- Author
-
Milke, J. A.
- Title
- Monitoring Multiple Aspects of Fire signatures for Discriminating Fire Detection. BFRL Fire Research Seminar. VHS Video.
- Coporate
- Maryland Univ., College Park
- Report
-
Video
June 20, 2000
- Keywords
-
fire detection
- Abstract
- Incorporating intelligence into a fire detector to recognize signature patterns is intended to provide the capability for prompt fire detection while discriminating between signatures from fire and non-fire sources. The primary purpose of this preliminary study is to investigate the patterns of signatures associated with fire and environmental sources via small-scale experiments. Products are generated from a wide range of conditions: flaming or pyrolyzing samples, heated samples and with an atomizer. Measurements of gas concentrations, light obscuration and temperature are collected to characterize the products. As a result of an analysis of the data, trends are identified from which an elementary expert system can be formulated to identify the source of the air-borne products. The following patterns are evident: the maximum CO2 concentrations achieved during experiments with flaming fires are significantly greater than the maximum CO2 concentrations from experiments with non-flaming fires ( pyrolyzing fires, heated liquids and environmental odors). The non-flaming sources can be identified based on the CO and metal oxide sensor peak measurements. Except for three experiments with pyrolyzing solids, the peak CO concentration is greater, tho ugh the Taguchi detector response less, for non-flaming fires than for environmental sources. Subsequent application of a neural network properly classifies all except one pyrolyzing fire.