Applications of intelligent techniques in modeling geothermal heat pumps: an updated review

Abstract

Regarding the significance of energy efficiency in cooling and heating systems, it is crucial to apply proper technologies. Heat pumps are among the systems with ability of efficient performance applicable in different operating conditions. These technologies can be coupled with renewable energy sources such as solar and geothermal, which cause lower energy consumption and emission of greenhouse gases. In the present work, studies considered utilization of intelligent techniques in modeling performance of geothermal heat pumps (GHPs) are reviewed. The main findings of the reviewed works reveal that intelligent techniques are able to model heat pumps output with significant and remarkable exactness; for instance, in some cases, R2 of the models proposed that the coefficient of performance of the ground sources heat pumps is around 0.9999, revealing closeness of the predicted data and actual quantities. The precision of the models, based on the intelligent methods, is affected by different elements including the used function, algorithm and architecture. Furthermore, it is observed that using optimization algorithms for tuning the hyperparameters of intelligent techniques cause higher estimation exactness. In addition to performance prediction, some other parameters related to the GHPs such as well temperature and thermal conductivity of the soil layers could be predicted by utilization of intelligent methods.