Neural network learning for blind source separation with application in dam safety monitoring

Abstract

Usually, dam monitoring systems are based on both boundary conditions (temperature, rainfall, water level, etc.) and structural responses. Statistical analysis tools are widely used to determine eventual unwanted behaviors. The main drawback of this approach is that the structural response quantities are related to the external loads using analytical functions, whose parameters do not have physical meaning. In this paper a new approach to solve this problem, based on a neural network learning rule for Blind Source Separation (BSS), to find out the contributions of the dam external loads is presented and applied in a case study for a concrete dam. © 2012 Springer-Verlag.