An Enhanced Long Range Pipeline Online Monitoring Scheme that uses Minimum Variance Cepstrum

Water transmission mains are crucial infrastructures. Most of them are underground, which makes hard to detect pipeline faults. Primary aim of pipeline management might be preventing or detecting in advance catastrophic accidents. Big amount of impact phenomena, such as pressure transients, have been known as major contributors in pipeline failures. Moreover, impact damages, even they are small, could worsen the faults in pipelines. However, detecting an impact signal is not easy because in general the environmental noise significantly corrupts the signal. In this study, a method to enhance detectability is proposed to localize where the impact source is. The proposed method leverages the minimum variance cepstrum to reduce background noise and directly provide the arrival time difference between monitoring sensors in quefrency domain. Computer simulations and field experiments demonstrate that the proposed approach is able to detect the impact location.