An-Najah National University

Baker Abdalhaq

 

 
  • Bookmark and Share Email
     
  • Sunday, January 25, 2009
  • Between classical and ideal: enhancing wildland fire prediction using cluster computing
  • Published at:Not Found
  • Abstract One of the challenges still open to wildland fire simulators is the capacity of working under real-time constrains with the aim of providing fire spread predictions that could be useful in fire mitigation interventions. We propose going one step beyond the classical wildland fire prediction by linking evolutionary optimization strategies to the traditional scheme with the aim of emulating an “ideal” fire propagation model as much as possible. In order to accelerate the fire prediction, this enhanced prediction scheme has been developed in a fashion on a Linux cluster using MPI. Furthermore, a sensitivity analysis has been carried out to determine the input parameters that we can fix to their typical values in order to reduce the search-space involved in the optimization process and, therefore, accelerates the whole prediction strategy. Journal Cluster Computing Publisher Springer Netherlands ISSN 1386-7857 (Print) 1573-7543 (Online) Issue Volume 9, Number 3 / July, 2006 DOI 10.1007/s10586-006-9745-4 Pages 329-343 Subject Collection Computer Science SpringerLink Date Tuesday, September 19, 2006
     
  • Bookmark and Share Email
     
Leave a Comment

Attachments

  • No Attachments Found for this Article

PROFILE

Baker Khaldoun Abdalhaq
Cmputer Science
 
Show Full ProfileEnglish CV
 
 
 
Please do not email me if you do not know me
Please do not e-mail me if you do not know me