Development of a Machine Learning Model to Estimate Refrigeration in an Underground Mine

The rise in temperature in an underground mine is one of the critical challenges. The main causes of increasing mine air temperature include geothermal gradients, machinery operation, depth, width of the airway, and limited natural ventilation. High temperatures cause discomfort and adverse health effects among miners and also reduce worker efficiency, impair equipment performance, increase energy consumption, and lower overall productivity. Efficient cooling or refrigeration and ventilation systems are key to maintaining safe and sustainable working conditions. The study deals with the simulation-based parametric study and machine learning techniques that can be used in order to reduce heat in underground mine air. The air velocity, mine depths, and width of airways have been considered for understanding air circulation and heat transfer in varying conditions. Simulation-based data were gathered from a number of cases to study variations in airflow, heat buildup, and cooling demand. These datasets were then processed using machine-learning algorithms to develop predictive models that estimate cooling demand and determine optimal ventilation strategies. Using machine learning, thermal behaviour can be forecast from past data; hence, it is a smart and efficient method for improving ventilation design. The machine learning model takes into consideration several input parameters simultaneously and can give indications about the cooling methods that consume the lowest amount of energy while keeping air temperature within acceptable limits. It is also observed that simulation-based analysis and machine learning form a powerful framework for understanding and determining the refrigeration to control heat in underground mines. The results of this study evidence that data-driven prediction can be highly effective in improving cooling efficiency by providing proper refrigeration, decreasing operational costs, and contributing to a safer, more sustainable dee

View article page

Download PDF