The Estimation Model Research On Core Index Of Coal Quality In Coal Mining Face

In the increasingly competitive coal industry,strengthening the management and control of coal quality in the process of coal production is the key to improving the efficiency of coal enterprises,and it is also a powerful guarantee for coal enterprise to be in an invincible position in market competition.Based on the actual demand of coal quality management in coal enterprises,the paper takes Kriging interpolation method as the theoretical foundation,introduces the differential evolution algorithm to carry on the optimization solution,and explores the theory and the method of coal quality core index estimation in coal mining face.The specific content of the study is as follows:1.By analyzing and studying the differential evolution algorithm,the scaling factor in the differential evolution algorithm is redesigned,so that its mutation process can dynamically modify the evolution direction and avoid the occurrence of "premature" phenomenon,besides it solves the problem that the convergence speed slows down in the late of evolution.Through the simulation experiment of solving the optimal parameters of the function,the advantages of the adaptive differential evolution algorithm(UMDE),which can modify the mutation direction in parameter optimization are verified.2.The UMDE algorithm is used to optimize the solution of three model parameters(nugget,partial abutment,and range)in the variogram function,and then an optimized Kriging interpolation model is constructed,which is applied to estimate the core index of coal quality.Simulation experiments on real coal quality data show that the variogram function in this model can describe the current geological spatial features more accurately than the Kriging interpolation models constructed by other parameter optimization schemes,the estimation accuracy of coal quality core index is higher.3.Four characteristics of geological structure change are selected,including slip distance,drop,distance from fault and location of fault.The UMDE algorithm is used to optimize the selection of support vector regression parameter pairs(insensitive loss coefficient,penalty coefficient,width coefficient of kernel function).Simulation experiments on real coal quality data show that,compared with other regression methods,the error correction model based on optimized support vector regression(UMDE-SVR)can effectively reduce the error of coal quality core index estimation model in the area of geological structure change.This paper studies the optimization scheme of Kriging in geostatistics,constructs the best coal quality estimation model,improves the accuracy and automation degree of coal quality core index estimation in coal face.An error correction model is established in the changing area of geological structure to solve the problem of decreasing estimation precision caused by the change of geological structure.It provides technical support for coal enterprises to realize " intelligent mining,quality mining,and transparent mining face ".