EXPERIMENTAL INVESTIGATION AND NEURAL NETWORK PREDICTION OF THE PERFORMANCE OF A MIXED MODE SOLAR DRYER FOR COCONUT

Abstract

The shelf life of agricultural food products may be enhanced by reducing their moisture contents, by means of a drying process. The present work aims at drying coconut yielding copra. This paper presents the design, analysis of a mixed mode solar dryer for food preservation and energy saving. In the mixed mode solar dryer, the drying cabinet absorbs solar energy directly through the transparent roof and during the same time the heated air from a solar collector is passed through a tray. Various measurements like solar radiation, mass flow rate, and moisture content and relative humidity have been observed. From previous literature four different models (Newton, Page, Henderson & Pabis and Wang & Singh) are chosen for testing the performance of mixed mode solar dryer. Selected models are evaluated by using EMD, ERMS, R2 and 𝜒2 and it is concluded that page model is more suitable for the fabricated cabinet solar dryer at air flow rate 0.009Kg/s based on the experimental analysis. The direct radiant solar energy and a convective hot air stream dry the products, resulting in longer life for the products which are also free from impurities. The experimental results are utilized to evolve a suitable mathematical model, among the different models that are chosen, for copra. This will help in designing suitable dryers for actual users. Also, a multilayer neural network approach has been used to predict the performance of a mixed mode solar dryer for drying coconut. The simulation of neural network is based on the feed forward back propagation algorithm.