Cigar Lake Mine Convolutional Neural Network

A Convolutional Neural Network (CNN) was developed for Cigar Lake Mine, Saskatchewan, Canada, to predict tunnel liner yield. The mine experiences a complex time-dependent ground squeezing behaviour resulting from the poor geological conditions and the artificial ground freezing implemented to stabilize the ore cavities and to control ground water during the ore extraction process. Four inputs were used inn the CNN to make this prediction: geotechnical zone mapping, primary support class, ground freezing pattern, and measured tunnel displacement. A sensitivity analysis of the CNN training parameters, called hyperparameters, was completed to optimize the final CNN performance. Hyperparameters analyzed include: the amount of training data, the convolutional filter size, and the error weighting scheme. Two final models were developed, one balanced model able to accurately predict tunnel liner yield across all classes of severity, and one targeted model that is calibrated to predict the higher classes of tunnel liner yield particularly well. Input Variable Selection (IVS) was applied to examine how the CNN used the given data, or inputs, to forecast rock mass behaviour. The three IVS methods investigated were Channel Activation Strength (CAS), Input Omission (IO), and Partial Correlation (PC). The IO and PC approaches proposed are novel for CNNs using a spatial and temporal geomechanical dataset.