Location of leaks in pipelines using parameter identification tools

TL;DR

This paper presents a method for leak detection in pipelines by identifying model parameters using techniques like the Prediction Error Method and extended Kalman filters, ensuring conditions for effective implementation through frequency analysis.

Contribution

It introduces a parameter identification approach for leak localization in pipelines, combining established methods with frequency evaluation for improved accuracy.

Findings

Successful application of identification methods for leak detection

Frequency analysis confirms conditions for method implementation

Enhanced accuracy in locating pipeline leaks

Abstract

This work proposes an approach to locate leaks by identifying the parameters of finite models associated with these fault events. The identification problem is attacked by using well-known identification methods such as the Prediction Error Method and extended Kalman filters. In addition, a frequency evaluation is realized to check the conditions for implementing any method which require an excitation condition.