Bibliographic citations
Medrano, B., (2020). Desarrollo del modelo neuronal para la mejora de la eficiencia de una línea de conversión de rollos de papel [Trabajo de suficiencia profesional, Universidad Nacional de Ingeniería]. http://hdl.handle.net/20.500.14076/21221
Medrano, B., Desarrollo del modelo neuronal para la mejora de la eficiencia de una línea de conversión de rollos de papel [Trabajo de suficiencia profesional]. PE: Universidad Nacional de Ingeniería; 2020. http://hdl.handle.net/20.500.14076/21221
@misc{renati/711107,
title = "Desarrollo del modelo neuronal para la mejora de la eficiencia de una línea de conversión de rollos de papel",
author = "Medrano Cochachi, Brayan Anthony",
publisher = "Universidad Nacional de Ingeniería",
year = "2020"
}
The present research work is based on the study and application of recurrent neural networks in predicting the production of a paper roll conversion line in a paper company to improve efficiency. Biological systems are the basis of all kinds of neural networks, since they have interconnections between all the neurons of which they are part. This characteristic is reflected in the behavior of the recurrent neural networks; which have the ability, due to their type of processing, to predict values based on past information. The volume of paper roll production is the variable to predict because it is the result of control variables such as cleaning time, change of base paper, change of format, scheduled, routine stops, quality defects, etc. The variables are the inputs and outputs of the red neuronal. Once the reliability of the prediction is obtained, new values of production line stop times are entered to predict the increase in production of the paper roll converting line. The simulation of the predictive system is carried out in KERAS, a high-level Python library, and the engine that trains the neural network is the Google implementation, called TENSORFLOW. This provides the percentage of error of the production predictions and the real ones, which is minimal and tends towards the value of zero. As this is the case, in addition to corroborating the reliability of the prediction, production is increased, the efficiency of the line conversion and there is a positive impact on the economic income of the paper company.
This item is licensed under a Creative Commons License