主管：中华人民共和国应急管理部
主办：应急管理部天津消防研究所
ISSN 1009-0029  CN 12-1311/TU
消防科学与技术

消防科学与技术 ›› 2021, Vol. 40 ›› Issue (2): 164-168.

• • 上一篇    下一篇

受限空间火灾探测数据融合算法研究

邓力1,2，刘全义1，贺元骅1，王海斌1，胡林1   

  1. 1. 中国民用航空飞行学院民航安全工程学院，四川广汉618307；2. 南京理工大学电光学院，江苏南京210094
  • 出版日期:2021-02-15 发布日期:2021-02-15
  • 通讯作者: 贺元骅（1965-），男，四川广汉人，中国民用航空飞行学院民航安全工程学院教授。
  • 作者简介:邓力（1986-），男，四川成都人，中国民用航空飞行学院民航安全工程学院讲师，主要从事民航飞机火灾探测及防火工程研究，四川省广汉市南昌路4 段46号，618307。
  • 基金资助:
    国家自然科学基金项目（U1933105）；四川省科技计划项目（2020YFG0447）；中国民用航空飞行学院基金项目（J2020-120）

Research on data fusion algorithm of fire detection signal in confined space

DENG Li1,2, LIU Quan-yi1, HE Yuan-hua1, WANG Hai-bin1, HU Lin1   

  1. 1.College of Civil Aviation Safety Engineering,Civil Aviation Flight University of China,Sichuan Guanghan 618307, China; 2. College of Optical and Electronics Engineering, Nanjing University of China, Jiangsu Nanjing 210094,China
  • Online:2021-02-15 Published:2021-02-15

摘要：

为了实现对受限空间火灾发生情况多传感器自动探测信号处理，建立了基于CO 浓度、烟气颗粒浓度、红外视频图像等多传感器探测的火灾探测数据融合分析系统。在受限空间内搭建基于多传感器火灾探测的数据融合分析实验平台。根据火灾红外视频图像的燃烧区域面积及圆形度等特征参数介绍了红外图像识别算法。在三种火灾探测方式特征参数的基础上提出了固定门限功率谱检测算法，并完成数据融合。在分析比较火灾探测系统响应时间、报警准确率等系统性能评价方法的基础上，说明了采用多传感器探测信号进行数据融合并建立灾情判据的方法。实验结果表明：采用数据融合算法后探测系统误报率低于单一探测方式，报警时间略高于单一探测报警方式。该数据融合系统基本满足火灾自动探测报警系统的稳定可靠、低误报率等要求。

关键词: 受限空间, 火灾探测, 数据融合, 红外视频

Abstract:

In order to realize the multi- sensor automatic detection signal processing for the fire occurrence in confined space, a fire detection data fusion analysis system was developed based on multi- sensor detection of CO concentration, smoke particle concentration, infrared video image, etc. First, build a data fusion analysis experimental platform based on multi- sensor fire detection in a confined space. Then, the infrared image recognition algorithm is introduced according to the characteristic parameters such as the area and circularity of the fire infrared video image. Then, based on the characteristic parameters of the three fire detection methods, a fixed threshold power spectrum detection method is proposed, and data fusion is completed. Finally, based on the analysis and comparison of the system performance evaluation methods such as the response time and alarm accuracy of the fire detection system, the method of multi- sensor detection signals for data fusion and the establishment of disaster criteria is explained. The experimental results show that: after using the data fusion algorithm, the false alarm rate of the detection system is lower than the single detection mode, and the alarm time is slightly higher than the single detection alarm mode. The data fusion system basically meets the requirements of stable and reliable automatic fire detection and alarm system and low false alarm rate.

Key words:  confined space, fire detection, data fusion, infrared video