This paper details an experimental program that gathered photoelectric and ionization analog output smoke sensor, thermal sensor, and an electrochemical CO sensor output signals during exposure to various fire and non-fire stimuli reproduced in the fire emulator/detector evaluator. Three fire sources: flaming fire, smoldering cotton (Similar to EN54 TF3) and pyrolyzing wood blocks (similar to EN54 TF2) were emulated in the fire emulator/detector evaluator (FE/DE) along with five nuisance sources: cigarette smoke, smoke from bread toasting, dust, high humidity/condensing water vapor, and smoke from heated cooking oil, to represent typical household fire and nuisance sources. Both the bread toasting and heated cooking oil were allowed to progress to a point well beyond that required for cooking activities, and subsequently transitioned to hazardous conditions that might be experienced during unattended cooking (heavy smoke and CO production). The sensor signals were used to develop a combined CO/smoke detection algorithm using fuzzy logic classification that provided early detection in the fire cases or hazardous cooking conditions and improved discrimination of the nuisance sources.
Citation: Special Publication (NIST SP) - 965Report Number:
NIST Pub Series: Special Publication (NIST SP)
Pub Type: NIST Pubs
fire detection, fire detection systems, smoke, carbon monoxide, sensors, algorithms, signals