基于多元线性回归的煤矿瓦斯含量分析与影响因素研究

Gas content and influential factors of coal mine based on multiple linear regression

  • 摘要: 【目的及方法】目前定量研究分析煤矿影响因素与煤层瓦斯含量之间的复杂关系多采用Langmuir吸附动力学经典模型,但该公式不能很好地针对煤质复杂、特殊的地质条件或有开采工程因素影响的情况,其公式中的变量在上述情况影响下会变得不够准确,进而干扰预测结果。故基于Langmuir公式和工业分析内容,在潞安矿区的不同矿井进行现场取样分析,采用相关性分析、多元线性回归和对比分析相结合的方法进行研究,将影响瓦斯含量的影响因子纳入多元线性回归模型,建立了瓦斯含量与影响因子的数学预测模型,最后,将该模型的预测值与实测数据进行对比验证。【结果】结果表明,在潞安矿区中,其瓦斯压力对含量影响最为显著,其次是吸附常数a,最后是挥发分,得到的数学回归模型为W=-0.224Vdaf+1.848P+0.242a-2.235。【结论】所得结果可为矿井瓦斯含量数据预测提供可靠支撑,进一步补充和完善现有知识体系,同时为相关研究提供了新的思路和方法。

     

    Abstract: At present, quantitative research and analysis of the complex relationship between coal mine influential factors and coal seam gas content mostly use the Langmuir adsorption dynamics classical model, but the formula can not well target at the case of complex coal quality, special geological conditions or the influence of mining engineering factors, and the variables in the formula will become inaccurate under the influence of the above conditions, which will then interfere with the prediction results.Based on Langmuir's formula and industrial analysis, this paper carries out on-site sampling and analysis in different mines in Lu'an Mining Area, and adopts the methods of correlation analysis, multiple linear regression and comparative analysis to study, and incorporates the influential factors affecting the gas content into the multiple linear regression model, so as to establish a mathematical prediction model for the gas content and the influential factors, and finally, compares and verifies the predicted value of the model with the actual data.The results showed that in Lu'an mine, its gas pressure has the most significant effect on the content, followed by the adsorption constant a, and finally the volatile fraction, and the obtained mathematical regression model is W=-0.224Vdaf+1.848P+0.242a-2.235.The obtained results can provide reliable support for the prdiction of mine gas content data, further supplement and improve the existing knowledge system in the field, and meanwhile offer new ideas and methods for related research.

     

/

返回文章
返回