End-to-End Autonomous Driving Through Dueling Double Deep Q-Network
作者机构:State Key Lab of Automotive Safety and EnergySchool of Vehicle and MobilityTsinghua UniversityBeijing 100084China Korea Advanced Institute of Science and Technology193Munji-roYuseong-guDaejeonKorea DiDi Autonomous Driving CompanyBeijing 100084China
出 版 物:《Automotive Innovation》 (汽车创新工程(英文))
年 卷 期:2021年第4卷第3期
页 面:328-337页
核心收录:
学科分类:08[工学] 082304[工学-载运工具运用工程] 080204[工学-车辆工程] 0802[工学-机械工程] 0823[工学-交通运输工程]
基 金:This work is supported by the National Key Research and Development Project of China under Grant 2018YFB1600600 Beijing Natural Science Foundation with JQ18010.The authors should also thank the support from Tsinghua University-Didi Joint Research Center for Future Mobility
主 题:End-to-end autonomous driving Reinforcement learning Deep Q-network Neural network
摘 要:Recent years have seen the rapid development of autonomous driving systems,which are typically designed in a hierarchical architecture or an end-to-end *** hierarchical architecture is always complicated and hard to design,while the end-to-end architecture is more promising due to its simple *** paper puts forward an end-to-end autonomous driving method through a deep reinforcement learning algorithm Dueling Double Deep Q-Network,making it possible for the vehicle to learn end-to-end driving by *** paper firstly proposes an architecture for the end-to-end lane-keeping *** the traditional image-only state space,the presented state space is composed of both camera images and vehicle motion *** corresponding dueling neural network structure is introduced,which reduces the variance and improves sampling ***,the proposed method is applied to The Open Racing Car Simulator(TORCS)to demonstrate its great performance,where it surpasses human ***,the saliency map of the neural network is visualized,which indicates the trained network drives by observing the lane lines.A video for the presented work is available online,https://***/76ciJ mIHMD8 or https://***/v_show/id_XNDM4 ODc0M TM4NA==.html.