中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 地理科学与资源研究所
  3. 中国科学院地理科学与资源研究所
HDRLM3D: A Deep Reinforcement Learning-Based Model with Human-like Perceptron and Policy for Crowd Evacuation in 3D Environments

文献类型：期刊论文

作者Zhang, Dong1,2; Li, Wenhang1; Gong, Jianhua1,2,3; Huang, Lin1; Zhang, Guoyong1; Shen, Shen4; Liu, Jiantao5; Ma, Haonan1,2
刊名ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
出版日期2022-04-01
卷号11期号:4页码:20
关键词crowd simulation agent-based model deep reinforcement learning perceptron policy
DOI10.3390/ijgi11040255
通讯作者Li, Wenhang(liwh@aircas.ac.cn)
英文摘要At present, a common drawback of crowd simulation models is that they are mainly simulated in (abstract) 2D environments, which limits the simulation of crowd behaviors observed in real 3D environments. Therefore, we propose a deep reinforcement learning-based model with human-like perceptron and policy for crowd evacuation in 3D environments (HDRLM3D). In HDRLM3D, we propose a vision-like ray perceptron (VLRP) and combine it with a redesigned global (or local) perceptron (GOLP) to form a human-like perception model. We propose a double-branch feature extraction and decision network (DBFED-Net) as the policy, which can extract features and make behavioral decisions. Moreover, we validate our method's ability to reproduce typical phenomena and behaviors through experiments in two different scenarios. In scenario I, we reproduce the bottleneck effect of crowds and verify the effectiveness and advantages of HDRLM3D by comparing it with real crowd experiments and classical methods in terms of density maps, fundamental diagrams, and evacuation times. In scenario II, we reproduce agents' navigation and obstacle avoidance behaviors and demonstrate the advantages of HDRLM3D for crowd simulation in unknown 3D environments by comparing it with other deep reinforcement learning-based models in terms of trajectories and numbers of collisions.
WOS关键词BEHAVIOR ; DRIVEN
资助项目National Natural Science Foundation of China[42171113] ; National Natural Science Foundation of China[41971361] ; National Key Technology R&D Program of China[2020YFC0833103] ; Pilot Fund of Frontier Science and Disruptive Technology of Aerospace Information Research Institute, Chinese Academy of Sciences[E0Z21101]
WOS研究方向Computer Science ; Physical Geography ; Remote Sensing
语种英语
WOS记录号WOS:000787951100001
出版者MDPI
资助机构National Natural Science Foundation of China ; National Key Technology R&D Program of China ; Pilot Fund of Frontier Science and Disruptive Technology of Aerospace Information Research Institute, Chinese Academy of Sciences
源URL[http://ir.igsnrr.ac.cn/handle/311030/175162]  
专题中国科学院地理科学与资源研究所
通讯作者Li, Wenhang
作者单位1.Chinese Acad Sci, Natl Engn Res Ctr Geoinformat, Aerosp Informat Res Inst, Beijing 100094, Peoples R China
2.Univ Chinese Acad Sci, Beijing 100049, Peoples R China
3.Zhejiang CAS Applicat Ctr Geoinformat, Jiaxing 314199, Peoples R China
4.Inst Geog Sci & Nat Resources Res, State Key Lab Resources & Environm Informat Syst, CAS, Beijing 100101, Peoples R China
5.Shandong Jianzhu Univ, Sch Surveying & Geoinformat, Jinan 250101, Peoples R China
推荐引用方式
GB/T 7714		 Zhang, Dong,Li, Wenhang,Gong, Jianhua,et al. HDRLM3D: A Deep Reinforcement Learning-Based Model with Human-like Perceptron and Policy for Crowd Evacuation in 3D Environments[J]. ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION,2022,11(4):20.
APA Zhang, Dong.,Li, Wenhang.,Gong, Jianhua.,Huang, Lin.,Zhang, Guoyong.,...&Ma, Haonan.(2022).HDRLM3D: A Deep Reinforcement Learning-Based Model with Human-like Perceptron and Policy for Crowd Evacuation in 3D Environments.ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION,11(4),20.
MLA Zhang, Dong,et al."HDRLM3D: A Deep Reinforcement Learning-Based Model with Human-like Perceptron and Policy for Crowd Evacuation in 3D Environments".ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION 11.4(2022):20.

入库方式： OAI收割

来源：地理科学与资源研究所

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。