Fire Source Localization method based on genetic algorithm

EasyChair Preprint 167

Abstract

In this paper, multiple fire source localization in building environments is approached based on energy distribution of fire by sensor array. A maximum likely-hood estimation method with genetic algorithm is proposed, that describes energy distribution of fire source as Gaussian distribution, and thus derives an energy multi-peak function model and an observation model of sensor nodes. The method facilitates the multiple fire source localization as a global optimization problem, which is explored by genetic algorithm with a global optimal solution. Finally, the simulation result shows the effectiveness of the proposed fire source localization method by the comparison of the traditional EM algorithm and genetic algorithm.

Keyphrases: Optimization, fire source, localization global

@booklet{EasyChair:167,
  author    = {Zheng Hengyi and Wang Xu and Wei Shanbi},
  title     = {Fire Source Localization method based on genetic algorithm},
  doi       = {10.29007/dltg},
  howpublished = {EasyChair Preprint 167},
  year      = {EasyChair, 2018}}