Stochastic modelling of Continuous Miner based underground coal mine production planning through curve-fitting and regression approach

In India, three major and effective technologies for underground coal mining are continuous miners (CM), shortwalls, and longwalls. According to the relevant literature, the available studies have not been able to successfully link the productivity of CM-based systems to the related independent parameters. However, this study makes an effort to find correlations between a number of independent variables and production-related attributes using regression modelling and curve fitting. A combined relationship was constructed considering all independent parameters as input and production-related parameter as output, and a relationship was made considering each input parameter separately with the production-related parameter as output. The idea of their strong association to represent the interrelationship conveyed by their statistical goodness of fit (R2 and RMSE) and the results of the t-test further support their relevance. The greatest R2 value was shown for seam thickness (0.8779), followed by pillar area (0.84), gradient (0.8112), and coal strength (0.6008). Besides geo-mining input parameters, the cutter motor power has shown a better correlation with production-related parameter, with its corresponding R2 values of 0.8033. The pertinent data from three CM panels (not considered during equations generation) were then incorporated into the derived interrelationships. These validation results were then compared using a t-test to the equivalent actual field data, which revealed a minor discrepancy. The interrelationships were found to be applicable and pertinent in CM–based underground coal mining situations.