Abstract
Today the method of treating multi-dimensional information is becoming increasingly important. There is a strong necessity to develop an effective methodology for this purpose. ln this paper,Mahalanobis distance was applied to develop a fire alarm sensing system as an example of treating multi-dimensional information. From an indoor space under normal conditions, i.e., no fire, data were collected to construct a Mahalanobis space as a base point. Additional data under simulated fire and smoke conditions were also collected to calculate the Mahalanobis distances of varying conditions for comparison. From the study, it was concluded that an effective fire alarm system can be designed through the utilization of Mahalanobis distance determined from the data under an abnormal indoor space.
