Long memory is important: A test study on deep-learning based car-following model

doi:10.1016/j.physa.2018.09.136

CORC > 自动化研究所 > 中国科学院自动化研究所 > 复杂系统管理与控制国家重点实验室 > 先进控制与自动化团队

	Long memory is important: A test study on deep-learning based car-following model
	Wang, Xiao 1; Jiang, Rui 2; Li, Li 3; Lin, Yi-Lun4 ; Wang, Fei-Yue4
刊名	PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
	2019-01-15
卷号	514 页码:786-795
关键词	Car-following Long memory Hysteresis Deep learning
ISSN号	0378-4371
DOI	10.1016/j.physa.2018.09.136
通讯作者	Jiang, Rui(jiangrui@bjtu.edu.cn) ; Li, Li(li-li@tsinghua.edu.cn)
英文摘要	Whether long memory effect plays an important role in car-following models remains unsolved. In this paper, we study the possible relationship between long memory effect and hysteresis phenomena observed in practice. Especially, we have compared the performance of different deep learning based car-following models that take various time-scale historical information as inputs. Test show that hysteresis phenomena can be correctly simulated only by car-following models with long memory. So, we argue that car-following models should embed long memory effect appropriately. (C) 2018 Elsevier B.V. All rights reserved.
资助项目	National Natural Science Foundation of China[61790565] ; Beijing Municipal Science and Technology Commission Program, China[D171100000317002] ; Beijing Municipal Commission of Transport, China[ZC179074Z] ; Natural Science Foundation of Beijing Municipality, China[9172013]
WOS关键词	CONGESTED TRAFFIC STATES ; PHASE-TRANSITIONS ; NEURAL-NETWORKS ; FLOW ; CALIBRATION ; HYSTERESIS ; STABILITY ; ALGORITHM ; DIAGRAM ; DELAY
WOS研究方向	Physics
语种	英语
出版者	ELSEVIER SCIENCE BV
WOS记录号	WOS:000450137000070
资助机构	National Natural Science Foundation of China ; Beijing Municipal Science and Technology Commission Program, China ; Beijing Municipal Commission of Transport, China ; Natural Science Foundation of Beijing Municipality, China
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/22587]
专题	自动化研究所_复杂系统管理与控制国家重点实验室_先进控制与自动化团队自动化研究所_复杂系统管理与控制国家重点实验室
通讯作者	Jiang, Rui; Li, Li
作者单位	1.Xi An Jiao Tong Univ, Dept Comp Sci & Technol, Xian 710049, Shaanxi, Peoples R China 2.Tsinghua Univ, Dept Automat, Beijing 100084, Peoples R China 3.Beijing Jiaotong Univ, Sch Traff & Transportat, Beijing 100044, Peoples R China 4.Chinese Acad Sci, State Key Lab Management & Control Complex Syst, Inst Automat, Beijing 100080, Peoples R China
推荐引用方式 GB/T 7714	Wang, Xiao,Jiang, Rui,Li, Li,et al. Long memory is important: A test study on deep-learning based car-following model[J]. PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS,2019,514:786-795.
APA	Wang, Xiao,Jiang, Rui,Li, Li,Lin, Yi-Lun,&Wang, Fei-Yue.(2019).Long memory is important: A test study on deep-learning based car-following model.PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS,514,786-795.
MLA	Wang, Xiao,et al."Long memory is important: A test study on deep-learning based car-following model".PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS 514(2019):786-795.