Click for new scientific resources and news about Corona[COVID-19]

Paper Information

Journal:   AGRICULTURAL RESEARCH   2007 , Volume 7 , Number 3; Page(s) 141 To 154.
 
Paper: 

INVESTIGATION ON IMPORTANT DRYING INDICES OF GRAPE IN HOT AIR FLOW BY USING ARTIFICIAL NEURAL NETWORKS

 
 
Author(s):  BEHROOZI KHAZAEI N., AMIRI CHAYJAN R., TAVAKOLI HASHJIN R., KHOSHTAGHAZA M.H.
 
* 
 
Abstract: 
In this research, prediction of drying time and quality parameters of final production (raisin) in drying process of grape carried out using Artificial Neural Networks (ANNs).
Input air temperature, input air velocity and pretreatment type of grape are important parameters in grape drying by hot air flow and were selected as independent variables for ANNs. For creating of training and testing patterns, drying experiments were, carried out by a laboratory dryer. Drying time and final product quality at the end of each experiment were obtained and considered as output of ANNs. Several networks as well as Levenberg- Marquardt (LM) training algorithm used for training of patterns. Results showed that the Elman network with topology of 3-6-3 and threshold function of logarithm sigmoid is able to predict the drying time and quality indices (lightness and redness to yellowness) by R2=0.9915, 0.9729 and 0.9934 and absolute error of 1.65, 0.39 and 0.026, respectively.
Noise application to input parameter of training patterns showed that the optimum selected network performance is acceptable, because of producing low training error.
 
Keyword(s): GRAPE, INTELLIGENT PREDICTION, ELMAN NETWORK, DRYING TIME, ANN
 
References: 
  • ندارد
 
  Persian Abstract Yearly Visit 82
 
Latest on Blog
Enter SID Blog