人群聚集场所的大型活动常常伴随拥挤踩踏事故的发生，造成严重的人员伤亡，研究人群聚集场所行为突变机理和防控方法具有重要的科学价值和现实意义。本项目提出大场景高密度下行人个体运动特征参数提取算法，分析人群聚集场所个体运动特征；通过实地问卷调查和实验研究的方法归纳分析高密度情况下个体感知外界压力产生的心理行为特征；分析人群聚集场所个体和群体危机行为的时空演变过程和特性变化，以此为基础研究个体至群体危机扩散机理，建立大规模人群危机行为演化的数学模型；结合拥挤踩踏事故案例视频，分析人群从稳定行走状态到拥挤踩踏状态的行为特征变化规律，研究高密度人群失稳的临界条件；基于上述研究成果，建立兼顾效率与精度，体现人群危机行为传播演化和高密度人群运动特性的疏散模型；以北京市某地铁站点为例，利用上述建立的疏散模型构建人群运动情景，研究人群聚集场所安全疏散及管控方案，进而为我国人群聚集场所应急管理提供科学支撑。

Large-scale activities in crowds gathered places often result in crowded stampede and cause serious casualties. The study of group behavior mutation diffusion mechanism and control methods in typical crowds gathered place has important scientific value and practical significance. Under the big scene high-density, a characteristic parameter extraction algorithm of pedestrian individual movement will be studied in this project, which is used to analyze the individual characteristic of the crowds gathering places. Then, by field survey and experimental research methods, we study the individual psychological behavior characteristics caused by outside pressure in high-density cases. And the evolution process and characteristic changes of space and time are analyzed for individual behavior and group crisis. On the basis of, one contagion mechanism from individual to groups is deeply studied; a mathematic model of the mass population crisis behavior is evaluated. For the crowded stampede case video, we study the change rule and critical conditions between stable state and crowded stampede state. Based on the above research results, an evacuation model is established, which are both efficiency and accuracy, and can reflect the population crisis spread behavior evolution and high density population movement characteristics. In the end, this project will study the safety evacuation and control scheme in the crowd gathering place, such as a typical subway stations in Beijing emergency management. We hope this project can provide a scientific basis for emergency management in crowded places.