Fire Sci. Eng. Search

CLOSE


Fire Sci. Eng. > Volume 28(5); 2014 > Article
Fire Science and Engineering 2014;28(5):64-70.
DOI: https://doi.org/10.7731/KIFSE.2014.28.5.064    Published online October 31, 2014.
연기입자의 광학적 특성과 신경망을 이용한 화원분석에 대한 연구
지승욱
영남대학교 대학원
A Study on the Fire Sources Analysis Using the Optical Characteristics of Smoke Particles and Neural Networks
Jee Seung-Wook
Graduate School, Yeungnam University
요약
신경망은 연기입자의 광학적 특징으로부터 화원을 분석할 수 있는 유용한 도구가 될 수 있다. UL268에서 연기감지기 시험에 사용되는 세 가지 화원(종이화원, 목재화원, 인화성 액체화원)들의 광학적 특징으로 신경망을 훈련시켰다. 또한, 소광과 산란을 동시에 측정할 수 있는 연기챔버를 제작하여 연기의 광학적 특징을 얻고 그 결과를 신경망에 입력하였다. 종이화원, 목재화원, 인화성 액체화원을 대상으로 한 실험에서 신경망은 화원을 정확하게 구별하였다. 또한, 종이-목재화원, 종이-인화성 액체화원, 목재-인화성 액체화원과 같은 복합화원을 대상으로 한 실험에서도 화원을 모두 구별하였다.
Abstract
The neural networks were able to be used by analyze fire source with the optical characteristics of smoke particles. The neural networks were learned the optical characteristics for three types test fire (paper, wood, flammable liquid). These test fires which were adopted in this study were also used to performance test of smoke detector according to UL268. A smoke chamber which was able to detect light extinction and scattering simultaneously was created. The optical characteristics of smoke particles were measured by the smoke chamber. And the results were used to input data for the neural networks. The neural networks distinguished the fire source accurately for paper fire, wood fire or flammable liquid fire. The neural networks distinguished accurately the combined fire source such as paper-wood fire, paper-flammable liquid fire or wood-flammable liquid fire.
Key Words: Smoke particles, Extinction, Scattering, Optical properties, Neural networks


ABOUT
BROWSE ARTICLES
EDITORIAL POLICY
AUTHOR INFORMATION
Editorial Office
Room 906, The Korea Science Technology Center The first building, 22, Teheran-ro 7 Gil, Gangnam-gu, Seoul, Republic of Korea
Tel: +82-2-555-2450/+82-2-555-2452    Fax: +82-2-3453-5855    E-mail: kifse@hanmail.net                

Copyright © 2024 by Korean Institute of Fire Science and Engineering.

Developed in M2PI

Close layer
prev next