In this work, an artificial neural network (ANN) was used to predict the final temperature of chicken carcasses in an industrial scale. For this purpose, temperatures of chicken carcasses were obtained in a slaughterhouse at the end of the cooling process. The variables considered to influence the temperature of carcasses were average carcass weight, prechiller and chillers velocities, prechiller and chillers average temperatures, the absence or presence of bubbles in prechiller and chiller and bubbles intensity, which were considered as input for the ANN. For training and validation of the feed forward neural network, the above inputs were combined to minimize the weighted sum of the squares of the residues, being tested 10 transfer functions, three training algorithms and two different architectures. The best result was obtained using an ANN composed of two hidden layer (10 nodes in each layer), radial bias as transfer function and gradient descent backpropagation algorithm for training. Using this architecture, the regression coefficient was 0.9265. Even with many variables affecting the industrial cooling process of the chicken carcasses, the ANN developed showed satisfactory fitting of the final temperature of chicken carcasses. This model can be further used for optimization purposes, improving the process quality of chicken carcasses as well as process profitability. PRACTICAL APPLICATIONThe model used in this study can be further used for control and optimization purposes, improving the process quality of chicken carcasses as well as process profitability.bs_bs_banner Journal of Food Process Engineering
In this work, correlation analysis was employed in order to determine how the process variables are associated with the temperature of chicken carcasses after the cooling process by immersion. Based on the correlation analysis, a second‐order polynomial model was proposed to predict the temperature of chicken carcasses given the values for the independent variables. The correlation analysis showed to be an important tool to be applied industrially because it enabled the choice of the variables that truly affect the cooling process of chicken carcasses. In addition, the correlation analysis demonstrated that first‐order interaction and quadratic terms of independent variables also affect the process and should be considered in the model. From these findings, a quadratic model capable of explaining about 45% of the variation of cooling process was proposed. This model can be used as a tool for making quick decisions in the industry or for quickly predicting the behavior of the cooling process given some specific conditions. Practical Applications This work describes a simple and practical procedure for statistical analysis of industrial process data, guiding the development of a quadratic model that can be further used for control and optimization purposes, improving the process quality of chicken carcasses as well as process profitability.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
334 Leonard St
Brooklyn, NY 11211
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.