustered, and shortcut; current available algorithms are applied to filter apparent point outliers for the first three categories, and a novel filtering approach is proposed for link outliers of the fourth category in urban areas using spatial reasoning rules without ancillary data. The proposed approach first measures specific geometric properties of links from trajectory databases and then evaluates the similarities of geometric measures among the links, following a set of spatial reasoning rules to determine link outliers. We tested this approach using taxi trajectory datasets for Beijing with a built-in sampling interval of 50 to 65 s. The results show that clustered links (27.14%) account for the majority of link outliers, followed by shortcut (6.53%), radial (3.91%), and drift (0.62%) outliers.

Understanding the urban mobility community by taxi travel trajectory

With the increase of urban population and the expansion of urban scale, understanding the urban structure could provide intellectual support for urban planning, traffic congestion, and even the spread of diseases. Little research has addressed the relationship between urban structure and human mobility. In this study, the community division method is applied to the itinerary network generated by taxi trajectory data, and the itinerary network is divided into several modules (called urban mobility communities). Our results suggest that the urban mobility community has different associated characteristics with travel distance and administrative area. When the travel distance increases, many adjacent communities gradually merge into large ones in the central area, while the communities remain similar in the suburban area. Moreover, we observed some communities around some pivotal roads are distributed as ribbons. The quantitative analysis of community division results demonstrates a phase transition between urban mobility community and travel distance. In particular, inspired by heterogeneous travel distance distribution and community boundary restrictions, we first construct a model that reflects the non-trivial relationship between urban mobility community and travel distance and reveals the phase transition between them. Our study represents the first attempt to prove that human mobility and urban structure can be reflected by each other.

Beyond absolute space: An exploration of relative and relational space in Shanghai using taxi trajectory data

The nature of urban space has long-drawn geographers' interest and David Harvey's conceptual framework of multiple spaces (i.e., absolute, relative, and relational) within cities has been widely adopted and developed. With its high spatial and temporal resolution, geospatial big data plays an increasingly important role in our understanding of urban structure. Taxi trajectory data is particularly useful in travel purpose estimation and allows for more granular insights into urban mobility due to the door-to-door nature of these trips. This article utilizes taxi trajectory data and explores the interaction among absolute space, relative space, and relational space in Harvey's framework using Structural Equation Modeling (SEM). Through an empirical study of Shanghai's downtown area, this paper highlights the importance of Harvey's framework in understanding cities' dynamic structure and argues for changes in urban planning and development to better coordinate land use and travel demand. We find an insignificant relationship between relative and relational space in Shanghai due to a mismatch between urban mobility and the built environment. This mismatch concentrates the transportation flow near the city's core area, transforming the polycentric structure of Shanghai's built environment in absolute space to a single-node structure in relational space. After identifying the contributing factors to this problem in Shanghai, this article suggests combining Harvey's conceptual framework of multiple spaces with geospatial big data to inform planning strategies that address the challenges of rapid urbanization.

Discovery of arbitrarily shaped significant clusters in spatial point data with noise

Spatial point data is an important data source that represents the locations of spatial events (such as crime, disease cases, and earthquakes). Detecting clusters I n spatial point data plays a key role in exploratory spatial data analysis. Although much attention has been paid to the clustering of spatial points, how to automatically and efficiently discover the statistically significant clusters with irregular shapes is still a challenging work. On that account, an automatic method to detect the statistically significant high-density clusters in spatial point data with noise is proposed in this paper. First, the Voronoi diagram of the spatial points is constructed, and the densities of spatial points are defined by the areas of the Voronoi cells. Then, high-density points are automatically detected using spatial hotspot statistics analysis, and a density-based clustering strategy is further adapted to combine the neighboring high-density points into candidate clusters. Finally, a statistical significance test is proposed to evaluate the significance of the candidate clusters under the spatially homogeneous or heterogeneous distribution assumption. We tested the proposed method with the simulated data sets and the real-world taxi trajectory data for detecting the pick-up hotspot regions in Wuhan, China. Results show that the proposed method can successfully find arbitrarily shaped significant clusters that existing state-of-the-art clustering algorithms may fail to find in spatial point data with noise.

An offline map matching algorithm based on shortest paths

ABSTRACT Offline map matching identifies corresponding roads to a GPS trajectory represented by a series of recorded geographic coordinates (GPS points) to the road network. This paper defines matching error as cost on the corresponding road-link to matched GPS points and formulates the offline map matching problem as a shortest path problem with resource constraints. By regarding matched points on one link as a type of resource consumed, the resource constraint indicates that the number of matched GPS points equals the total number of points in the given trajectory. We propose an offline map matching algorithm based on shortest paths by calculating the matching error on each link and extending the classic label-setting shortest path algorithm to find the path with the minimum total matching error for all GPS points. We use real-world taxi trajectories to compare our algorithm with three state-of-the-art map matching algorithms. Our algorithm outperforms all benchmark algorithms in terms of both matching accuracy and computational efficiency. Our algorithm achieves greater matched length (5.36 to 12.27% larger) and lower mis-matched length (3.72 to 75.30% smaller) at a very high matching speed (60.59 points per second on average over thirteen sampling intervals).

Inferring Urban Land Use from Multi-Source Urban Mobility Data Using Latent Multi-View Subspace Clustering

In the era of big data, vast urban mobility data introduce new opportunities to infer urban land use from the perspective of social function. Most existing works only derive land use information from a single type of urban mobility dataset, which is typically biased and results in difficulty obtaining a comprehensive view of urban land use. It remains challenging to fuse high-dimensional and noisy multi-source urban mobility data to infer urban land use. This study aimed to infer urban land use from multi-source urban mobility data using latent multi-view subspace clustering. The variation in the number of origin/destination points over time was initially used to characterize land use types. Then, a latent multi-view representation was applied to construct the common underlying structure shared by multi-source urban mobility data and effectively deal with noise. Finally, based on the latent multi-view representation, the subspace clustering method was used to infer the land use types. Experiments on taxi trajectory data and bus smart card data in Beijing reveal that, compared with the method using a single type of urban mobility dataset and the weighted fusion method, the approach presented in this study obtains the highest detection rate of land use. The urban land use inferred in this study provides calibration and reference for urban planning.