Adaptive estimation of the hydraulic gradient for the location of multiple leaks in pipelines

Leak diagnosis in pipelines using a combined artificial neural network approach

“…In addition, an Extended Kalman Filter (EKF) was also used to detect and locate leaks, incorporating differential equations based on the dynamic model of the pipeline. Here, the friction factor is a function of the flow rate and is determined iteratively through the use of the Swamee-Jain equation (Delgado-Aguiñaga et al, 2016;Verde & Rojas, 2017;Santos-Ruiz et al, 2018a;Rojas & Verde, 2020).…”

Two Simultaneous Leak Diagnosis in Pipelines Based on Input–Output Numerical Differentiation

“…On the other hand, the multi-leak case study has also been considered from two different perspectives: (i) for sequential leaks (non-concurrent case) [ 7 , 8 , 9 ] and (ii) for the simultaneous leak problem (concurrent case) [ 10 ].…”

“…Experimental results show the potential of this approach by allowing to monitor each new leak, no matter where it appears. Following this direction, a scheme is proposed in [ 8 ] for detecting and locating multiple sequential leaks based on a combination of an adaptive observer to identify the hydraulic gradient in real time and a leak location observer to estimate the leak position and its outflow. Experimental results of a pilot pipeline showed a satisfactory estimation in spite of operational changes and leaks.…”

Leak diagnosis in pipelines based on a Kalman filter for Linear Parameter Varying systems

“…(c) because the design is based on the LMI approach, additional convergence constraints based on the Lyapunov theory can be added as proposed in Pletschen and Diepold (2017). Those aspects are the main differences with respect to the EKF widely used in leak diagnosis: (Navarro et al, 2017;Santos-Ruiz et al, 2018;Torres, Besançon, Georges, Navarro, & Begovich, 2011) for single leak cases and Delgado-Aguiñaga, Besançon, Begovich, and Carvajal (2016) and Rojas and Verde (2020) for multi-leak problems, but also for real life applications (Delgado-Aguiñaga & Begovich, 2017), for instance, and finally, (d) a quantitative analysis evidences that the LPV Kalman filter outperforms the classical EKF. This fact, according to the authors' experience, is decisive to reduce the cost of repairing procedures as it is discussed in Delgado-Aguiñaga and , where it is presented a real life leak identification in an aqueduct that is located in Guadalajara city in Mexico, on the basis of an EKF.…”