Learning driver-specific behavior for overtaking : a combined learning framework

Learning-based methods have gained increasing attention in the intelligent vehicle community for developing highly autonomous vehicles and advanced driving assistance systems (ADAS). However, traditional offline learning methods lack the ability to adapt to individual driving behavior. To overcome t...

Full description

Saved in:
Bibliographic Details
Main Authors: Lu, Chao, Wang, Huaji, Lv, Chen, Gong, Jianwei, Xi, Junqiang, Cao, Dongpu
Other Authors: School of Electrical and Electronic Engineering
Format: Article
Language:English
Published: 2020
Subjects:
Online Access:https://hdl.handle.net/10356/142721
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-142721
record_format dspace
spelling sg-ntu-dr.10356-1427212020-06-29T03:38:04Z Learning driver-specific behavior for overtaking : a combined learning framework Lu, Chao Wang, Huaji Lv, Chen Gong, Jianwei Xi, Junqiang Cao, Dongpu School of Electrical and Electronic Engineering School of Mechanical and Aerospace Engineering Engineering::Electrical and electronic engineering Driving Behaviour Natural Actor Critic Learning-based methods have gained increasing attention in the intelligent vehicle community for developing highly autonomous vehicles and advanced driving assistance systems (ADAS). However, traditional offline learning methods lack the ability to adapt to individual driving behavior. To overcome this limitation, a combined learning framework (CLF) based on the Natural Actor Critic (NAC) learning and general regression neural network (GRNN) is developed in this paper. GRNN can be trained offline based on the historical data, while NAC is carried out online. In this way, the general behavior learned by the offline module can be reused and adjusted by the online module to capture the driver-specific behavior. Driving data collected from human drivers through a driving simulator are used to test the proposed learning framework. The complex overtaking behavior is selected to formulate the learning problem and test scenarios. Experimental results show that the proposed system performs well on learning driver-specific behavior for overtaking, and compared with the Gaussian mixture model-maximum-a-posterior method, CLF shows a more flexible performance when newly-involved drivers are considered. 2020-06-29T03:38:04Z 2020-06-29T03:38:04Z 2018 Journal Article Lu, C., Wang, H., Lv, C., Gong, J., Xi, J., & Cao, D. (2018). Learning driver-specific behavior for overtaking : a combined learning framework. IEEE Transactions on Vehicular Technology, 67(8), 6788 - 6802. doi:10.1109/TVT.2018.2820002 0018-9545 https://hdl.handle.net/10356/142721 10.1109/TVT.2018.2820002 2-s2.0-85044846242 8 67 6788 6802 en IEEE Transactions on Vehicular Technology © 2018 IEEE. All rights reserved.
institution Nanyang Technological University
building NTU Library
country Singapore
collection DR-NTU
language English
topic Engineering::Electrical and electronic engineering
Driving Behaviour
Natural Actor Critic
spellingShingle Engineering::Electrical and electronic engineering
Driving Behaviour
Natural Actor Critic
Lu, Chao
Wang, Huaji
Lv, Chen
Gong, Jianwei
Xi, Junqiang
Cao, Dongpu
Learning driver-specific behavior for overtaking : a combined learning framework
description Learning-based methods have gained increasing attention in the intelligent vehicle community for developing highly autonomous vehicles and advanced driving assistance systems (ADAS). However, traditional offline learning methods lack the ability to adapt to individual driving behavior. To overcome this limitation, a combined learning framework (CLF) based on the Natural Actor Critic (NAC) learning and general regression neural network (GRNN) is developed in this paper. GRNN can be trained offline based on the historical data, while NAC is carried out online. In this way, the general behavior learned by the offline module can be reused and adjusted by the online module to capture the driver-specific behavior. Driving data collected from human drivers through a driving simulator are used to test the proposed learning framework. The complex overtaking behavior is selected to formulate the learning problem and test scenarios. Experimental results show that the proposed system performs well on learning driver-specific behavior for overtaking, and compared with the Gaussian mixture model-maximum-a-posterior method, CLF shows a more flexible performance when newly-involved drivers are considered.
author2 School of Electrical and Electronic Engineering
author_facet School of Electrical and Electronic Engineering
Lu, Chao
Wang, Huaji
Lv, Chen
Gong, Jianwei
Xi, Junqiang
Cao, Dongpu
format Article
author Lu, Chao
Wang, Huaji
Lv, Chen
Gong, Jianwei
Xi, Junqiang
Cao, Dongpu
author_sort Lu, Chao
title Learning driver-specific behavior for overtaking : a combined learning framework
title_short Learning driver-specific behavior for overtaking : a combined learning framework
title_full Learning driver-specific behavior for overtaking : a combined learning framework
title_fullStr Learning driver-specific behavior for overtaking : a combined learning framework
title_full_unstemmed Learning driver-specific behavior for overtaking : a combined learning framework
title_sort learning driver-specific behavior for overtaking : a combined learning framework
publishDate 2020
url https://hdl.handle.net/10356/142721
_version_ 1681059317562736640