A carregar...
Publicação
Modelo de Simulação para Avaliar a Prontidão de Bombeiros: Um Estudo de Caso
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 14.39 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Em Portugal, a maioria dos corpos de bombeiros são associações humanitárias que
surgiram da necessidade e da vontade popular. Consequentemente, não existiram estudos
anteriores à criação destes corpos de bombeiros de modo a perceber-se quais seriam
as localizações em que se deveriam construir os quartéis de cada corpo. Isto originou
uma discrepância nas várias localizações dos quartéis, existindo algumas zonas do país
com excesso de quartéis e outras onde que faltam quartéis. Esta dissertação surge da
necessidade de se perceber qual a eficiência das localizações atuais dos quartéis bem como
quais as zonas que se encontram mais desprotegidas. Para tal, implementou-se um modelo
de simulação que modela o processo de resposta dos bombeiros e o comportamento dos
incêndios urbanos de modo a se poder analisar o impacto das diferentes localizações dos
quartéis na eficiência da resposta. Foram analisadas as atuais localizações dos quartéis
assim como possíveis novas localizações determinadas por modelos de otimização. Foram
ainda testadas diferentes regras para a alocação dos incêndios aos quartéis.
No modelo de simulação foram utilizados dois métodos para gerar os tempos de
viagens, sendo um deles com recurso às distâncias euclidianas e o outro utilizando as
distâncias obtidas através das Google Maps APIs, chegando-se à conclusão que ambos
os métodos providenciam resultados consideravelmente diferentes. O modelo premitiu
ainda analisar o impacto da utilização de uma distribuição dos incêndios tendo em conta
a densidade populacional ao invés da habitual distribuição uniforme das ocorrências
utilizada na maioria da literatura.
O modelo de simulação foi aplicado ao distrito do Porto, sendo que os principais
resultados mostram que existe uma distribuição incorreta dos quartéis de bombeiros
pelo distrito, verificando-se que a percentagem de incêndios cobertos é no melhor dos
casos de 65%. Por outro lado, mostrou-se que é possível aumentar em cerca de 10% a
cobertura alterando as regras de alocação. Relativamente aos métodos utilizados para
distribuir as ocorrências verificou-se que estas apresentam diferenças significativas na
cobertura verificada, sendo esta diferença de 10%. No entanto, devido à falta de dados
não foi possível concluir qual o método que melhor se adequa ao sistema em estudo. Por
fim conclui-se que as localizações provenientes dos modelos de otimização não são mais eficazes que as atuais localizações dos quartéis e como tal estes modelos precisam de ser
melhorados.
This work comes from the need to understand the efficiency of the current locations of fire stations in Portugal, as well as which areas are mostly unprotected. To this end, a simulation model was implemented to review the firefighters response process and behavior of urban fires, in order to analyze the impact and response efficiency, according to their different locations. The current fire station locations were analyzed, as well as possible new locations determined by optimization models. New rules for the allocation were also tested. Two methods were used in the simulation model in order to generate the travel times, one of them uses the euclidean distances and the other method is by using the distances obtained through Google Maps APIs, coming to the conclusion that both methods provide considerably different results. With this model it was also analyzed the impact of using distribution of fires considering the population density instead of the usual method used in most of the literature. This simulation model was applied in the Porto district, the main results showed what was already expected, there was a wrong distribution of fire stations through the district, where the percentage of contained fires is 65% in the best case. On the other hand, it was determined that it is possible to increase coverage by 10% by changing the allocation rules. Regarding the methods used to distribute the occurrences, it was found that they present significant differences in the coverage verified, this difference being 10%, however due to the lack of data it was not possible to conclude which is the best method to use in the simulation. Finally, it is concluded that the locations from the optimization models are not more effective than the current fire station locations, and as such, these models need to be improved.
This work comes from the need to understand the efficiency of the current locations of fire stations in Portugal, as well as which areas are mostly unprotected. To this end, a simulation model was implemented to review the firefighters response process and behavior of urban fires, in order to analyze the impact and response efficiency, according to their different locations. The current fire station locations were analyzed, as well as possible new locations determined by optimization models. New rules for the allocation were also tested. Two methods were used in the simulation model in order to generate the travel times, one of them uses the euclidean distances and the other method is by using the distances obtained through Google Maps APIs, coming to the conclusion that both methods provide considerably different results. With this model it was also analyzed the impact of using distribution of fires considering the population density instead of the usual method used in most of the literature. This simulation model was applied in the Porto district, the main results showed what was already expected, there was a wrong distribution of fire stations through the district, where the percentage of contained fires is 65% in the best case. On the other hand, it was determined that it is possible to increase coverage by 10% by changing the allocation rules. Regarding the methods used to distribute the occurrences, it was found that they present significant differences in the coverage verified, this difference being 10%, however due to the lack of data it was not possible to conclude which is the best method to use in the simulation. Finally, it is concluded that the locations from the optimization models are not more effective than the current fire station locations, and as such, these models need to be improved.
Descrição
Palavras-chave
Simulação Quartéis de Bombeiros Problema de Localização