Pipeline monitoring using acoustic principal component analysis recognition with the Mel scale

摘要: In modern cities, many important pipelines are laid underground. order to prevent these lifeline infrastructures from accidental damage, monitoring systems becoming indispensable. Third party activities were shown by recent reports be a major cause of pipeline damage. Potential damage threat the can identified detecting dangerous construction equipment nearby studying surrounding noise. Sound recognition technologies used identify them their sounds, which easily captured small sensors deployed along pipelines. Pattern classification methods based on principal component analysis (PCA) recognize sounds road cutters. this paper, Mel residual, i.e.?the PCA residual in scale, is proposed feature. Determining if sound belongs cutter only requires checking how large its is. Experiments conducted and results showed that Mel-residual-based worked very well. The method will useful for breakage ensure safety underground infrastructures.