Optimization of district heating system aided by geothermal heat pump: A novel multistage with multilevel ANN modelling

The aim of this study is to obtain the optimum design of geothermal heat pump aided district heating system (GHPDHS) by using a novel ANN model that composed of multistage with multilevel. For this purpose, the acquiring of the best design were performed by utilizing the back-propagation learning algorithm with three different variants which were Levenberg-Marquardt (LM), Pola-Ribiere Conjugate Gradient (CGP), and Scaled Conjugate Gradient (SCG). In this aim, the proposed ANN model was mainly formed from two stages. The first one has a single level whereas the second one composed of three levels in this new ANN model. According to results, the maximum rate of the error occurred in the Pump 2 as % 3.0092 and the minimum of that was obtained from COPsys with % 0.0018. The best R2 value of the third level of the second stage network structure was calculated as 1 for LM-20. As a consequent, this study showed that the multistage with multi-level ANN model could be easily applied to other energy systems in order to save more time and simplicity.