Ali Mohammad Ahmadi - Physics Department, Sharif University of Technology, Tehran,Iran.
Aboozar Gaffari - Electronics Department, Sharif University of Technology, Tehran, Iran.
Ahmad Amjadi - Physics Department, Sharif University of Technology, Tehran,Iran.
Ali Reza Bahrampour - Physics Department, Sharif University of Technology, Tehran,Iran.

Leak detection and localization are critical issues in gas transportation and gas distribution industry. Small leakages can be determined and localized by NDT techniques that makes saving process more safely. In this paper, time delay estimation using regional and global similarity measure based on residual complexity is employed for acoustical leak detection. As a similarity measure, the residual complexity (RC) leads to robust time delay estimation against correlated noises as well as low samples number. Although, the RC essentially, considers global interdependency in the sequence of signals, the regional interdependency can be employed because the leakage signal is not long range correlated. The sensitivity of regional method to length of window and controlling parameter is investigated. This approach yields robustness as well as calculation simplicity which is useful for leak detection in distribution networks with huge amount of calculation.

Non-destructive Testing, Leak Detection, Acoustic Emission, Time Delay Estimation, Correlated Noise, Residual Complexity, and Regional Similarity Measure