Saved in:
| Main Author: | |
|---|---|
| Format: | Artículo científico |
| Language: | en |
| Published: |
Sociedad Mexicana de Nutrición y Tecnología de Alimentos
2001
|
| Subjects: | |
| Online Access: | https://www.redalyc.org/articulo.oa?id=72430102 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- Aplicación de redes neuronales artificiales en la modelización del tratamiento térmico de alimentos W. E. Luera Peña L. A. Minim Química modelling neural networks JB backpropagation thermal treatments In this work, the modelling technique by artificial neural networks was used to predict the temperature profile in the center of a product packed in cans during the thermal processing, as a function of the retort temperature, initial temperature of the productand processing time. A recurrent-type network, comprised of five nodes in the input layer, 10 nodes in the hidden layer and one node in the output layer, was developed. The algorithm JB-backpropagation was used in the learning process. Data used in the training, validation and test phases were obtained through the computational implementation of a mathematical model resulted from theapplication of the differential balance of energy in cylindrical coordinates holding the food. Initially, the process was analysed using time-temperature data with constant autoclave temperature. Later, time-temperature data with varying autoclave temperature were utilized, thus simulating a more realistic process. In both situations, the neural network technique showed good performance in the prediction of the temperature profile in the center of the product, indicating that it is a feasible alternative of modelling in time-varying processes. 2001 artículo científico 1135-8122 https://www.redalyc.org/articulo.oa?id=72430102 en http://www.redalyc.org/revista.oa?id=724 Ciencia y Tecnología Alimentaria application/pdf Sociedad Mexicana de Nutrición y Tecnología de Alimentos Ciencia y Tecnología Alimentaria (México) Num.2 Vol.3