Tool Face Angle Prediction for Directional Drilling in Underground Coal Mine

The trajectory of directional drilling borehole is liable to deviate due to drilling parameters, formation characteristics and workers’ mis judgment. In order to explore the impact of various factors on the slope rate of directional drilling and to achieve machine intelligent decision-making tool face angle. A deep learning based neural network model is proposed to predict tool face angles, taking into account variables such as drilling inclination angle, azimuth angle, drilling speed, rock hardness, and screw motor bending angle. The model was trained and validated using actual drilling data from two coal mines in Yulin, northern Shaanxi, China., the prediction model based on deep learning for tool face angle prediction, after training, the maximum error of prediction is 4.7%.