Selection of appropriate variables in the prediction of the coefficient of performance of absorption heat transformers

This research consists of a proposal of three methods that analyze the contribution of each operational variable in the prediction of the coefficient of performance of three different absorption systems aimed at reducing the independent variables for a faster and more efficient prediction. Methods considered in this study are correlation analysis, principal component analysis combined with correlation analysis, and Carson's method. The experimental information includes (i) an absorption heat transformer with duplex components, (ii) an absorption single-state heat transformer, and (iii) a double-absorption heat transformer. For each case, the three methods were applied, the results were discussed in order to find the coincidences and discrepancies.