Application of ANN Artificial Neural Networks to Predict Air Temperature in Longwall Workings of Vietnamese Anthracite Mines

Abstract

To meet Vietnam's growing energy demand, mines are increasing coal extraction, which results in growing depth of exploitation and
establishing new mining levels at greater depths. An increase in the intensity and efficiency of mining can be achieved through efficient
longwall complexes with high electrical power of machines and devices. Climatic conditions in Vietnam, geological conditions of coal
seams, exploitation depth, use of machines with higher electrical power, and mechanization of mining work contribute to the increase in
air temperature in underground mine workings. To ensure the required working conditions for miners, the efficiency of mine workings
ventilation should be increased. Unfortunately, this is not always sufficient to ensure the required conditions, and it is necessary to
use air cooling systems using air conditioning systems. Changes in air temperature in mining excavations are influenced by many
natural, technical, and organizational factors, which are difficult to capture using analytical methods. Therefore, a method based on
the ANN artificial neural network model was proposed for temperature forecasting, which enables forecasting the air temperature in
mechanized and non-mechanized longwall workings. The air temperature forecast results were compared with measurement data. The
analyses show that the actual and forecast data correspond with each other. Therefore, the presented method can be used as a tool for
mining services in the fight against the climate threat in underground mines.