The following information was submitted:
Transactions: WSEAS TRANSACTIONS ON CIRCUITS AND SYSTEMS
Transactions ID Number: 27-648
Full Name: Marllene Daneti
Position: Ph.D. Candidate
Age: ON
Sex: Female
Address: 1 Decembrie 11 Ap.3 300231, Timisoara
Country: ROMANIA
Tel: +40-356-441763
Tel prefix: +40-356
Fax:
E-mail address: marllene.daneti@etc.upt.ro
Other E-mails: marllenedaneti@gmail.com
Title of the Paper: A Second Order Autoregressive Based Technique for Pipeline Leak Detection
Authors as they appear in the Paper: Marllene Daneti
Email addresses of all the authors: marllene.daneti@etc.upt.ro
Number of paper pages: 10
Abstract: Efficient leak detection techniques need to be characterized both by rapidity and robustness. This paper studies a simple detection method based on the second order autoregressive (AR) parameters of the pipeline signals- a trade-off between the two required characteristics. The theoretical geometrical positions defined by the AR coefficients are developed for some particular cases of interest. The resulted model proves to be in concordance with the experimental data. The algorithm is next tested under non-stationary burst-type conditions induced both soft and hard, indicating a good stability in comparison with the basic most inexpensive computational detection method. The area under the Receiver Operating Characteristics (ROC) curves indicates also a better performance of the proposed detector.
Keywords: Leak detection,Pipeline monitoring, AR models, ROC curves, Non-stationary perturbations
EXTENSION of the file: .pdf
Special (Invited) Session: On Using Autoregressive Second Order Models for Pipeline Leak Detection
Organizer of the Session: 591-472
How Did you learn about congress:
IP ADDRESS: 86.127.190.134
