Article
Genetic Algorithm Based Simulation–optimization For Fighting Wildfires Open Access Deposited
Wildfire is one of the most significant disturbances responsible for reshaping the terrain and changing the ecosystem of a particular region. Its detrimental effects on environment as well as human lives and properties, and growing trend in terms of frequency and intensity of wildfires over the last decade have necessitated the development of efficient forest fire management techniques. During the last three decades, Forest Fire Decision Support Systems (FFDSS) have been developed to help in the decision-making processes during forest fires by providing necessary information on fire detection, their status and behavior, and other aspects of forest fires. However, most of these decision support systems lack the capability of developing intelligent fire suppression strategies based upon current status and predicted behavior of forest fire. This paper presents an approach for development of efficient fireline building strategies via intelligent resource allocation. A Genetic Algorithm based approach has been proposed in this paper for resource allocation and optimum fireline building that minimizes the total damage due to wildland fires. The approach is based on a simulation–optimization technique in which the Genetic Algorithm uses advanced forest fire propagation models based upon Huygens principles for evaluation of cost index of its solutions. Both homogeneous and heterogeneous environmental conditions have been considered. Uncertainties in weather conditions as well as imperfect knowledge about exact vegetation and topographical conditions make exact prediction of wildfires very difficult. The paper incorporates Monte-Carlo simulations to develop robust strategies in uncertain conditions. Extensive simulations demonstrate the effectiveness of the proposed approach in efficient resource allocation for fighting
complex wildfires in uncertain and dynamic conditions.
- Creator
- License
- Submitter
- College
- Department
- Date Created
- Journal Title
- International Journal of Computational Methods
- Language
- Note
This work was part of a pilot "mediated-deposit model" where library staff found potential works, later submitted for faculty review
Digital Object Identifier (DOI)
Identifier: 10.1142/S0219876213500357
Link: https://doi.org/10.1142/S0219876213500357
This DOI link is the best way for others to cite your work.
Items
Thumbnail | Title | Date Uploaded | Visibility | Actions |
---|---|---|---|---|
|
J2013_C.pdf | 2017-02-13 | Open Access |
|
Permanent link to this page: https://scholar.uc.edu/show/bc386v31q