Feature Parameters Computation of Underground Coal Mining Space Coupled Probability Integral Model With SBAS-InSAR Subsidence Data

With the development of ground observation technology, non-contact localization of disordered goaf can be realized, which is the key to suppressing ground disasters and guiding government supervision. However, most of the existing localization methods are geophysical techniques, whose efficiency seems to be a challenge. By defining the mining space that is goaf approximately as a cuboid, the localization can be transferred to feature parameters computation of the cuboid. In this paper, considering for the first time the case of multi-seam mining, which means a working face with old goafs in the upper layer, a parameters computation method of subcritical underground long goaf based on probability integral model and small baseline subset interferometry synthetic aperture radar (SBAS-InSAR) is proposed. Firstly, the paper integrates SBAS-InSAR and Offset-Tracking method to obtain the line-of-sight (LOS) deformation as the real ground deformation of the goaf. Next, we propose a fusion of simplified parametric generalized probability integral model (SGPIM) with residual deformation, named RD-SGPIM, to develop a connection between ground deformation and goaf mining, which accumulates the residual deformation of the upper old goaf to calculate the simulated ground deformation. Finally, a multi-strategy improved sparrow search algorithm (TLGSSA) was utilized to determine localization results. The results show that the maximum errors of both simulated and actual experiments appear in the inclination angle, and the average relative errors of the remaining parameters are 0.19% and 3.71%, respectively, which verifies that this method is effective.