The application provides a self-adaptive
smoke fire detection method suitable for ancient
building environment, and relates to the technical field of
fire detection, and comprises the following steps: collecting
smoke concentration data of a
smoke detector in an ancient
building environment, removing
invalid data by using an SPC
statistical process control method to obtain an optimal
data set; based on the optimal
data set, calculating the correlation support degree between each smoke
detector, and constructing a
support matrix; the application effectively removes
invalid data by using an SPC
control chart, objectively evaluates the reliability of the
detector based on the correlation support degree, dynamically integrates multi-sensor information by using a self-adaptive weighted fusion
algorithm, and designs a trend threshold
algorithm and a field model analysis
algorithm for single detector and multi-detector scenes respectively, so that the fire determination strategy and the alarm threshold can be adaptively adjusted according to real-time environmental parameters and interference characteristics, the high sensitivity to real fire is retained, and frequent false alarms caused by
traditional use of fire, ventilation
airflow and
humidity change are effectively inhibited.