- Author
- Ono, T. | Ishii, H. | Muroi, N.
- Title
- Frequency Response Required for an Intelligent Fire Alarm System.
- Coporate
- Nihon Univ., Tokyo, Japan
- Journal
- Bulletin of Japanese Association of Fire Science and Engineering, Vol. 35, No. 2, 23-30, 1986
- Keywords
- fire alarm systems | response time | sensors | response time
- Abstract
- [ABSTRACT IN ENGLISH] In order to detect a compartment fire in the early stage and provide highly reliable fire information, it is necessary first to sense the fire phenomena as they change moment by moment and then make a judgment about the fire based on the data obtained. The time-domain data of a fire cannot be attained by conventional sensors which are actuated at a fixed threshold level. To obtain such data, analog output type sensors should be used. However, because the behavior of a fire is not usually steady in the course of its progression, the main stream of the fire phenomena may not be precisely extracted if the analog sensor output data are used as they are for fire judgment. In the design of an intelligent fire alarm system, therefore, consideration should be given to the frequency response of the system. Using the results of the analysis of the output data of temperature sensors and smoke sensors obtained from fire experiments, we have investigated the frequency response required for the intelligent system. Our results show that in the early stage of a fire, the main frequency components of the fire phenomena in the course of its progression fall into the frequency band of from DC to 44 mHz for the temperature and from DC to 15 mHz for the smoke density. This implies that the temperature sensor output and the smoke sensor output should be filtered individually through these frequency bands. We also investigated the requirements for digital filters that give the above frequency responses, and have clarified the relationship between the frequency response of the sensor itself, the cut-off frequency of the digital filter, and the data sampling interval. The study shows that precise fire judgment can be made by an intelligent system where the sensor output data are preprocessed, as described above, before being sent to the fire judgment algorithm.