Optimal Sensor Placement for Leak Location in Water Distribution Networks using Evolutionary Algorithms

Abstract

In this paper, a sensor placement approach to improve the leak location in water distribution networks is proposed when the leak signature space (LSS) method is used. The sensor placement problem is formulated as an integer optimization problem where the criterion to be minimized is the number of overlapping signature domains computed from the original LSS representation. First, a semi-exhaustive search approach based on a lazy evaluation mechanism ensures optimal placement in the case of low complexity scenarios. For more complex cases, a stochastic optimization process is proposed, based on either the genetic algorithms (GAs) or particle swarm optimization (PSO). Experiments on two different networks are used to evaluate the performance of the resolution methods, as well as the efficiency achieved in the leak location when using the sensor placement results.

Sensor Placement for Leak Location in Water Distribution Networks using the Leak Signature Space

“…As future work, we would like to extend this work to more complex WDN as the Barcelona network used in (Casillas et al, 2014a), consisting of 3320 nodes. Finally, we have seen that even with the stochastic optimization tools proposed in this paper, the application of the time horizon analysis was not tractable for the Limassol network with more than 3 sensors.…”

“…In practice, for a given network, the demand usually varies along the time and it is important to carry out the leak location analysis taking into account a given time horizon (Casillas et al, 2014a). To address this point, it is proposed to record, for each potential leak node j, the various signatures it has along a time horizon that corresponds to one cycle of the demand pattern.…”

“…This error is not modeled in the theoretical approximation but it will be taken into account in the realistic cases analysed in the experiments. The S matrix has been used in a variety of works (Pérez et al, 2011;Casillas et al, 2014aCasillas et al, , 2012.…”

“…Closer to this work, Pérez et al (2011) used thresholds on the differences of pressures measured to obtain binary matrices that were used to translate the sensor placement problem to an integer programming optimization problem. In Casillas et al (2013), a new approach for sensor placement for leak location in WDN is proposed that is based on the projection-based location scheme proposed in Casillas et al (2012Casillas et al ( , 2014a. Actually, leak location and sensor placement should be considered together since the best placement depends on the method that is used to locate the potential leaks and the efficiency of the leak location depends on the sensor placement.…”

Efficient Sensor Placement for Leak Localization Considering Uncertainties

“…Currently, Casillas et al (2015) placed sensors in an optimal way by minimizing the overlapping signatures in the leak signature space.…”