Neural network analysis of fin-tube refrigerating heat exchanger with limited experimental data

We consider the problem of accuracy in heal late estimations from artificia l neural network (ANN) models of heat exchangers used for refrigeration app lications. Limited experimental measurements from a manufacturer are used t o show the capability of the neural network technique in modeling the heat transfer phenomena in these systems. A well-trained network correlates the data with errors of the same order as the uncertainty of the measurements. It is also shown that the number and distribution of the training data are linked to the performance of the network when estimating the heat rates und er different operating conditions, and that networks trained from few tests may give large errors. A methodology based on the cross-validation techniq ue is presented to find regions where not enough data are available to cons truct a reliable neural network. The results from three tests show that the proposed methodology gives an upper bound of the estimated error in the he at rates. The procedure outlined here can also help the manufacturer to fin d where new measurements are needed. (C) 2001 Elsevier Science Ltd. All rig hts reserved.