A review of prediction methods for moving target trajectories

With the rapid emergence of mobile terminal equipment in intelligent transportation system, the deep understanding and accurate prediction of moving target trajectories are capable of reducing the traffic accident probability, and promoting the location service-based intelligent transportation appli...

Full description

Saved in:
Bibliographic Details
Main Authors: Wen LIU, Kunlin HU, Yan LI, Zhao LIU
Format: Article
Language:zho
Published: POSTS&TELECOM PRESS Co., LTD 2021-06-01
Series:智能科学与技术学报
Subjects:
Online Access:http://www.cjist.com.cn/thesisDetails#10.11959/j.issn.2096-6652.202115
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:With the rapid emergence of mobile terminal equipment in intelligent transportation system, the deep understanding and accurate prediction of moving target trajectories are capable of reducing the traffic accident probability, and promoting the location service-based intelligent transportation applications.The trajectory prediction methods prediction methods for moving target trajectories were reviewed from the data-driven prediction methods and the behavior-driven trajectories prediction methods.Firstly, the data-driven prediction methods were reviewed, including probabilistic statistics, neural networks, deep learning, and hybrid modeling.Then, the basic conceptions of target behavior-driven trajectories prediction methods were analyzed.The corresponding dynamical modeling and intention recognition methods were reviewed.The trajectory prediction applications were briefly analyzed and reviewed, such as target trajectory reconstruction, target abnormal behavior identification, and navigation route planning.Finally, the main problems and development directions related to prediction of moving target trajectories were discussed.
ISSN:2096-6652