基于PLS回归分析建立岩移参数预测模型

Based on PLS regression to construct forecasting model concerning strata-moving parameters

根据《煤矿测量手册》提供的关于岩层移动角的观测资料 ,论文试图用普通最小二乘回归建立岩移参数 β关于煤层倾角、煤厚、采深等参数因子的回归模型 ,结果没有成功 ,按照《煤矿岩移与地表移动》的介绍 ,回归的结果不符合实际情况。在这种情况下 ,作者应用偏最小二乘回归建立岩移参数 β的统计模型。结果表明 ,所建的统计模型与《煤矿岩移与地表移动》中的相比 。

On the basis of the surveying data of strata- moving angle which are provided by reference [1] ,this paper use ordinary least squares regression to construct a regression model which is strata moving parameter β concerning the coal bed obliquity,coal thickness,mining depth,etc,the regression is unsuccessful.According to reference [4] ,this regression result is not inconsistent with fact.In terms of this situation,this paper use partial least squares regression (PLS regression) to construct the statistic model of strata moving parameter β .The result shows it's reasonable.