Identification of Comulative Fruit Responses during Storage Using Neural Networks

Abstract

Neural networks are useful to identify complex nonlinear relationships between input and output of a system. Cumulative fruit responses such as water losses and ripening during storage are characterized non-linearly. For identification, several patterns of these cumulative responses, as affectef by environmental factors, are often conducted by repeating the experiment several times under different enviromental conditions. It is not well-known how many response patterns (training data sets) are necessary for an acceptable identifiaction. This research explores an affective way to identify the cumulative responses of tomato during storage using neural networks. Firstly, data for identification were obtained from a mathematical model. Secondly, the relationship between the number of response pattern and the estimation error were investigated. The estimated error becomes smaller when the number of response pattern is three or more. This suggests that three types of response patterns allow cumulative responses to be succesfully identified. Besides, an addition of linear data (1,2,..,N) as input variable significantly improves the identification accuracy of the cumulative response. Finally, the identification of actual was implemented based on these results and satisfactory results will be obtained.