Study on the Modified Model of Probability Integra Method

The fast convergence of boundary is a major problem of probability integral method, especially for the prediction of mining under thick alluvium, abandoned mine goaf, etc. To solve this problem and improve the accuracy of probability integral method, a modified model is constructed for surface movement and deformation based on theoretical analysis. In which, three weighted values i.e. 1, -1/t1, 1/t2 are set for the mining movement and deformation of three different major influence radius i.e. r1, R1, R2, respectively, then the parameters are selected as following: (1) Parameters are inversed based on the observation data, in which the subsidence factor q, deviation of inflection point s, displacement factor b are considered as generic parameters, the tangent of major influence angle tanß is modified as tanß2, that is the major influence radius r is modified as r2. (2) Inversion parameters of tangent of major influenceangle tanß or major influence radius r, based on the observation station edge data which beyond major influenc eradius distance from work-face mining boundary. The inversed tanß and r will be used as tangent of major influence angle tanß1 and major influence radius r1, respectively. (3) Other relevant parameters i.e. t1, t2, R1, R2, are calculated with the algorithm of this paper. A comparison experiment between the traditional and modified models has been conducted using the observation data from a observation station, the results show that in its central area, 58.6 percent of observation points accuracy is improved and 36.6 percent is decreased, while in its edge area, 61.5 percent is improved and 15.4 percent is decreased, totally, 68.6 percent is improved and 30 percent is decreased. Therefore, we can draw the conclusion that our modified model can improve the prediction predication accuracy, in particular, the boundary region.