Development and evaluation of an adaptive neuro fuzzy interface models to predict performance of a solar dryer

Abstract

 This research is carried out to predict energy efficiency of a solar dryer by adaptive neuro-fuzzy inference system (ANFIS) model. In this model, temperatures in the collector inlet, collector outlet and in the dry chamber exit and also absorbed heat energy by collector and necessary energy for evaporation of product moisture were considered as an ANFIS network inputs. To investigate the capability of ANFIS models in prediction of dryer efficiency, empirical model and regression analysis were used and their results were compared by ANFIS models. To evaluate an accuracy ANFIS models, statistical parameters such as mean absolute error, mean squared error, sum squared error, correlation coefficient (R) and probability (P) were calculated. Results indicated that coefficient of determination for ANFIS model was higher than empirical model and regression analysis whereas amounts of SSE and MSE were lower. From the results of this research, it is concluded that ANFIS model represent energy efficiency better than empirical model and regression analysis. Finally, it can be stated that the ANFIS model could be efficient in to determining the energy efficiency in a forced-convection solar dryer.