Method for detecting pipeline leaks using neural networks and dynamic transient modeling to improve the integrity of water supply systems

Summary

A method for detecting pipeline leaks using neural networks and dynamic transient modeling to improve the integrity of water supply systems is presented. The leak detection and localization method combines an approach based on real-time transient models with artificial neural networks. The method provides for combining the analytical data available before the system startup with real-time data represented by measured process signals. Real-time data can be presented before and during the system operation providing for the flexible information processing.

Key words

water supply system , leak monitoring , transient modeling , artificial neural networks , data base