基于地面场粒子群优化算法的高密度人群应急疏散建模

发布时间：2018-02-02 04:20

本文关键词： 地面场模型粒子群优化模型密集人群信息物理系统仿真推演　出处：《计算机应用》2017年12期 　论文类型：期刊论文

【摘要】：针对非常规突发事件环境下高密度人群的拥挤管理和快速疏散问题,提出一种由感知层、传输层、计算层和应用层构成的多层结构人群疏散信息物理系统(E-CPS)体系框架。在E-CPS体系框架计算层中将静态地面场(FF)建模规则引入经典粒子群优化(PSO)模型,提出地面场PSO(FF-PSO)人群疏散模型,该模型同时具备静态场规则简单、计算快和PSO模型快速搜索、快速收敛的优点。此外,FF-PSO模型中构建了一种新的适应度函数,实现了疏散策略的动态选择,并通过数值仿真及实例仿真验证了FF-PSO模型在拥挤管理中的可行性和有效性。国家会展中心(上海)的实例仿真结果表明,考虑拥堵管理比仅考虑距离最短平均每分钟可多疏散66人,疏散时间节省19 min,疏散效率提升13.4%。
[Abstract]:In order to solve the problem of high density crowd congestion management and rapid evacuation in unconventional emergency environment, a new method is proposed, which is composed of perceptual layer and transmission layer. An E-CPS (E-CPS) system, which is composed of computing layer and application layer, is used to calculate the static ground field in the framework of E-CPS. The classical particle swarm optimization (PSO) model is introduced into the modeling rules. A crowd evacuation model for ground field PSO-FF-PSOs is proposed. The model has the advantages of simple static field rules, fast computation and fast search of PSO model. In addition, the model has the advantages of fast convergence. A new fitness function is constructed in the FF-PSO model, which realizes the dynamic selection of evacuation strategy. The feasibility and effectiveness of FF-PSO model in congestion management are verified by numerical simulation and case simulation. The simulation results of the National Convention and Exhibition Center (Shanghai) show that the model is feasible and effective. Considering congestion management, 66 more people can be evacuated per minute than the shortest distance, and the evacuation time is saved 19 minutes, and the efficiency of evacuation is improved by 13.4.
【作者单位】：同济大学CIMS研究中心;
【基金】：国家自然科学基金资助项目(71573190)~~
【分类号】：TP18;TU998.1
【正文快照】： 0引言随着城市居民人口密度的不断增大,城市公共场所的人群安全问题日益突出。在大型交通枢纽、公共展馆等人数众多的场所,若突发火灾、暴恐等突发事件,将造成严重后果。因此,研究突发事件下的高密度人群疏散问题至关重要至关重要。目前,人群疏散领域的研究方法主要有两种:1)

【相似文献】