Citas bibligráficas
García, G., (2021). Sistema de monitoreo y predicción del desgaste de los revestimientos de un molino de bolas haciendo uso de un gemelo digital [Tesis, Universidad de Ingeniería y Tecnología]. https://hdl.handle.net/20.500.12815/294
García, G., Sistema de monitoreo y predicción del desgaste de los revestimientos de un molino de bolas haciendo uso de un gemelo digital [Tesis]. PE: Universidad de Ingeniería y Tecnología; 2021. https://hdl.handle.net/20.500.12815/294
@misc{renati/230563,
title = "Sistema de monitoreo y predicción del desgaste de los revestimientos de un molino de bolas haciendo uso de un gemelo digital",
author = "García Adrianzén, Grimaldo Ernesto",
publisher = "Universidad de Ingeniería y Tecnología",
year = "2021"
}
Currently, the challenge in mining is linked to reduce the costs of its various processes involved in the value chain; in this sense, predicting the failures of different equipment, for example the liners of a ball mill, involved in the processing of different minerals can be very beneficial; on the one hand, reducing costs, making the most of the investment that is given to this equipment; on the other hand, reducing or minimizing emergency stops, which causes production losses and thus monetary losses. Current control systems allow making projections of when the equipment could be changed before failure; however, they are not very accurate and are inefficient, since there are many factors that intervene in the life of the equipment; for example, the mechanical properties of the rocks that enter the mill or the types of materials that are manufactured. This thesis will model a control system that uses the wear database of a ball mill to develop a predictive neural network type algorithm, which has the ability to learn and allow the establishment of an output function that best fits the wear data evaluated. In addition, the Python programming language will be used, with the help of different AI libraries such as Tensorflow, Keras and data analysis libraries and mathematical functions such as Numpy and Pandas. Finally, a Digital Twin of the wear profiles will be modeled, which will allow to see and evaluate the behavior of the coatings over time and will allow comparisons with traditional monitoring systems; likewise, the time and money savings that the implementation of this system would provide will be evaluated.
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons