对瓦斯含量预测数据的精确处理

Study on the precision treatment of gas content forecast data

为了证实本文数据处理方法的精准度,对张继豹等发表的"线性回归分析在邯峰矿区2#煤层瓦斯含量预测中的应用"一文中的挥发分和煤层倾角数据进行"线性化"处理,从而得到一个修正后多元线性回归分析方程。通过分别计算未修正前与修正后的多元线性回归方程计算值对比,证实对某些原始数据进行"线性化"修正,可以得到多元线性回归分析的相对平均误差仅为原来相对平均误差的44%。

To certify the accuracy of the data processing method in this article,by the linearization processing of volatile matters and coal seam dip angles which excerpted from "Application of linear regression analysis in Handan-Fengfeng mining area No. 2 coal gas content prediction",a modified multiple linear regression analysis equation was obtained. By comparing the calculated values between the unmodified and the modified multiple linear regression equation,it was proved that the relative average error of multiple linear regression analysis is reduced by 44% by linearizing some original data.