American Society of Civil Engineers


Robust Estimation of Integrated Hydraulics and Parameters in Water Distribution Systems


by Mietek A. Brdys, (School of Electronic and Electrical Engineering, The University of Birmingham, Edgbaston Birmingham B15 2TT, UK; Technical University of Gdansk, Electrical and Control Engineering Faculty, Control Engineering Department, Narutowicza 11/12, 80-952 Gdansk, Poland), Kazimierz Duzinkiewicz, (Technical University of Gdansk, Electrical and Control Engineering Faculty, Control Engineering Department, Narutowicza 11/12, 80-952 Gdansk, Poland), Michal Grochowski, (Technical University of Gdansk, Electrical and Control Engineering Faculty, Control Engineering Department, Narutowicza 11/12, 80-952 Gdansk, Poland), and Tomasz Rutkowski, (Technical University of Gdansk, Electrical and Control Engineering Faculty, Control Engineering Department, Narutowicza 11/12, 80-952 Gdansk, Poland)
Section: Water Distribution Systems Analysis, pp. 1-12, (doi:  http://dx.doi.org/10.1061/40569(2001)405)

     Access full text
     Purchase Subscription
     Permissions for Reuse  

Document type: Conference Proceeding Paper
Part of: Bridging the Gap: Meeting the World’s Water and Environmental Resources Challenges
Abstract: The paper applies and further develops an algorithm for a recursive robust integrated estimation of flows, nodal hydraulic heads, tank levels and model parameters in drinking-water distribution systems. The algorithm was proposed by Brdys. The parameters and variables are estimated simultaneously. Robustness of the estimates is achieved through non-probabilistic set-bounded modeling of uncertainty in the measurement and modeling errors. The nonlinear non-convex estimation problem is transformed into sparse mixed integer linear programming problem by piecewise linearization in order to guarantee the global optima can be efficiently achieved. The uniform approach to piecewise linearization in two and three dimensional spaces is proposed. It is well known from an existing experience that the algorithm efficiency and quality of estimates heavily depend on: choice of points for the hydraulic nonlinear model piecewise linearisation and length of moving information windows. The paper addresses these issues by using theoretical and simulation based techniques. The results are illustrated by simulation study of physical drinking-water distribution system.


ASCE Subject Headings:
Head (fluid mechanics)
Integrated systems
Parameters
Pipe flow
Water distribution systems
Water tanks