Application of Kriging Algorithms for Solving some Water Nets Management Tasks

Abstract

The kriging algorithms have been developed by Krige (1951), Sichel (1952) and Matheron (1962) as some methods of spatial statistics and in the beginning they were intended for applications in the mining for identification of raw material resources. Their essential advantages are simplicity, efficiency and reliability and also a big robustness on the measurement disturbances. The kriging algorithms are used to make the 2D and 3D approximation of the variables investigated and their main mathematical apparatus is a method of static optimization applied to model a discreet curve won from the measurements data. In the System Research Institute (IBS PAN) some successful studies have been made while using these algorithms to approximate some environmental data like the rain and snow falls, the soil composition and air temperature (Bogdan, Studzinski, 2006A,B,C,D, 2007). Parallel with these calculations an information system to complex management of communal water networks have been developed in IBS PAN. It consists of three standard programs, i.e. of a GIS (Geographical Information System) system, SCADA (System of Control And Diagnostics Analysis) system and a CIS (Customers Information System) system and of several mathematical models of the water net objects and processes. The main idea of this information system is to solve the tasks connected with the water net operation with the help of these different mathematical models (Studzinski, 2006, 2007). While developing the system and independently the kriging algorithms the idea arose to use this kind of approximation to help the solution of some problems of the water net management. The first approach was to calculate with the kriging algorithms the distribution of the water net flows and pressures (Bogdan, Studzinski, 2007). After the successful results have been obtained the application of the kriging approximation in the waterworks branch has been extended and some new findings in this field are reported in this paper.