(Influencing Factors And Time Modeling Of Crowd Emergency Evacuation)

In recent years, disaster events frequently occur around the world, which cause great casualties and property losses. China is an accident and disaster-prone country. It is estimated that there were360,000national security incidents in2010alone and nearly80,000deaths. However, the accident loss will be greatly reduced if the crew can safely evacuate in the shortest possible time. Therefore, the study of emergency evacuation is of great theoretical and practical value.People tend to gather in groups to form into crowd in the course of action. Crowd consists of individuals with social attributes, which leads to some difference between crowd behavior and individual behavior. Since emergency evacuation often generates the crowd behavior, it is necessary to study the laws of crowd behavior.In general, there are three different analytical dimensions to study the evacuaton behavior. The first is to study the physical location of the evacuation, which mainly refers to the ambience of evacuation scenarios. The second is to examin the evacuation management, including the policies, procedures and evacuation control strategy. The third is to judge the influence of social psychology and social organization characteristics over crowd emergency evacuation. This paper focuses mainly on the analysis of the third dimension. With the knowledge in several disciplinary fields such as safety science, statistics, social psychology and dynamics, three methods below are mainly employed.The first is'System Analysis Method', by which the principles, data processing, tools of modeling and the operation mechanism are discussed.The second is the method of'Comparative Analysis'. With this method, the regression model of evacuation time was compared with the classical model. It shows that social factors of individuals have a significant impact on the emergency evacuation time of crowd.The third is the method of social science, including the research of organizational behavior, statistics and psychology. A regression model of evacuation time was established through investigating and sampling by which two samples that contained216and185sets of data respectively were founded. There are two main purposes in this study:one as revealing how the attributes (including natural and social attributes) of individual affect the evacuation time of group, the other as analyzing and optimizing the internationally accepted model of evacuation time.This study is combined with the Plan of the'Eleventh Five-Year' Emergency Response system of China. Based on the funded projects of the university affiliated with ministry of education (No.1401-2010112), the simulation of crowd evacuation efficiency is studied, which mainly includes:1) Aiming at improving the simulation degree of evacuation time model (briefly called empirical model) that generally accepted by the world, the influences of individual attributes (including natural and social attributes) are analyzed.2) Based on regression analysis and the evacuation investigation on a large university gymnasium, a185-set-data sample is established. Combined with the factors that affect evacuation behavior and time of individuls, the regression model of crowd evacuation time is built with the statistical software SPSS17.0. Since the Artificial Neural Network (ANN) is a nonlinear dynamic system, similar to Logistic Regression analysis, it was used to test the validity of the results of Logit Regression3) The individual attributes are consolidated into crowd attributes and the correction coefficient'k'is solved on the premise of regression model. The empirical model is revised and examined through the use of orrection coefficient'k'. The results show that with the use of regression analysis and correction coefficient'k', the empirical model weakness that lacking of considering the influences of crowd attributes can be effectively filled up, and both the prediction accuracy and application value of evacuation time model will be improved.4) Based on the optimized empirical model, a fire evacuation design of a university stadium project is analyzed. The original fire evacuation partition is improved with the use of empirical model according to the requirements of performance-based fire design. Each evacuation time of the two evacuation plans is obtained, with the use of the optimized empirical model. Finally, the differences of evacuation efficiency between the two plans are also explored.The results show that individual attributes (including natural and social attributes) can affect the properties of the whole group, and ultimately affect the evacuation efficiency and time. However, the individual attributes can't fully explaine the behavioral characteristics of evacuation. It can be concluded that the classic model is idealized as it lacks of consideration of the influence of group properties over evacuation efficiency. The correction coefficient'k', which depends on the establishment of regression model, can be used to amend the classic model. It's clear that with the use of optimized empirical model, fire evacuation designs of large public buildings can be analyzed and the fire evacuation plans can be improved effectively.