基于灰建模的瓦斯含量多变量预测模型研究

研究了GM(1,N)、GM(0,N)瓦斯含量预测模型的数学原理,收集郑煤集团告成矿地质勘探期间及生产期间的瓦斯含量实测资料,获得16个可靠点,选取基岩厚度、新生界厚度、煤层厚度、煤层水分、煤层灰分、50m顶板含砂率6个因素作灰色建模预测的指标,分别建立了GM(1,6)和GM(0,6)瓦斯含量多变量预测模型。根据计算和评价结果,GM(1,6)和GM(0,6)瓦斯含量预测模型精度均能够满足工程精度的要求,说明利用灰色模型来预测瓦斯含量是可行的。由于前者精度略高于后者,故建议告成矿采用GM(1,6)模型来进行未知地区煤层瓦斯含量的预测。需要注意,由于模型没有考虑构造的影响,在实际预测时,还应根据构造对待预测区的影响关系和影响程度对模型的预测结果进行修正。

Study on Multivariate Forecast Model of Gas Content Based on Grey System Theory Modeling

The mathematic principles of GM(1, N)and GM(0, N)for gas contents were firstly studied. Then, the actual measurement data of gas contents during geological prospecting and mining of Gaocheng mine were collected, and sixteen reliable dots were gained. By selecting 6 factors including bedrock thickness, cenozoic thickness, coal seam thickness, coal seam moisture, coal seam ashes and the sand-contained rate of 50 m top surface as the gray modeling forecast indicators, the multivariate forecast models of GM(1, 6)and GM(0, 6)for gas contents were respectively constructed. The result calculated by use of the two models shows that the precision meets the requirements of engineering and the models can be used for predicting gas contents. Meanwhile, it is suggested to use GM(1, 6)to forecast the gas contents of unknown area in Gaocheng mine due to that it is more precise than GM(0, 6). Finally, It is emphasized that the forecast result must be amended because the geological structure is not taken into consideration in these models.