Publicação
Prepositioning of resources for fire suppression problem under uncertainty
| Resumo: | Wildfires are natural events that, in recent decades, have been increasingly more frequent and severe, and when not contained, can result in great losses. The operational management of firefighting resources is essential in effective wildfire responses. This management is assisted by different optimisation techniques, which means that the development of suitable optimisation models is crucial. This document presents the study done on the use of different optimisation approaches and their appplicability in the prepositioning of resources under uncertainty. The uncertainty present in this problem is mainly represented by the wind and, in order to create scenarios that describe the uncertainty of the climate characteristics, we carried out a statistical analysis of real data obtained by a weather station near the case study area. The approaches studied were: stochastic optimisation, robust optimisation and distributionally robust optimisation. In the stochastic optimisation approach that was taken the expected burned area is minimized. The robust optimisation approach prepositions the firefighting resources such that the burned area in the worst-case is minimized. Lastly, the distributionally robust optimisation approach includes traits from both of the previous techniques by minimizing the impact of the worst-case probability distribution (of the scenarios). In order to perform computational tests, we generated instances based on real data and the scenarios obtained by the statistical analysis. For every model, we were able to obtain the location of the optimal prepositioning for the firefighting resources. |
|---|---|
| Autores principais: | Marques, Francisco José Coelho |
| Assunto: | Wildfire Firefighting problem Uncertainty Prepositioning Optimisation Wind modelling |
| Ano: | 2024 |
| País: | Portugal |
| Tipo de documento: | dissertação de mestrado |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade de Aveiro |
| Idioma: | inglês |
| Origem: | RIA - Repositório Institucional da Universidade de Aveiro |
| Resumo: | Wildfires are natural events that, in recent decades, have been increasingly more frequent and severe, and when not contained, can result in great losses. The operational management of firefighting resources is essential in effective wildfire responses. This management is assisted by different optimisation techniques, which means that the development of suitable optimisation models is crucial. This document presents the study done on the use of different optimisation approaches and their appplicability in the prepositioning of resources under uncertainty. The uncertainty present in this problem is mainly represented by the wind and, in order to create scenarios that describe the uncertainty of the climate characteristics, we carried out a statistical analysis of real data obtained by a weather station near the case study area. The approaches studied were: stochastic optimisation, robust optimisation and distributionally robust optimisation. In the stochastic optimisation approach that was taken the expected burned area is minimized. The robust optimisation approach prepositions the firefighting resources such that the burned area in the worst-case is minimized. Lastly, the distributionally robust optimisation approach includes traits from both of the previous techniques by minimizing the impact of the worst-case probability distribution (of the scenarios). In order to perform computational tests, we generated instances based on real data and the scenarios obtained by the statistical analysis. For every model, we were able to obtain the location of the optimal prepositioning for the firefighting resources. |
|---|