Determination of Output Parameters of a Thermoelectric Module using Artificial Neural Networks


Ciylan B.

ELEKTRONIKA IR ELEKTROTECHNIKA, no.10, pp.63-66, 2011 (SCI-Expanded) identifier identifier

Abstract

Determination of instant dynamic output parameters of thermoelectric module which is worked in any system is very important. Despite of the new methods this process takes a lot of times. In this study, two artificial neural network (ANN) models are designed for the estimation of dynamic output parameters at any desired moment of the thermoelectric modules. MATLAB-ANN tools and an ANN simulator program are used for creating the models. Experimental dynamic output parameters data which obtained from eight different thermal load conditions were used for training the ANN Models. On the designed ANN models which were created to estimate instant dynamic output parameters of the thermoelectric module, the Levenberg-Marquardt (LM) learning algorithm has been used. The results obtained with these ANN models, compared with the experimental data and it was shown in graphs