作者: Hikmet Esen , Mustafa Inalli
DOI: 10.1016/J.ESWA.2009.01.055
关键词: Materials science 、 Heat exchanger 、 Conjugate gradient method 、 Control theory 、 Heat pump 、 System identification 、 Artificial neural network 、 Sigmoid transfer function 、 Systems modeling 、 Tangent
摘要: This paper describes the applicability of artificial neural networks (ANNs) to estimate performance a vertical ground coupled heat pump (VGCHP) system used for cooling and heating purposes experimentally. The involved three exchangers in different depths at 30 (VB1), 60 (VB2) 90 (VB3)m. experimental results were obtained seasons 2006-2007. ANNs have been varied applications they shown be particularly useful modeling identification. In this study, back-propagation learning algorithm with variants, namely Levenberg-Marguardt (LM), Pola-Ribiere conjugate gradient (CGP), scaled (SCG), tangent sigmoid transfer function network so that best approach could found. most suitable neuron number hidden layer found as LM 8 neurons both modes.