ADVANCED FIRE DETECTION BASED ON COMBINED CONVENTIONAL SMOKE AND CO SENSORS

CURRENT SMOKE DETECTION TECHNOLOGIES SUFFER FROM SIGNIFICANT FALSE ALARM RATES. THESE PROBLEMS ARE CURRENTLY BEING ADDRESSED BY REDUCING THE SENSITIVITY OF EXISTING PARTICULATE DETECTORS. IN THIS WORK, THE FEASIBILITY OF A COMBINED PARTICULATE AND CARBON MONOXIDE DETECTOR WILL BE EVALUATED. THE USE OF MULTIPLE FIRE SIGNATURES HAS THE POTENTIAL FOR REDUCING THE FALSE ALARM PROBLEMS WHILE INCREASING THE SESNSITIVITY OF THE DETECTOR. AN EXPERIMENTAL PROGRAM OF FIRE SIGNATURE MEASUREMENTS IN RESPONSE TO BOTH FIRE AND NON-FIRE SOURCES WILL BE CARRIED OUT. ADVANCED SIGNAL PROCESSING METHODS COUPLED WITH THE MULTIPLE SIGNATURE MEASUREMENTS WILL BE USED TO IMPROVE DETECTOR PERFORMANCE. THE RESULT WILL BE A NEW SMOKE DETECTION TECHNOLOGY WITH THE CAPABILITY OF IMPROVED SESNSITIVITY AND SUBSTANTIALLY REDUCED FALSE ALARM RATES.