The Study On Optimization Of Sub-district Emergency Evacuation Based On Dynamic Network Flow

This thesis illustrates the study background of Sub-district emergency evacuation from three aspects: macro demand,realistic demand and research demand.Generally speaking,the economic and human losses caused by disasters are increasing;the evacuation plan level fails to correspond to the relevant shelter planning;and the researches on Sub-district evacuation are quite inadequate.The purpose of the study is to propose the optimization strategy and corresponding measures of Sub-district emergency evacuation based on the study results,and then strengthen the detailed and flexible deployment.From the composition,classification,network model and some existing practical experience at home and abroad of Sub-district emergency evacuation,this thesis analyzes its basic content and existing problems as the basis of dynamic network scenario modeling of Sub-district evacuation.Sub-district emergency evacuation should be regarded as a dynamic system,which consists of evacuee,evacuation source node,evacuation route,intermediate node and evacuation destination.Besides,it can be divided into two types: disaster oriented and behavior oriented.From the perspective of route form,Sub-district evacuation network model can be divided into service radius evacuation network and real route evacuation network,and the latter one will be the main research trend.In the part of theoretical research,it is proposed to build the Sub-district evacuation model based on the theory of dynamic network flow,and to divide different evacuation situations,and finally to solve the relative optimal evacuation process of the Sub-district by CCRP algorithm.The modeling optimization and parameter acquisition method of Sub-district evacuation dynamic network is an important focus of the theoretical research,specifically comparing different modeling ideas and summarizing the empirical model of parameter acquisition.Another focus is the Sub-district evacuation situation,which can be divided into four emergency evacuation situations: day / night on weekdays and day / night on non weekdays according to the evacuation time period;divided by the evacuation stage,it includes emergency evacuation,secondary evacuation and tertiary evacuation.In the empirical part,based on the two-tier standards-evacuation allocation gap and Sub-district population density,“Datang Sub-district” is selected from 18 Sub-districts in Yuexiu District as the research scope.Associated with the guidance of theoretical research,the CCRP algorithm will be used to solve the dynamic network of Sub-district emergency evacuation to get evacuation route schedule and time.As far as the results are concerned,the evacuation route schedule is an important basis for optimizing strategies and measures,which can reflect the relatively optimal evacuation process of the Sub-district;the evacuation time reflects the results of different evacuation situations,which is helpful to compare the evacuation laws of them.Therefore,the optimization idea of the study is to find out the relative optimal evacuation process and results,and then put forward further optimization strategies and measures.The empirical study shows that the dynamic network flow theory is applicable to the study of street scale crowd evacuation.It can solve the relative optimal evacuation process and time,and analyze whether the whole evacuation path and the capacity allocation of refuge are reasonable.The evacuation time of Datang in the day and night situation is not different as a whole,but the evacuation time in the daytime on weekdays is longer than that in other situations.Emergency evacuation takes more time than other stages,but there is not a big gap between them.Finally,the optimization strategies and measures are proposed from four aspects: evacuation plan,evacuation path,evacuation safety and evacuation situation.