Regularization of an Inverse Problem for Parameter Estimation in Water Distribution Networks

Abstract

An accurate hydraulic model of a water distribution network (WDN) is a critical prerequisite for a multitude of operational, optimisation and planning tasks. The accuracy of a hydraulic model can only be maintained through its periodic calibration and validation with acquired pressure and flow data from a WDN. It is important that this process is robust and computationally efficient. This paper describes the regularisation of an inverse problem to deal with data uncertainties and ill-posedness of parameter estimation problems in WDNs. A novel data driven strategy is presented for tuning the regularisation hyper-parameter for the inverse problem and also for validating the results on an independent set of operational hydraulic data. A limited-memory quasi-Newton method (L-BFGS-B) is implemented for solving the resulting regularised nonlinear inverse problem. Furthermore, the implemented method utilises either the Darcy-Weisbach or Hazen-Williams head loss formulae, and is investigated both with and without pipe grouping. An extensive experimental programme was carried out to acquire unique hydraulic data from an operational WDN in order to investigate the robustness of the proposed parameter estimation method. The hydraulic model of the operational 1 Waldron, January 30, 2020 WDN and the acquired hydraulic data are provided as supplementary data to enable the comparison of hydraulic model calibration methods with operational data and encourage reproducible research.

Optimization-Based Selection of Hydrants and Valves Control in Water Distribution Networks for Fire Incidents Management

“…We apply the proposed optimization-based heuristic on the BWFL network shown in Figure 4. The BWFL network is an operational water distribution network within the city of Bristol (UK) and it has been used as a case study in multiple published works: [18], [3], [31], [32], [33], [34], [35]. The network is comprised of 2371 links and 2310 nodes.…”

Adaptive Model Predictive Control for Fire Incidents in Water Distribution Networks

“…The tank constraints (26) and the energy balance constraints (27) are the same as with NC. Variable bounds are set by constraints ( 28)- (34).…”

“…The BWFL network is an operational water supply network, which is part of the water distribution network of the city of Bristol in the United Kingdom, and it serves approximately 8,000 customers. It has been used as a case study in multiple published works [39,38,25,1,22,34]. The BWFL network, shown in Figure 4, is comprised of 2371 links, 2310 nodes, 2 sources of fixed head, 2 (fixed-speed) pumps, two tanks, three pressure reducing valves (PRVs) and two boundary control valves (BVs).…”

Optimal design-for-control of self-cleaning water distribution networks using a convex multi-start algorithm

“…It currently operates with dynamically controlled PRV and bidirectional DBV valves, which modulate flow and pressure between adjacent zones to minimize AZP. BWFLnet's current control valve configuration was applied to model existing conditions -for more details on BWFLnet see Wright et al (2014) and Waldron et al (2020). Moreover, existing kept-shut boundary valves were opened in order to test the full capabilities of the proposed solution algorithm.…”