Tuesday , August 14 2018

Leakage Detection and Localisation in Drinking Water Distribution Networks by MultiRegional PCA

Kazimierz DUZINKIEWICZ1, Adam BOROWA1, Krzysztof MAZUR1, Michal GROCHOWSKI1, Mietek A. BRDYS2, Krzysztof JEZIOR1
1 Faculty of Electrical and Control Engineering,
Gdansk University of Technology
ul. Narutowicza 11/12, 80-952 Gdansk, Poland
{k.duzinkiewicz; aborowa; k.mazur; m.grochowski; m.brdysl; k.jezior}@ely.pg.gda.pl
2 Department of Electronic, Electrical and Computer Engineering,

University of Birmingham, Birmingham B15 2TT, U.K.
m.brdys@bham.ac.uk

Abstract: Monitoring is one of the most important steps in advanced control of complex dynamic systems. Precise information about systems behaviour, including faults indicating, enables for efficient control. The paper describes an approach to detection and localisation of pipe leakage in Drinking Water Distribution Systems (DWDS) representing complex and distributed dynamic system of large scale. Proposed MultiRegional Principal Component Analysis (MR-PCA) skilfully takes full advantage of well known PCA method and enables not only for detecting the leakages but also supports their localisation. The main idea of MR-PCA is presented on example of small water network. Next the method is applied to DWDS in Chojnice, northern Poland. DWDS Chojnice is decomposed into suitable subnetworks what makes that the monitoring process is easier and require less sensors. The subnetworks and corresponding PCA monitoring models are selected based on the network operational knowledge and information regarding its topology.

Keywords: Monitoring, large-scale systems, network systems, fault detection algorithms, water leakage detection, statistical methods.

>>Full text
CITE THIS PAPER AS:
Kazimierz DUZINKIEWICZ, Adam BOROWA, Krzysztof MAZUR, Michal GROCHOWSKI, Mietek A. BRDYS, Krzysztof JEZIOR, Leakage Detection and Localisation in Drinking Water Distribution Networks by MultiRegional PCA, Studies in Informatics and Control, ISSN 1220-1766, vol. 17 (2), pp. 135-152, 2008.

1. Introduction

Nowadays, monitoring systems besides data gathering are able to pre-process the data, to recover and estimate not directly measured variables. However, in large scale systems there is very large quantity of information that are hard to handle and sometimes almost impossible to properly process and hence to efficiently utilised it in the control process. An example of such systems is Drinking Water Distribution Systems (DWDS) the representatives of the class of network systems. The DWDS are usually, very complex (lots of pipes, connecting nodes, pumps, tanks etc.) and distributed (in space). It entails measuring of very large number of variables necessity, in order to possess information about the system state that is necessary for efficient system control. In such situations special methods enabling for analysis of large amount of data (e.g. faults detection and isolation) are required. Advanced monitoring systems should not only visualize desired data but also be able to detect devices faults and/or the unusual system behaviour. The paper proposes an approach to detecting and localisation of water leakage in pipes by using the Principal Components Analysis (PCA) method [1]. The PCA is a method that looks for multidimensional correlation between the variables and uses it to reduce the dimensionality of problems simultaneously remaining most of original information. Mostly, large amount of real data process do not provide large amount of important information. Hence, PCA explores data to find out very meaningful ones and include them into statistical models. Moreover, these models clearly indicate the abnormal state of the system thanks to specially calculated measures (T2 and SPE). In case of DWDS such a situation might be caused by device faults (e.g. sensor or pump break down), water leakage in pipe, significant increasing of the water uptake (e.g. caused by fire brigades) etc. Detecting of fault is important however, in case of DWDS the system operator still does not know its type and localisation. Leakages detection and localisation issue is a very important and complex problem that has been widely investigated [2] – [6]. However, available active leakage control methods are basically unpractical due to costs or long leak detection and location time [4]. In the paper the novel approach the MultiRegional Principal Component Analysis (MR-PCA) method is used to detect and to locate the water leakage based on measurements from limited number of measuring devices [5]. MR-PCA tries to join operational experience of staff working in water companies and advanced mathematical analysis. Moreover, this method compromises between detection efficiency and a number of measuring devices.

The method is explained based on simple water network and followed with its application to real town case study DWDS Chojnice (northern Poland).

REFERENCES

  1. Jackson, J.E., A User’s Guide to Principal Components, Wiley-Interscience (New York), 1991.
  2. Vítkovský, J.P., Simpson, A.R. and Lambert, M.F., Leak Detection and Calibration Using Transients and Genetic Algorithms, J. Water Resour. Plng. and Mgmt 126, 2000.
  3. Poulakis, Z., Valougeorgis, D. and Papadimitriou, C., Leakage Detection in Water Pipe Networks Using a Bayesian Probabilistic Framework, Probabilistic Engineering Mechanics 18, 2003.
  4. Misiunas, D., Failure Monitoring and Asset Condition Assessment in Water Supply Systems. Ph.D thesis, Department of Industrial Electrical Engineering and Automation, Lund University (Lund), 2005.
  1. Borowa, A., Mazur, K., Grochowski, M., Brdyś, M.A. and Jezior, K., MultiRegional PCA for Leakage Detection and Localisation in DWDS – Approach, (Eds.) Korbicz J., Patan K. and Kowal M., Fault Diagnosis and Fault Tolerant Control, Academic Publishing House EXIT (Warsaw), 2007.
  2. Jezior, K., Mazur, K., Borowa, A., Grochowski, M. and Brdys, M.A., MultiRegional PCA for Leakage Detection and Localisation in DWDS – Chojnice Case Study, (Eds.) Korbicz J., Patan K. and Kowal M., Fault Diagnosis and Fault Tolerant Control, Academic Publishing House EXIT (Warsaw), 2007.
  3. Venkatasubramanian, V., Rengaswamy, R., Kavuri, S.N. and Yin, K., A Review of Process Fault Detection and Diagnosis Part III: Process History Based Methods, Computers and Chemical Engineering 27, 2003.
  4. Zhang, X., Parisini, T. and Polycarpou, M., Adaptive Fault-tolerant Control of Nonlinear Uncertain Systems an Information Based Diagnostic Approach, IEEE Transactions on Automatic Control 49, August 2004.
  5. Brdys, M.A., Grochowski, M., Gminski, T., Konarczak, K. and Drewa, M., Hierarchical Predictive Control of Integrated Wastewater Treatment Systems, Control Engineering Practice 16, 2008.
  6. Venkatasubramanian, V., Rengaswamy, R., Kavuri, S.N. and Yin, K., A Review of Process Fault Detection and Diagnosis Part I: Quantitative Model-based Methods, Computers and Chemical Engineering 27, 2003.
  7. Venkatasubramanian, V., Rengaswamy, R. and Kavuri, S.N., A Review of Process Fault Detection and Diagnosis Part II: Qualitative Models and Search Strategies, Computers and Chemical Engineering 27, 2003.
  8. Li, W., Yue, H.H., Valle-Cervantes, S. and Qin, S.J., Recursive PCA for Adaptive Process Monitoring, J. Process Control 10, 2000.
  9. Jackson, J.E. and Mudholkar, G., Control Procedures for Residuals Associated with Principal Component Analysis, Technometrics 21, 1979.
  10. Rosen, C. and Lennox, J. A., Multivariate and Multiscale Monitoring of Wastewater Treatment Operation, Wat. Res. 35, 2001.
  11. Lennox, J. A. and Rosen, C., Adaptive Multiscale Principal Components Analysis for Online Monitoring of Wastewater Treatment, Water Science Technology 45, 2002.
  12. Borowa, A., Mazur, K. and Brdyś, M.A., Identyfikacja i diagnostyka procesu oczyszczania ścieków przy użyciu metody MS-PCA, PAK 9 (BIS), 2005. (in Polish)
  13. Mazur, K., Borowa, A. and Brdyś, M.A., Diagnostyka i identyfikacja procesu przy użyciu AdMS-PCA, PAK 9 (BIS), 2005. (in Polish)
  14. Mazur, K., Borowa, A. and Brdys, M.A., Condition Monitoring Using PCA Based Method and Application to Watewater Treatment Plant Operation, Proc. 1st IFAC Workshop on Applications of Large Scale Industrial Systems, ALSIS´06, Helsinki-Stockholm cruise liner, 2006.
  15. Borowa, A., Brdys, M.A. and Mazur, K., Detection of Unmeasured Process Abnormalities in Wastewater Treatment Process Using MS-PCA, Proc. 11th IFAC/IFORS/IMACS/IFIP Symposium on Large Scale Systems Theory and Applications, LSS’07, Gdansk, Poland, 2007.
  16. Qiu, B., Prinet, V., Perrier, E. and Monga, O., Multi-Block PCA Method for Image Change Detection, Proc. 12th International Conference on Image Analysis and Processing, ICIAP’03, Mantova, Italy, 2003.
  17. Westerhuis, J.A., Kourti, T. and MacGregor, J.F., Analysis of multiblock and hierarchical PCA and PLS models, Journal of Chemometrics 12, 1998.
  18. Brdys, M.A., Duzinkiewicz, K., Chang, T., Polycarpou, M., Wang, Z., Uber, J., Propato, M., Hierarchical control of integrated quality and quantity in drinking water distribution systems, Proc. 1st IFAC International Conference on Technology, Automation and Control of Wastewater and Drinking Water Systems, TiASWiK’02, Gdansk Sobieszewo, Poland, 2002.
  19. Duzinkiewicz, K., Integrated Control of Drinking Water Supply and Distribution Systems. D. Sc. thesis, AGH University of Science and Technology Publishers, 2005. (in Polish)
  20. Brdys, M.A., Duzinkiewicz, K., Grochowski, M, Rutkowski, T., Robust Estimation of Integrated Hydraulics and Parameters in Water Distribution Systems, Proc. 4th ASCE Annual Water Distribution Systems Analysis, World Water and Environmental Resources Congress, Orlando, 2001.
  21. Brdys, M.A. and Ulanicki, B., Operational Control of Water Systems: Structures, Algorithms and Applications, Prentice Hall (New York, London, Toronto, Sydney, Tokyo), 1994.
  22. Mays, L.W., Water Distribution Systems Handbook, McGraw-Hill (New York), 2000.
  23. Rossman, L. A., EPANET 2 users manual, United States Environment Protection Agency, 2000.
  24. Poulakis, Z., Valougeorgis, D. and Papadimitriou, C., Leakage Detection in Water Pipe Networks Using a Bayesian Probabilistic Framework, Probabilistic Engineering Mechanics 18, 2003.
  25. Ewald, G., Kurek, W. and Brdys, M.A., Grid Implementation of Parallel Multiobjective Genetic Algorithm for Optimized Allocation of Chlorination Stations in Drinking Water Distribution Systems: Chojnice case study, IEEE Transactions on System, Man, and Cybernetics-Part C: Applications and Reviews (submitted in 2007).
  26. Zierolf, M., Polycarpou, M. and Uber, J., Development and Autocalibration of an Input-Output Model of Chlorine Transport in Drinking Water Distribution Systems, IEEE Transaction on Control Systems Technology 6, 1998.
  27. Dunia, R. and Qin, S.J., Joint Diagnosis of Process and Sensor Faults Using Principal Component Analysis, Control Engineering Practice 6, 1998.