ABSTRACT
Underground bauxite mining exploitations is a challenging environment for ventilation. A controlled underground ventilation system can significantly improve the environmental and working conditions at the mines. In this paper, the modelling of a section of an existing complex underground ventilation network assisted by machine learning (ML) techniques and more particularly by the use of Artificial Neural Network (ANN). The developed ANN is focusing in the prediction of NOx concentration at a selected mine site in order to model its operating characteristics so that they can be automatically adjusted to the existing conditions, ensuring better working conditions and creating a safer and controlled underground environment. The above model can make prediction that are accurate and respond to actual conditions and can be the basis for a further improvement of the Ventilation on Demand (VoD) technology.
