Signal processing algorithms for fire localization using temperature sensor arrays

This paper proposes signal processing algorithms to resolve the problem of robust fire source localization under planar or circular wavefronts of hot gas conditions. The approach is based on the signal time delay estimates and replicas of a source signal at spatially distributed temperature sensor arrays. According to different planar and circular wavefronts, the far-field and near-field algorithm can be used, respectively, to identify the position of the fire source. On the one hand, experiments indicate that the far-field method provides acceptable errors for fire localization. On the other, by using a near-field algorithm, we can increase the probability of correct fire source localization compared to using only the far-field approach.