Abstract:
Directional drilling technology can accurately control the tracks of borehole and effectively improve the depth of borehole,which has been more and more widely used in gas control of coal mine and geological structure exploration. However,directional drilling tracks in underground coal mine are often different from the design tracks in the construction process. When multiple drilling holes exist in the same working face,the phenomenon of holes collision may occur. Therefore the effect of gas extraction is affected and resulting in potential safety risks,or hole collapse lead to economic losses. For this problem,an anti-collision hole prevention method based on double objective twin support vector machines for directional drilling in coal mine is proposed,and is verified by directional drilling experiment in a coal mine. The method can be integrated into measurement while drilling (MWD) software to judge whether the hole collision may occur in real time by using the measured data of drilling,and real-time guidance to the site construction is provided at the same time. Firstly,the absolute coordinate system is established by the zero-point borehole,all boreholes are contained on the same working face. Then,the dip Angle,azimuth angle and tool facing angle of the newly completed drilling measuring points are selected as input parameters,and the regression model is established by the double-objective twin support vector regression algorithm. Finally,the absolute coordinates of the predicted points are cut by the x axial direction,and the distance between the intersection points and the predicted points is obtained to determine whether there is a risk of preventing the hole collision. The experiments show that this method can guide the construction of directional drilling under coal mine, effectively avoid the risk of hole collision, eliminate the blind area of extraction and mining, and provide technical support for the improvement of the construction quality and drilling efficiency of directional drilling.