Different methods for modeling absorption heat transformer powered by solar pond

Abstract

Solar ponds are a type of solar collector used for storing solar energy at temperature below 90°C. Absorption heat transformers (AHTs) are devices used to increase the temperature of moderately warm fluid to a more useful temperature level. In this study, a theoretical modelling of an absorption heat transformer for the temperature range obtained from an experimental solar pond with dimensions 3.5 × 3.5 × 2 m is presented. The working fluid pair in the absorption heat transformer is aqueous ternary hydroxide fluid consisting of sodium, potassium and caesium hydroxides in the proportions 40:36:24 (NaOH:KOH:CsOH). Different methods such as linear regression (LR), pace regression (PR), sequential minimal optimization (SMO), M5 model tree, M5′ rules, decision table and back propagation neural network (BPNN) are used for modelling the absorption heat transformer. The best results were obtained by the back propagation neural network model. A new formulation based on the BPNN is presented to determine the flow ratio (FR) and the coefficient of performance (COP) of the absorption heat transformer. The BPNN procedure is more accurate and requires significantly less computation time than the other methods.