[Home ] [Archive]   [ فارسی ]  
:: Main :: About :: Current Issue :: Archive :: Search :: Submit :: Contact ::
:: Volume 10, Issue 3 (Autumn 2015) ::
2015, 10(3): 67-77 Back to browse issues page
Prediction of Fundamental Rheological Properties of Dough with Artificial Neural Networks-Genetic Algorithm
H Abbasi . *, MA Mohammadifar
Islamic Azad University, Isfahan (Khorasgan) Branch
Abstract:   (4909 Views)
Background and Objectives: Dough is an intermediate product of bakery production that represents physicochemical properties of raw materials and also qualitative aspects of the final product.Due to the high performance of artificial neural networks in educability and parallel processing of data with non-linear relations, the purpose of present research is providing suitable models for predicting rheological properties of dough based on the physicochemical properties of flour. Materials and Methods: A range of flour samples produced in the factories of various regions and provinces of Iran were collected and seven physicochemical properties of them were evaluated. Oscillatory tests of stain and frequency sweep were done on the dough of samples, and two important parameters of them were selected for modeling. After training, and determining the specifications of networks with genetic algorithm and testing them, the sensitivity of output to input parameters was evaluated. Results: Developed networks are four-layer perceptrons in which the first network, with removing gluten index and particle size index of flour, has 5 neurons in input layer and 15 neurons in the first and second layers for prediction of the incept, and the second has 7 neurons of input layer, 24 neurons in the first hidden layer and 17 neuron in the second hidden layer for predicting the slope of fitted model on frequency sweep. Developed networks predict rheological properties of dough with correlation coefficient more than 97%. Gluten index and zeleny number are introduced as important parameters on changing the rheological properties of dough. Conclusion: Artificial neural network-genetic algorithm is a powerful method in predicting dough's rheological properties. Sensitivity analyses of optimum network indicate in the importance of flour's physicochemical properties in predicting the changes in the fundamental rheological properties of dough. Keywords: Artificial neural network, Genetic Algorithm, Dough rheology, Physicochemical properties of flour
Keywords: Artificial neural network, Genetic Algorithm, Dough rheology, Physicochemical properties of flour
Full-Text [PDF 711 kb]   (1985 Downloads)    
Type of Study: Research | Subject: Food Science
Received: 2015/03/14 | Accepted: 2015/07/26 | Published: 2015/09/26
Send email to the article author

Add your comments about this article
Your username or Email:

CAPTCHA


XML   Persian Abstract   Print


Download citation:
BibTeX | RIS | EndNote | Medlars | ProCite | Reference Manager | RefWorks
Send citation to:

Abbasi H, Mohammadifar M. Prediction of Fundamental Rheological Properties of Dough with Artificial Neural Networks-Genetic Algorithm . Iranian Journal of Nutrition Sciences & Food Technology. 2015; 10 (3) :67-77
URL: http://nsft.sbmu.ac.ir/article-1-1882-en.html


Volume 10, Issue 3 (Autumn 2015) Back to browse issues page
Iranian Journal of  Nutrition Sciences & Food  Technology
Persian site map - English site map - Created in 0.05 seconds with 32 queries by YEKTAWEB 4075