Search

Maintaining a good quality of service under a wide range of operational management is challenging for water utilities. One of the significant challenges is the location of water leaks in the large-scale water distribution networks (WDN) due to limited data information throughout the system, generally having only flow sensors at the entrance of the system and some pressure sensors in some selected nodes. In addition, most systems do not have a hydraulic model of the network. Therefore, when using the hydraulic model, the presence of model errors such as nodal demand uncertainty and measurement noise can interfere with the performance of the leak location method. This work presents a fully data-driven technique to reduce the area of the leak localization in the WDN, using Graph theory to represent the network. To do so, we have developed a distance clustering with pre-defined centroids that are the sensor pressure information and some selected nodes. Furthermore, some extra pressure information of leaks events in the selected centroids is studied to develop a correlation between the pressure measurement and the event. Finally, the approach is evaluated in real-world water systems and discusses graphical results and key performance indicators.

The detection and localization of leaks in water distribution networks (WDNs) is one of the major concerns of water utilities, due to the necessity of an efficient operation that satisfies the worldwide growing demand for water. There exists a wide range of methods, from equipment-based techniques that rely only on hardware devices to software-based methods that exploit models and algorithms as well. Model-based approaches provide an effective performance but rely on the availability of an hydraulic model of the WDN, while data-driven techniques only require measurements from the network operation but may produce less accurate results. This paper proposes two methodologies: a model-based approach that uses the hydraulic model of the network, as well as pressure and demand information; and a fully data-driven method based on graph interpolation and a new candidate selection criteria. Their complementary application was successfully applied to the Battle of the Leakage Detection and Isolation Methods (BattLeDIM) 2020 challenge, and the achieved results are presented in this paper to demonstrate the suitability of the methods., The authors want to thank the Spanish national project DEOCS (DPI2016-76493-C3-3-R) and L-BEST (Ref. PID2020-115905RB-C21), as well as the Spanish State Research Agency through the María de Maeztu Seal of Excellence to IRI (MDM-2016-0656).

This work presents the design of a three-layer control strategy to regulate the water levels in inland waterways. The upper layer takes into account the tidal period, which defines two different operating modes. A controller is designed for each mode in the intermediate layer, as well as an observer that estimates the states and the disturbances. Finally, the lower layer solves an optimization problem that yields the scheduling of a set of discrete actuators that best approximates the optimal reference. A real case study based on part of the inland waterways in the north of France is used to test the proposed approach and demonstrate its effectiveness

This work presents a distributed model predictive control strategy as an alternative to conventional centralized approaches, which often suffer from implementation issues when applied to large-scale systems. The overall system is partitioned into minimally coupled subsystems based on its structural properties. Then, the coordination among the subproblems is achieved by means of a communication protocol, which allows each local controller to broadcast its solution to the rest of controllers with a coupled variable. The proposed approach is tested on the quadruple-tank process, and its effectiveness is proved by comparing the obtained results to those documented in an existing benchmark.

Inland navigation networks are composed of several artificial canals, usually characterized by no slope. These canals are large-scale systems that can be accurately described by means of the nonlinear Saint-Venant partial differential equations. However, the lack of an analytical solution for these equations is one of the reasons why simplified models have been developed. In this work, the Integrator-Delay-Zero model is used as the input-output model to link the water depths to the discharges. A modeling extension for a canal with inflows and outflows along the water stream is proposed: an overlapping problem is discussed and a criterion choice is defined in order to obtain the general model. With regard to this criterion, a calibration step is needed, both to overcome the natural limitations of the simplified model and to ensure the correct computation of the general model. The application of this modeling approach to a part of the inland navigation network in the north of France serves as the case study for this work. Les voies navigables sont des réseaux de grandes tailles composés de plusieurs rivières canalisées mais également de canaux artificiels. Ces canaux, délimités par des écluses en amont et en aval, sont souvent caractérisés par l'absence de pente. Ils sont donc particulièrement sujets à des phénomènes de résonance qui se caractérisent par l'apparition de vagues de grandes amplitudes provoquées par l'activation des écluses. Pour modéliser ces phénomènes, il est possible d'utiliser les équations de Saint Venant (Chow, 1959). Cependant, leur utilisation pour la conception de lois de commande ou de stratégies de diagnostic des ouvrages reste difficile par la nature même du système considéré. Une alternative consiste en la conception de modèles simplifiés tels que le modèle IDZ (Integrator-Delay-Zero) proposé dans la littérature (Litrico & Fromion, 2004). Il s'agit d'un modèle permettant de reproduire la dynamique entrée-sortie des canaux en prédisant l'effet des débits en amont et en aval du canal étudié sur ses hauteurs d'eau mesurées aux mêmes endroits. Dans cet article, une extension de ce modèle est proposée avec l'objectif de pouvoir prendre en compte plusieurs entrées et sorties situées le long d'un même canal. La méthode permettant l'identification des paramètres du modèle IDZ conduit alors à des redondances de calcul. Ce problème est mis en évidence et discuté, puis une solution basée sur un critère de sélection des paramètres est proposée. Dans un deuxième temps, afin de pouvoir reproduire au mieux la dynamique réelle des canaux de navigation sans pente, en particulier la forte amplitude du premier pic de résonance, une phase de calage du modèle est présentée. Elle consiste à introduire deux coefficients conduisant à une modification légère des valeurs du Zéro et de l'Intégrateur du modèle en tenant compte des sorties du modèle et des mesures disponibles. Le bief de voies navigables Cuinchy-Fontinettes situé dans les Hauts-de-France permet d'illustrer cette démarche, puisqu'il a la particularité de pouvoir être alimenté par l'intermédiaire d'une vanne située environ au centre du bief. Les performances du modèle proposé sont illustrées en reproduisant par simulation un scénario de fonctionnement réel

Inland navigation networks are composed of several artificial canals, usually characterized by no slope. These canals are large-scale systems that can be accurately described by means of the nonlinear Saint-Venant partial differential equations. However, the lack of an analytical solution for these equations is one of the reasons why simplified models have been developed. In this work, the Integrator Delay Zero model is used to obtain the general input–output model for a canal with inflows and outflows along the water stream. An overlapping problem is discussed and a criterion choice is defined in order to obtain the general model. With regard to this criterion, a calibration step is needed, both to overcome the natural limitations of the simplified model and to ensure the correct computation of the general model. The application of this modeling approach to a part of the inland navigation network in the north of France serves as the case study for this work.

Trabajo presentado a la USCID Conference on Planning, Operation and Automation of Irrigation Delivery Systems celebrada en Phoenix, Arizona (US) del 2 al 5 de diciembre de 2014., Automated control of water systems (irrigation canals, navigation canals, rivers etc.) relies on the measured data. The control action is calculated, in case of feedback controller, directly from the on-line measured data. If the measured data is corrupted, the calculated control action will have a different effect than it is desired. Therefore, it is crucial that the feedback controller receives good quality measurement data. On-line fault detection techniques can be applied in order to detect the faulty data and correct it. After the detection and correction of the sensor data, the controller should be able to still maintain the set point of the system. In this paper this principle using the sensor fault masking is applied to model predictive control of open channels. A case study of a reach of the northwest of the inland navigation network of France is presented. Model predictive control and water level sensor masking is applied.

International audience; abstract simple