Conference Proceedings

A Machine Learning Approach for the Performance Prediction of GCHPs with Horizontal Ground Heat Exchangers

Yu Zhou, Guillermo Narsilio, Nicolas Makasis, Lu Aye, NP LopezAcosta (ed.), E MartinezHernandez (ed.), AL EspinosaSantiago (ed.), JA MendozaPromotor (ed.), AO Lopez (ed.)

GEOTECHNICAL ENGINEERING IN THE XXI CENTURY: LESSONS LEARNED AND FUTURE CHALLENGES | IOS PRESS | Published : 2019

Abstract

This study aims to provide a machine learning approach to predict the performance of Ground Coupled Heat Pumps (GCHPs) with horizontal Ground Heat Exchangers (GHEs). Specifically, an ANN model was developed for this purpose which can potentially be generally applied to similar sites at different locations and climate conditions, with even limited types of input data. In this example, a TRNSYS model regarding a typical horizontal trench within a rural farm in Australia, has been developed and verified, covering over 50 different yearly loading patterns under 3 different climate conditions. The simulated performance data is then used to train the artificial neural network. As results, the trai..

View full abstract

University of Melbourne Researchers