中国安全科学学报 ›› 2019, Vol. 29 ›› Issue (4): 51-57.doi: 10.16265/j.cnki.issn1003-3033.2019.04.009

• 安全系统学 • 上一篇    下一篇

基于元胞自动机的应急疏散最短路径优化算法

何梦男1, 付瑜玲1,2, 陈诚**1 工程师, 何湖滨1, 林育青1, 陈求稳1 研究员   

  1. 1 南京水利科学研究院 生态环境研究中心,江苏 南京 210029;
    2 重庆交通大学 河海学院,重庆 400074
  • 收稿日期:2018-12-19 修回日期:2019-01-28 出版日期:2019-04-28 发布日期:2020-11-04
  • 通讯作者: ** 陈 诚(1990—),男,江苏宿迁人,硕士,工程师,从事城市水环境和生态环境遥感研究。E-mail: chencheng@nhri.cn。
  • 作者简介:何梦男 (1993—),男,重庆人,硕士,工程师,主要从事洪水灾害人员疏散和城市水环境模型研究。E-mail: mnhe@nhri.cn。
  • 基金资助:
    国家自然科学基金青年基金资助(51609142);中央级公益性科研院所基金资助(Y918011,Y919006);江苏省双创团队项目(SC917001)。

Shortest path optimal algorithm for emergency evacuation based on cellular automata

HE Mengnan1, FU Yuling1,2, CHEN Cheng1, HE Hubin1, LIN Yuqing1, CHEN Qiuwen1   

  1. 1 Center for Eco-Environmental Research, Naning Hydraulic Research Institute, Nanjing Jiangsu 210029, China;
    2 College of Hehai, Chongqing Jiaotong University, Chongqing 400074, China
  • Received:2018-12-19 Revised:2019-01-28 Online:2019-04-28 Published:2020-11-04

摘要: 为科学、高效地规划应急疏散最短路径,建立人员对障碍物的绕行距离场,提出一种基于元胞自动机的全局最短路径(GSPCA)优化算法;设置GSPCA_2D和GSPCA_3D等2种求解模式,并将该算法与基于欧氏距离的启发式搜索算法(A*算法)进行最短路径规划对比试验,分析GSPCA算法的特点和适用性。结果表明:与A*算法相比,GSPCA优化算法搜索结果全局性更好,可避免出现贴壁绕行现象,其最短路径平均缩短12.39 %(GSPCA_2D)、16.82 %(GSPCA_3D);GSPCA算法能考虑地形因素,可根据地形特征和实际应用需求合理选择求解模式。

关键词: 应急疏散, 基于元胞自动机的全局最短路径(GSPCA), A*算法, 欧氏距离, 绕行距离场

Abstract: To plan the shortest emergency evacuation route scientifically and efficiently, an optimal algorithm for GSPCA was put forward by establishing a detour distance field against obstacles, and two solving modes of GSPCA_2D and GSPCA_3D were developed. To analyze the characteristics and applicability of GSPCA, this algorithm was compared with the heuristic search algorithm based on Euclidean distance (A* algorithm) in the shortest path planning experiment. The results show that compared with A* algorithm, GSPCA optimal algorithm is more comprehensive which can avoid the phenomenon of detours by clinging to the wall, and the shortest path is shortened by 12.39 % (GSPCA_2D) and 16.82 % (GSPCA_3D) on average; more importantly, with the terrain factors taken into consideration, GSPCA could select reasonable solution modes according to the terrain characteristics and practical application requirements.

Key words: emergency evacuation, global shortest path based on cellular automata (GSPCA), A* algorithm, Euclidean distance, detour distance field

中图分类号: