Multiple logistic regression based prediction of heat flow direction in an intake incline of shallow depth by integrating thermal flywheel effect: A case study

Abstract

The diurnal changes of psychrometric conditions in a shallow intake incline of an Indian coal mine were investigated. On different days spread over thirteen months, 24-hour studies were conducted. Based on the findings, the cyclic behavior of the heat load was documented. The Kendall rank correlation analysis between heat parameters of surface air and the airway reveals that the wet-bulb depression of inlet air is the most important factor for air to rock heat transfer, while sensible heat of inlet air is dominating factor for rock to air heat transfer. A multiple logistic regression-based model was proposed for predicting the direction of heat flow. Using Youden's index, the optimal cut-off value for the model was determined to be 0.478. Performance evaluation of the proposed model on the testing dataset shows an F-score of 0.8182 and an accuracy of 80%.