Development of Modular Neural Networks with Fuzzy Logic Response Integration for Monitoring the Quince Osmotic Dehydration-Part Ⅱ Prediction
This research was conducted to forecast osmotic dehydration characteristics of quince by modular neural network with fuzzy logic response integration (MNFLRI).Solid gain, water loss and moisture content were implemented as outputs, whereas drying time and classified images were put as inputs.The minimum ％MRE (18.153) for water loss (WL) was obtained when applying two hidden layers with 6 neurons per each two layers.As well as, the lowest ％MRE (35.5335) for solid gain (SG) was acquired when utilizing 6 and 8 neurons per first and second layer, correspondingly.It can be stated that ％MRE was at least (7.4759) for moisture content (MC) by 6 and 5 neurons per first and second layer, respectively.This innovative technique could be successfully applied for quantitative monitoring of quince quality changes undergoing osmotic dehydration process.
neural networks quince fuzzy integration osmotic dehydration prediction
M.Shafafi Zenoozian M.Irani H.Irani
Department of Food Science and Technology,Sabzevar Branch ,Islamic AzadUniversity,Sabzevar,Iran Department of Computer Engineering,Islamic Azad University,Mashad Branch,Mashad,Iran
国际会议
The 7th Asia-Pacific Drying Conference(第七届亚太地区干燥会议 ADC2011)
天津
英文
1-8
2011-09-18（万方平台首次上网日期，不代表论文的发表时间）