Performance prediction of porous bed solar air heater using MLP and GRNN model- A comparative study

Abstract

In present work, two different types of neural model have been used to predict the thermal performance of unidirectional flow porous bed solar air heater (SAH). These models are multi-layer perceptron (MLP) and generalized regression neural network (GRNN). Total 96 data were used in neural model. The neural model developed with six input parameters: mass flow rate, wind speed, ambient temperature, inlet air temperature, air mean temperature and solar intensity, thermal efficiency is used as output variable. In MLP model, LM with 13 neurons was optimal model and in case of GRNN model, maximum accuracy in prediction has been obtained at spread value- 0.8.  The comparative analysis shows that the GRNN is the best model as compared to MLP due to less error and highest value of R2. These results show that the GRNN model is appropriate model for predicting the thermal performance SAH.