An Investigation with Neural Network of Heat Loss for Optimum Insulation

In this study, two different artificial neural network models were used for insulation and non-insulation of the heating pipes used for heating in buildings and two different artificial neural networks (YSA) models for the insulated and non-insulated states of the building walls. 3-layer forward feed in YSA models designed for these situations and a back-propagation model is preferred. The sigmoid transfer function is used in the hidden layer and the linear transfer function is used in the output layer. Back propagation artificial neural network topology is preferred as YSA model and the data were presented to the network in normalized form. The temperature values obtained from the network are compared with the measured temperature values and the results are very close to one another. In this way, the use of artificial neural network method for estimation of 4 different internal models, definition of models and the prediction power has increased. In the random and periodic time interval, the inner plaster thickness is 2 cm, the outer plaster thickness is 3 cm and according to the wall width of 17 cm, 10 cm thick insulation (xps material insulated) and according to the non-insulated wall parameters The statistical data generated from this table that is not based on a nonlinear formula, ie, YSA, is introduced to the network structure and the results obtained by testing from the YSA model in the Matlab environment after training were compared and values very close to each other were determined. Again, in a random and periodic time interval insulated with 100 mm pipe size (insulated stapler material) and the values obtained from the table according to the uninsulated pipe parameters and the results from the YSA model were compared and compared very close values have been determined.