Bibliographic citations
Pinedo, J., Grados, F. (2023). Modelo predictivo de planificación operativa bajo el enfoque de machine learning para gestionar los tiempos de ciclo del volquete en una mina subterránea [Trabajo de suficiencia profesional, Universidad Peruana de Ciencias Aplicadas (UPC)]. http://hdl.handle.net/10757/670723
Pinedo, J., Grados, F. Modelo predictivo de planificación operativa bajo el enfoque de machine learning para gestionar los tiempos de ciclo del volquete en una mina subterránea [Trabajo de suficiencia profesional]. PE: Universidad Peruana de Ciencias Aplicadas (UPC); 2023. http://hdl.handle.net/10757/670723
@misc{renati/410784,
title = "Modelo predictivo de planificación operativa bajo el enfoque de machine learning para gestionar los tiempos de ciclo del volquete en una mina subterránea",
author = "Grados Acosta, Francisco Jesus",
publisher = "Universidad Peruana de Ciencias Aplicadas (UPC)",
year = "2023"
}
The haulage in subway mines usually involves a high operational cost, so new methodologies that contribute to the data management of the dumper are needed to optimize the haulage cycle. In this sense, based on the treatment of data from the fleet management system of the subway mine located in central Peru, the machine learning automation technique was applied to estimate the dumper cycle times. It is important to mention that an inventory of 590,088 data comprising date, shift, loading time, unloading time, dump truck cycle time, polygon, bench, material, loading equipment, haulage equipment, unloading location, trips, haulage/return distance were used for the prediction. This information was taken from March through July of this year. Likewise, within the selection process of the type of machine learning algorithm to be used, the backpropagation (BP) neural network was selected, which is composed of an input layer with 17 neurons, a hidden layer with 100 neurons and an output layer with 1 neuron. In addition, the BP neural network programming was carried out using 80% of the values for training, 10% for predictive model validation and finally 10% of the total data set for testing. The predictive model developed had an RMSE value of 0.26 in its validation, which means that the model has a good performance and predictive capacity. Therefore, the results obtained in the subsequent test can be used by the operations area to map routes more efficiently. In that sense, the route starting from Pit 590, located at sublevel 4055 towards the unloading location called “Cancha Sur Marginal 2“ is a critical route, because its actual value of haul cycle time exceeds the predicted value by both the upper and lower limit, this may indicate possible operational problems or challenges that need to be addressed to improve the efficiency and performance of the CV 31 dump truck cycle on the identified route.
This item is licensed under a Creative Commons License