On using autoregressive second order models for pipeline leak detection

Authors:
Marllene Daneti
Affiliations:
Faculty of Electronics and Communication, Applied Electronics Department, "Politehnica" University of Timisoara, Timisoara, Romania
Venue:
ICC'08 Proceedings of the 12th WSEAS international conference on Circuits
Year:
2008

Citing 2
Cited 0

Detection, Estimation, and Modulation Theory: Radar-Sonar Signal Processing and Gaussian Signals in Noise

Detection, Estimation, and Modulation Theory: Radar-Sonar Signal Processing and Gaussian Signals in Noise
Discrete Random Signals and Statistical Signal Processing

Discrete Random Signals and Statistical Signal Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Signals produced by fluids passing through various pipeline transport systems carry information about the system's state along with a certain degree of disturbing noise. This paper addresses the problem of detecting possible leaks in a pipeline system by using the autoregressive (AR) second order parameters of the random signals acquired from the pipeline installation. A case study for some possible situations is developed in order to find the geometrical positions of the AR coefficients. The resulted theoretical model is then compared with the experimental data. Based on these observations, a detection algorithm is developed. The area under the Receiver Operating Characteristics (ROC) curves describes the performance of the proposed detector.