Study on Dynamic Prediction of Mining Subsidence with Segmented Weibull Time Function

Abstract

The maximum subsidence speed moment τ of the surface point was set as cut-off point to divide the dynamic subsidence process into two phases of approximate symmetry. By combining it with deviation correction and growth function model to eliminate the oretical error, a segmented Weibull function was proposed, the regression equation of τ and the distance x between the surface point and the open hole was established, and the determination method of lithologic parameters c and k was discussed, which effectively improved the deficiency of Weibull time function in dynamic prediction. On the basis of the measured data of 3214 working face in a mine in Shanxi Province, the subsidence values of single point and main section were predicted and verified, and compared with those obtained by Weibull, Knothe and optimized segmented Knothe functions. From the analysis it was found that the prediction accuracy of piecewise Weibull time function is 41.09%, 51.03%, and 3.47% higher than above functions, respactively, at a single point. For the main section, the relative error in the first peried is the largest, only 5.25%. The segmented Weibull time function has certain reliability and can be used for dynamic prediction with higher accuracy in subsidence areas.

Keywords

• segmented weibull
• dynamic prediction
• mining subsidence
• time function