Origin-destination Flow Prediction with Vehicle Trajectory Data and Semi-supervised Recurrent Neural Network

Huang, Tao; Ma, Yintai; Qin, Zhiwei; Zheng, Jianfeng; Liu, Henry; Zhu, Hongtu; Ye, Jieping

doi:10.1109/bigdata47090.2019.9005972

Cited by 10 publications

(4 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our ANN consists of three layers with units [256, 256,

N

], where

N

is the number of stations. We fed all the extracted features to it, choosing ReLU as the activation function of all layers. XGBoost 49 : XGBoost represents a prediction method using ensemble learning method for regression problems, generating a prediction model in the form of an ensemble of basic learning models, typically decision trees. We fed all the extracted features from this article into the model. CASCNN 30 (channel‐wise attentive split convolutional neural network): CASCNN is a CNN‐based deep learning model designed for predicting short‐term OD matrices in a metro system.…”

Section: Experiments Results and Analysismentioning

confidence: 99%

STRmt: A state transition based model for real‐time crowd counting in a metro system

Sun,

Zhao,

Zhang

et al. 2024

Concurrency and Computation

View full text Add to dashboard Cite

SummaryReal‐time estimation of crowd counting in underground metro systems, constrained by limited space, is crucial for managing heightened pedestrian volumes and responding promptly to emergencies. To address this challenge, we propose a passenger state transition‐based model, called STRmt, designed for the seamless and continuous monitoring of real‐time crowd movement within service areas of stations and trains, leveraging auto fare collection systems (AFC) as a comprehensive sensor network. Our innovation lies in modeling the dynamic movement of passengers within a metro system over time as a state transition process aligned with the train schedule. To achieve this, we introduce a spatio‐temporal deep learning framework, denoted as STnet, designed to dynamically predict these state transitions. The performance of our method is rigorously assessed through extensive experiments conducted spanning 2 years in Shenzhen, China, utilizing AFC data, train schedule data, and weather data. The results demonstrate that the proposed method surpasses baseline methods, achieving an estimation precision of 0.92.

show abstract

“…Our ANN consists of three layers with units [256, 256,

N

], where

N

Section: Experiments Results and Analysismentioning

confidence: 99%

STRmt: A state transition based model for real‐time crowd counting in a metro system

Sun,

Zhao,

Zhang

et al. 2024

Concurrency and Computation

View full text Add to dashboard Cite

show abstract

“…It suggests that a higher number of ofces in origin TAZs are associated with lower travel fow diferences between peak hours. However, in previous studies, ofces are usually thought as major sources of travel demand [39,43,59]. However, since this paper focuses on the travel fow diferences, ofces tend to generate large but relatively equal commuting travel demand as people need to go to work in the morning peak and leave ofce in evening peak.…”

Section: Efects Of Explanatorymentioning

confidence: 93%

“…Te results revealed six variations of the travel demand on weekdays and weekends. Huang et al [39] proposed a semisupervised deep learning based model that appropriately combines both AVI and smartphone trajectory data during training. Te model can provide OD estimation and prediction services on larger spatial areas beyond the limited spatial coverage of AVI data.…”

Section: Avi Data In Travel Flowmentioning

confidence: 99%

Investigating Travel Flow Differences between Peak Hours with Spatial Model with Endogenous Weight Matrix Using Automatic Vehicle Identification Data

Zhou

Chen

et al. 2022

Journal of Advanced Transportation

View full text Add to dashboard Cite

The rapid urbanization has brought great challenges to the transportation network. However, travel flow at peak hours is not always the same. It is important to investigate how travel flow differs between peak hours to capture travel flow patterns and influential factors to facilitate traffic management and urban planning. This paper establishes a spatial model with endogenous weight matrix (SARBP-EWM) to investigate the travel flow differences between morning and evening peaks on both weekday and weekend based on automatic vehicle identification (AVI) data and point of interest (POI) data in Xuancheng, China. The results confirm strong spatial effects and endogeneity issue. Besides, facility variables such as number of offices and number of clinics reveal strong negative impacts on travel flow differences on both weekday and weekend, while the number of middle school shows significantly positive relation with travel flow differences. In addition, the endogenous weight matrix on both weekday and weekend is successfully estimated and compared. It is found that TAZ pairs tend to be clustered with lower spatial weights on weekday, while they are more randomly distributed with higher spatial weights at weekend. Based on the results above, the policies proposed from Xuancheng 14th Five-Year Plan are evaluated and discussed. The above empirical analysis quantifies impacts from key factors on urban travel flow differences between peak hours and provides important references for urban planning and policy making.

show abstract

“…The second is intelligence-based control, which is mainly based on deep learning, reinforcement learning, and other AI-related techniques, and mainly uses the function approximation capabilities of deep neural network models to solve complex decision processes. Intelligent-control-related models have been applied to traffic flow prediction [99][100][101][102]. Hou et al [103] adopted the fusion of random forest and AdaBoost to make lane change decisions via the first lane change model established through NGSIM.…”

Section: Control Modelsmentioning

confidence: 99%

Vehicle Lane Change Models—A Historical Review

Liu,

Hong,

Lin

2023

Applied Sciences

View full text Add to dashboard Cite

Lane changing is a complex operation that has a significant impact on traffic safety. The accurate identification and assessment of potential risks in the driving environment before lane changing is crucial for the safe and smooth completion of a lane change. In this paper, the research status of vehicle lane change models is reviewed. Firstly, various factors affecting lane change models are analyzed. Different drivers will be affected by vehicle dynamic parameters, vehicle driving states, and driver characteristics under various road environments. Secondly, the vehicle lane change models are divided into four types: the empirical model of lane changing, the physical model of lane changing, the cognitive model of lane changing, and the mixed model of lane changing. The advantages and disadvantages of different types of lane change models are analyzed, and the key problems to be solved by different lane change models are expounded, respectively, from the aspects of input variables and reasoning algorithms. Finally, according to the advantages and disadvantages of different lane change models, a future research direction is proposed.

show abstract

Origin-destination Flow Prediction with Vehicle Trajectory Data and Semi-supervised Recurrent Neural Network

Cited by 10 publications

References 33 publications

STRmt: A state transition based model for real‐time crowd counting in a metro system

STRmt: A state transition based model for real‐time crowd counting in a metro system

Investigating Travel Flow Differences between Peak Hours with Spatial Model with Endogenous Weight Matrix Using Automatic Vehicle Identification Data

Vehicle Lane Change Models—A Historical Review

Contact Info

Product

Resources

About