An automatic pipeline monitoring system using sound information

摘要: In modern cities most pipelines such as those for oil, gas, and water supply networks are buried underground. order to prevent these lifeline infrastructures from being broken accidentally, monitoring systems becoming indispensable. Recent reports show that pipeline damage is caused by third-party activities. this paper, a novel automatic system proposed in accidental damage. study, potential threat integrity recognized detecting the existence of road cutters, which actually prelude ground construction. Sound recognition technologies used identify cutters sound, can easily be captured small sensors installed along pipelines. A pattern classification method based on Mel frequency cepstral coefficient (MFCC) feature study cutter sounds. The location well detected knowing sensor sounding an alarm relevant measures then expediently executed. Experiments were conducted resulting data analyzed. Results showed effectively MFCC distance. This deployed with low cost poses no significant privacy problems nearby residents. It help breakage thus ensuring longevity underground infrastructures. Copyright © 2008 John Wiley & Sons, Ltd.