煤岩界面预测灰模型

Grey model in coal-rock interface prediction

为提高煤岩界面记忆截割检测的自动化水平,基于灰建模改进方法,采用相同的初始数据序列,建立经典GM(1,1)、新数据GM(1,1)、新陈代谢GM(1,1)三种煤岩界面预测模型,并分析其数据残差和相对误差。结果表明,三种模型的残差和相对误差均随预测步数的增加而增大,新陈代谢GM(1,1)模型的残差和相对误差最小,模型精度较高,适用于煤岩界面预测。

This paper proposes a model designed for improving automatic standard of test approach to the cutting memory program of the coal-rock interface prediction. Based on the improvement of gray model, the model involves the adoption of the same original data sequence to establish such prediction model as the classical GM （ 1,1 ）, new date GM （ 1,1 ）, metabolic GM （ 1,1 ） and to analyze their data residual and relative error. The results prove that three models with the enlarging residual and relative error due to the increasing numbers of steps, and the minimum residual and relative error of metabolic GM （ 1,1 ）, give a greater accuracy and allow coal-rock interface prediction.