Fire Safety Science Digital Archive

The characteristics of an improved fire detector which promptly reacts to smoke while discriminating between airborne signatures from fire and non-fire sources can be identified by considering signature patterns of each group of sources. Discrimination is accomplished by comparing signature response patterns from fire: and environmental sources collected in small- and large-scale tests. Airborne signatures are produced in the tests from a variety of conditions: flaming, pyrolyzing and heated samples, and nuisance sources, such as aerosols, household products and cooked food. The signatures are described in terms of light obscuration, temperature, CO, C02, O2 concentrations and signals from metal oxide sensors. An expert system was developed by a neural network and a multivariate statistical method to distinguish between fire and non-fire sources. In addition, the presence of a flaming or nonflaming fire can be identified despite the interjection of signatures from nuisance sources which could mask the fire signatures.