Research on Detection Method of Uncertainty Fire Signal Based on Fire Scenario

For a safe evacuation of people in fire, which is a serious disaster to damage the security of people´s lives and possessions, the fire signal for alarm is accurately detected and recognized in a period of early or extremely early fire time which people wanted in possible. At the same time, the accuracy and validity of fire detection is greatly weakened by reason of increasing uncertainty of fire signal. Therefore a new approach is presented for the detection of uncertainty fire signal based on fire scenario, which its algorithm in the detection of one-dimension fire space is given in order to show the proposed method having fine fault-tolerance, robustness and accuracy. In this method, a certain character of fire space fractal in the detected fire space is obtained firstly by a performance analysis of fire scenario, and then a character of signal space fractal is acquired by using intelligent method based on rough set theory and trend fusion to process the multi-source signal detected from a group of fire detectors, when a logical coincident relationship between the two characters recognized, a real fire alarm is produced lastly