Optimising Pedestrian Flow Around Large Stadiums
Optimising Pedestrian Flow Around Large Stadiums
复制标题
DOI：
10.17815/cd.2021.117
复制
发表时间：
2021-12
影响因子：
--
通讯作者：
Yuming Dong;Xiaolu Jia;D. Yanagisawa;K. Nishinari
中科院分区：
文献类型：
--
作者：
Yuming Dong;Xiaolu Jia;D. Yanagisawa;K. Nishinari
本研究提出了一种结合元胞自动机模型和差分进化算法的方法来优化大型体育场周围的人流。通过元胞自动机模型构建了大型体育场及其周边区域的微型版本。应用特殊机制来影响从某个体育场大门离开的特工的行为。代理可能会被附近的商业设施吸引和/或被引导到不拥挤的区域。然后使用差分进化算法来确定每个体育场大门影响因素的最佳概率。主要目标是减少疏散时间，还考虑其他目标，例如减少影响主体行为的成本和个人疏散时间。我们发现，尽管在不同场景下的工作方式有所不同，但智能体的吸引和引导显着缩短了疏散时间。通过对商业设施的适度吸引力和对绕行路线的有力引导，实现了最佳疏散时间。结果表明，所提出的方法可以提供一种依赖于目标、特定于出口的策略，否则很难获得这种策略来优化行人流量。
This study proposes a method that combines the cellular automaton model and the differential evolution algorithm for optimising pedestrian flow around large stadiums. A miniature version of a large stadium and its surrounding areas is constructed via the cellular automaton model. Special mechanisms are applied to influence the behaviour of an agent that leaves from a certain stadium gate. The agent may be attracted to a nearby business facility and/or guided to uncongested areas. The differential evolution algorithm is then used to determine the optimal probabilities of the influencing agents for each stadium gate. The main goal is to reduce the evacuation time, and other goals such as reducing the costs for the influencing agents’ behaviours and the individual evacuation time are also considered. We found that, although they worked differently in different scenarios, the attraction and guidance of agents significantly reduced the evacuation time. The optimal evacuation time was achieved with moderate attraction to the business facilities and strong guidance to the detouring route. The results demonstrate that the proposed method can provide a goal-dependent, exit-specific strategy that is otherwise hard to acquire for optimising pedestrian flow.