Application of neural networks to predict the transient performance of a Run-Around Membrane Energy Exchanger for yearly non-stop operation

Application of soft computing methods (i.e. neural networks and genetic algorithms) for modeling and controlling the dynamic and transient behavior of systems has been increasing during the last decade. In this study, a neural network (NN) model is developed to predict the transient heat and moisture transfer performances (i.e., the sensible and latent effectivenesses) of a novel HVAC energy exchanger, called the Run-Around Membrane Energy Exchanger (RAMEE), which is able to transfer both heat and moisture between exhaust and supply air streams. The training data set for the NN model covers a wide range of outdoor conditions and system parameters and is produced using a Transient Numerical Model (TNM) that has been experimentally validated for some transient applications. Two separate NNs (one for sensible and one for latent energy transfer) each with 12 inputs and one output, are selected to represent the RAMEE. The ability of NN models to predict the performance of a given RAMEE design in different climates is numerically validated. The mean absolute difference (MAD) between the results of TNM and NN models for different locations are 0.5 °C for the sensible model and 0.2 gv/kga for the latent model, which indicates satisfactory agreement for energy exchange calculations. These NN models are very fast and easy to use therefore, they might be used for design purposes or estimating the annual energy savings in different buildings with continuous operation and a RAMEE in their HVAC system.