Day-to-day dynamic origin–destination flow estimation using connected vehicle trajectories and automatic vehicle identification data

Cao, Yumin; Tang, Keshuang; Sun, Jian; Ji, Yangbeibei

doi:10.1016/j.trc.2021.103241

Cited by 30 publications

(18 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For Xuhui data, all models perform worst in 'spatial' FM-1 case. This is because without data of their neighbor traffic analysis zones, imputing OD flows has to exploit the time dimension, where shows slight within-day consistency (Cao et al, 2021), thus the information loss along this axis has a significant impact. One possible response is to establish the space topology as a additional prior knowledge, e.g., Graph Laplacian regularization in (Yang et al, 2021a).…”

Section: Imputation Resultsmentioning

confidence: 99%

Truncated tensor Schatten p-norm based approach for spatiotemporal traffic data imputation with complicated missing patterns

Nie,

Qin,

Sun

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Rapid advances in sensor, wireless communication, cloud computing and data science have brought unprecedented amount of data to assist transportation engineers and researchers in making better decisions. However, traffic data in reality often has corrupted or incomplete values due to detector and communication malfunctions. Data imputation is thus required to ensure the effectiveness of downstream data-driven applications. To this end, numerous tensor-based methods treating the imputation problem as the low-rank tensor completion (LRTC) have been attempted in previous works. To tackle rank minimization, which is at the core of the LRTC, most of aforementioned methods utilize the tensor nuclear norm (NN) as a convex surrogate for the minimization. However, the over-relaxation issue in NN refrains it from desirable performance in practice. In this paper, we define an innovative nonconvex truncated Schatten p-norm for tensors (TSpN) to approximate tensor rank and impute missing spatiotemporal traffic data under the LRTC framework. We model traffic data into a thirdorder tensor structure of time intervals×locations (sensors)×days and introduce four complicated missing patterns, including random missing and three fiber-like missing cases according to the tensor mode-n fibers. Despite nonconvexity of the objective function in our model, we derive the global optimal solutions by integrating the alternating direction method of multipliers (ADMM) with generalized soft-thresholding (GST). In addition, we design a truncation rate decay strategy to deal with varying missing rate scenarios. Comprehensive experiments are finally conducted using four different types of real-world spatiotemporal traffic datasets, which demonstrate that the proposed LRTC-TSpN method performs well under various missing cases even with high rates of data loss, meanwhile outperforming other state-of-the-art tensor-based imputation models in almost all scenarios. In general, the performance of the proposed method achieves up to a 52% improvement in mean absolute error (MAE) compared to baseline methods even when the missing ratio is as high as 90%. Moreover, both theoretical and empirical evidences indicate that our method has robust and efficient convergence performances.

show abstract

Section: Imputation Resultsmentioning

confidence: 99%

Truncated tensor Schatten p-norm based approach for spatiotemporal traffic data imputation with complicated missing patterns

Nie,

Qin,

Sun

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…These locations are then incorporated into a Markov model that can be consulted for various applications in both single-user and collaborative scenarios. Cao et al proposed factorizing personalized Markov chain (FPMC), which combines matrix factorization and first-order Markov chains to capture long-term preferences and forecast the subsequent purchase behavior ( 11 ). Incorporating the softmax function to fuse the personalized Markov chain with the latent pattern, He et al furnish a Bayesian personalized ranking (BPR) approach and gives the ideal prediction location accordingly ( 12 ).…”

Section: Literature Reviewmentioning

confidence: 99%

“…In recent years, automatic vehicle identification (AVI) systems have been rapidly deployed in many countries, bringing new tasks for individual behavior prediction problems ( 11 ). The main components of the novel vehicle identification technology are AVI sensors, whose operation can be illustrated in Figure 1.…”

mentioning

confidence: 99%

When and Where to Go Next: Deep Learning Framework for Modeling Drivers’ Behaviors Using Automatic Vehicle Identification Data

Jin

Wang

et al. 2022

Transportation Research Record: Journal of the Transportation R

View full text Add to dashboard Cite

In recent years, automatic vehicle identification (AVI) systems have rapidly developed in many countries, which provides an excellent opportunity to understand drivers’ mobility patterns in urban road networks. Few efforts have been devoted to fully utilizing the AVI data to address the prediction issue in behaviors modeling. This paper proposes a deep learning (DL) framework driven by AVI data to model drivers’ behaviors and further incorporate travel time prediction in the next location prediction problem. Specifically, DeepWalk encoder and DeepWalk + Time long short-term memory (DT-LSTM) were proposed to capture the spatial and temporal correlations simultaneously. By learning the spatial relationship between the sensors from the historical trajectories, the DeepWalk encoder converted the sensors into low-dimensional numerical vectors essential for DL. As a vector form, the spatial relationship of the sensor can be adequately measured. Besides, a new long short-term memory (LSTM) variant, DT-LSTM, was designed to memorize drivers’ short-term and long-term interests, respectively. By adding additional time gates, DT-LSTM facilities the location and travel time consistency in the trajectories sequences with different time intervals. Experimental results demonstrated the effectiveness of the model, which achieved the state-of-the-art performance (SOTA) compared with other benchmark methods, with the Acc-1 of 82%, Acc-5 of 95%, and MAPE of 17%.

show abstract

“…However, path information cannot be fully detected, so researchers normally make use of observed flows or data on partial path, which can be captured by GPS, mobile phone, plate scanning, AVI, etc. [34][35][36][37][38][39]. For instance, to estimate static O-D demand, Hu et al [35] proposed link-based and pathbased models to estimate O-D demand based on traffic counts by vehicle detector sensors and license plate recognition.…”

Section: O-d Demand Estimation Using Multiple Sources Of Datamentioning

confidence: 99%

“…Krishnakumari et al [38] proposed a method without dynamic network loading using measured flows and speeds. Cao et al [39] estimated day-to-day dynamic O-D demand based on connected vehicle trajectories and automatic vehicle identification data.…”

Section: O-d Demand Estimation Using Multiple Sources Of Datamentioning

confidence: 99%

A Bayesian Method for Dynamic Origin–Destination Demand Estimation Synthesizing Multiple Sources of Data

Zhu

Yang

et al. 2021

Sensors

View full text Add to dashboard Cite

In this paper a Bayesian method is proposed to estimate dynamic origin–destination (O–D) demand. The proposed method can synthesize multiple sources of data collected by various sensors, including link counts, turning movements at intersections, flows, and travel times on partial paths. Time-dependent demand for each O–D pair at each departure time is assumed to satisfy the normal distribution. The connections among multiple sources of field data and O–D demands for all departure times are established by their variance-covariance matrices. Given the prior distribution of dynamic O–D demands, the posterior distribution is developed by updating the traffic count information. Then, based on the posterior distribution, both point estimation and the corresponding confidence intervals of O–D demand variables are estimated. Further, a stepwise algorithm that can avoid matrix inversion, in which traffic counts are updated one by one, is proposed. Finally, a numerical example is conducted on Nguyen–Dupuis network to demonstrate the effectiveness of the proposed Bayesian method and solution algorithm. Results show that the total O–D variance is decreasing with each added traffic count, implying that updating traffic counts reduces O–D demand uncertainty. Using the proposed method, both total error and source-specific errors between estimated and observed traffic counts decrease by iteration. Specifically, using 52 multiple sources of traffic counts, the relative errors of almost 50% traffic counts are less than 5%, the relative errors of 85% traffic counts are less than 10%, the total error between the estimated and “true” O–D demands is relatively small, and the O–D demand estimation accuracy can be improved by using more traffic counts. It concludes that the proposed Bayesian method can effectively synthesize multiple sources of data and estimate dynamic O–D demands with fine accuracy.

show abstract

Day-to-day dynamic origin–destination flow estimation using connected vehicle trajectories and automatic vehicle identification data

Cited by 30 publications

References 51 publications

Truncated tensor Schatten p-norm based approach for spatiotemporal traffic data imputation with complicated missing patterns

Truncated tensor Schatten p-norm based approach for spatiotemporal traffic data imputation with complicated missing patterns

When and Where to Go Next: Deep Learning Framework for Modeling Drivers’ Behaviors Using Automatic Vehicle Identification Data

A Bayesian Method for Dynamic Origin–Destination Demand Estimation Synthesizing Multiple Sources of Data

Contact Info

Product

Resources

About