Neural Network predictive process modeling: Application to food processing

Mesfin Agide
Currently, food processing industry is driven by several requirements. This requirement includes ensuring safety, meeting quality standard and customer expectation and reducing production cost to be competent in market. To achieve this requirement they have to operate at optimum process conditions all the time. In food processing, due to the nature of the process, it is difficult to find and operate at the best conditions solely by experience. The Ethiopia food industry is no coping up with
more » ... requirement due cost of optimization and low level of education of works operating in the production system. Thus, it is necessary modeling of the process or part of the process to capture the relation of between important process parameters and use the model to control and improve the process better. In addition, it is found necessary to make the model accessible for the operators working in Ethiopian industry. Using artificial neural network method is found to be very good modeling to tool to solve food engineering problems. In this thesis, therefore, artificial neural network method is used to model and tested for selected food industry engineering problems, specifically, water activity prediction, predictive food microbiology and control chart pattern recognition. The model is enclosed in interactive software so that it could also be used by people that do not have sophisticated mathematical and technical skills. The result obtained for all problems shows that neural network modeling can be used to model food process and to predict food process parameters with sufficient accuracy.
doi:10.20372/nadre/5275 fatcat:dfzhwz7ernd45nz2ppn7axrd7u