Fire source positioning method based on statistic characteristics of sensor array

The invention relates to a fire source positioning method based on statistic characteristics of a sensor array. A positioning result of the traditional method is directly output point by point and is scattered in picture display, and the uncertain range is large, so that the traditional method is difficult to directly apply fire-fighting direction. The method disclosed by the invention comprises the following steps of: carrying out grey correlation degree analysis on temperature signals collected in the sensor array so as to obtain smoke propagation delay; then obtaining azimuth angle estimation according to a propagation model and a geometrical relationship; and, regarding the azimuth angle estimation as a random vibration defined by a mean value and a variance on the basis of a statistical modeling method, so that fire source point estimation is converted to an area range estimation of a fire source; and finally, establishing an area range of fire source positioning estimation in a two-group array intersection manner, and taking an area center as point estimation of a position of the fire source. According to the expansion estimation based on a statistical modeling idea, overallhistory data can be used, an estimated result has a better stability along the increasing of data quantity, and random errors are effectively reduced.