Neural network-based prediction of effective thermal conductivity of loose multi-phase systems

The use of an artificial neural network (ANN) approach has increased in many areas of engineering problems in the past few years. The objective of this study is to develop an artificial neural network (ANN) model to predict the effective thermal conductivity (ETC) of loose multi-phase systems that is based on experimentally measured variables. A three-layer feedforward ANN structure is constructed and a backpropagation algorithm is used for the training of ANNs. ANN models are based on feedforward backpropagation network with training functions such as: Levenberg-Marquardt (LM), one-step secant (OSS), resilient backpropagation (RP), and scaled conjugate gradient backpropagation (SCG). The training algorithm for neurons and the hidden layers for different feedforward backpropagation networks at the uniform threshold function TANSIG-PURELIN is used and run for 100 epochs. The predicted effective thermal conductivities of loose multi-phase systems are in good agreement with the available experimental data. The study demonstrates the utility and efficiency of the ANN model for estimating the ETC of loose multi-phase systems. (author)