| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 4.01 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Com a expansão contínua da escala da cidade e o rápido crescimento da população urbana, as atividades sociais e económicas das pessoas e as atividades de intercâmbio cultural são cada vez mais frequentes. No entanto, devido aos ajuntamentos em grande escala e a situações imprevisíveis, ocorrem frequentemente numerosas vítimas em acidentes fatais. O planeamento da evacuação potencial de um grande número de
pessoas de uma zona devido a uma emergência ou ameaça é uma aplicação em que a modelação baseada em agentes tem sido utilizada em simulações para ajudar a compreender os potenciais problemas. A presente dissertação tem como objetivo a análise de diferentes estratégias de evacuação num centro de congressos/ eventos através de uma análise extensiva do espaço, recorrendo ao software Pathfinder a fim, de facilitar o processo, e identificar possíveis pontos de estrangulamento, analisar a evacuação em si, sugerindo melhorias e analisando os resultados. Foram assim analisadas 6 estratégias diferentes, para determinar o menor tempo de evacuação possível, identificando pontos de gargalo e locais de congestionamento dos ocupantes. O modelo foi utilizado para evacuar a ala poente do piso 2 da Alfândega do Porto com lotação máxima em todas as salas, fazendo várias combinações com as saídas disponíveis. Posteriormente, realizou-se o mesmo
para a zona este/ centro e oeste do piso em análise. Os resultados obtidos nas simulações permitem concluir que a evacuação mais eficiente é a situação em que temos todas as saídas disponíveis, uma vez que permite
um fluxo variado por todo o espaço, sendo que os ocupantes se espalham e não provocam congestionamento nas portas intermédias. Por vezes a indecisão das pessoas atrasa todo o processo; assim esta evacuação tem um tempo total de 678 segundos equivalente a 11 minutos aproximadamente. Conclui-se assim que a utilização de portas de emergência com portas duplas, permite um maior fluxo de pessoas por segundo, que certamente reduziria substancialmente o tempo total de evacuação.
With the continuous expansion of the city's scale and the rapid growth of the urban population, people's social and economic activities and cultural exchange activities are becoming more and more frequent. However, due to large-scale gatherings and unpredictable situations, there are often numerous victims in fatal accidents. Planning for the potential evacuation of large numbers of people from an area due to an emergency or threat is an application in which agent-based modeling has been used in simulations to help understand the potential problems. This dissertation aims to analyze different evacuation strategies in a congress/event center through an extensive analysis of the space, using Pathfinder software to facilitate the process, identify possible bottlenecks, analyze the evacuation itself, suggest improvements and analyze the results. Six different strategies were analyzed to determine the shortest possible evacuation time, identifying bottlenecks and places where occupants were congested. The model was used to evacuate the west wing of floor 2 of Alfândega do Porto with maximum capacity in all the rooms, making various combinations with the available exits. Subsequently, the same was done for the east/center and west areas of the floor under analysis. The results obtained in the simulations allow us to conclude that the most efficient evacuation is the situation in which we have all the exits available, since it allows a varied flow throughout the the entire space, with the occupants spreading out and not causing the intermediate doors. Sometimes people's indecision slows down the whole process, so this evacuation takes a total of 678 seconds, equivalent to approximately 11 minutes. It can therefore be concluded that the use of double emergency doors allows for a greater flow of people per second, which would certainly substantially reduce the total evacuation time.
With the continuous expansion of the city's scale and the rapid growth of the urban population, people's social and economic activities and cultural exchange activities are becoming more and more frequent. However, due to large-scale gatherings and unpredictable situations, there are often numerous victims in fatal accidents. Planning for the potential evacuation of large numbers of people from an area due to an emergency or threat is an application in which agent-based modeling has been used in simulations to help understand the potential problems. This dissertation aims to analyze different evacuation strategies in a congress/event center through an extensive analysis of the space, using Pathfinder software to facilitate the process, identify possible bottlenecks, analyze the evacuation itself, suggest improvements and analyze the results. Six different strategies were analyzed to determine the shortest possible evacuation time, identifying bottlenecks and places where occupants were congested. The model was used to evacuate the west wing of floor 2 of Alfândega do Porto with maximum capacity in all the rooms, making various combinations with the available exits. Subsequently, the same was done for the east/center and west areas of the floor under analysis. The results obtained in the simulations allow us to conclude that the most efficient evacuation is the situation in which we have all the exits available, since it allows a varied flow throughout the the entire space, with the occupants spreading out and not causing the intermediate doors. Sometimes people's indecision slows down the whole process, so this evacuation takes a total of 678 seconds, equivalent to approximately 11 minutes. It can therefore be concluded that the use of double emergency doors allows for a greater flow of people per second, which would certainly substantially reduce the total evacuation time.
Description
Keywords
Evacuation strategies Congress center Evacuation time PathFinder Simulations Estratégias de evacuação Centro de congressos Tempo de evacuação PathFinder Simulações