A fire sensor system based on the simultaneous detection
of CO, CO2, and smoke concentrations is demonstrated.
The rates of increase of these three components are used in
the fire alarm algorithm to determine the presence of a fire.
The algorithm monitors the rate of increase of smoke level,
and when this rate exceeds its threshold rate, the rates of
increase of CO and CO2 concentrations are checked. When
either the rate of increase of CO or CO2 concentration
exceeds its threshold rate, a fire alarm is initiated. The fire
detection system was found to perform better than a smoke
detector operating alone. In cases where the smoke
detector did not alarm, the algorithm was able detect the
fire. However, in cases where the smoke detector did alarm,
the algorithm detected the fire in a much shorter time. The
nuisance sources did not cause the fire detection system to
generate false alarms.
An advantage of the fire alarm algorithm is that the
signal to noise ratio for the temporal derivatives of
measurements improves more rapidly with increasing time
window length than does the signal-to-noise ratio for the
mean of data points within the time window; this allows the
alarm threshold to be set at a more sensitive value without
causing noise-generated false alarm. A second advantage is
that the alarm point is insensitive to constant offsets in the
gas concentration measurements.
Although, the fire detection system was originally
conceived to be deployed in cargo compartments of
aircraft, it can also be applied to buildings, ship compartments,
submarines, living compartments in space, and
concealed cavities used for running electrical wires and
plumbing. Future work will focus on the improvement of
the minimum detection of CO and CO2 concentrations by
using separate lasers for each chemical species, and
selecting stronger absorption lines where these two
chemical species do not overlap.