Neural network-based modeling of a thermal power plant feedwater pump

Abstract

Obtaining an accurate model of a real-world system using linear systems theory can prove to be a complex task due to the nonlinear characteristics that systems exhibit. Neural networks have the ability to reproduce the complex nonlinear relations which makes them a useful tool in system identification and modeling. The purpose of this paper is to obtain the model of a thermal power plant feedwater pump in order to test various control approaches. The neural network used in this paper is a multi-layer feed-forward network. The comparison of the results obtained by using this approach with the results obtained from a mathematical model confirms that the neural network-based model is a better approximation of the observed system.