基于灰色理论的回采工作面瓦斯涌出量动态预测研究

瓦斯涌出量的准确预测直接关系到煤矿企业的宏观决策及系统布局。为了提高回采工作面瓦斯涌出量的预测精度，提出了采用灰色预测法对瓦斯涌出量动态预测进行研究，以车集矿2316回采工作面为例，通过重组瓦斯监测数据构建了灰色GM（1，3）动态预测模型，并依据后验差检验比值c及小概率精度p对模型预测效果进行了分析。研究结果表明，数据重组后的GM（1，3）模型的动态预测值平均相对误差为5.65%，后验差检验比值c<0.35，小误差概率p>0.95，预测精度达到了1级，在对2316工作面后期的瓦斯涌出量动态预测结果与实测值十分接近，平均相对误差仅有2.26%，变化趋势也高度吻合，灰色GM（1，N）预测模型能够实现对工作面瓦斯涌出量的实时、动态、准确预测。

Dynamic prediction for gas emission in coal winning face based on grey theory

The accurately prediction for gas emission has influenced to macro decision-making and system layout.In order to improve the accuracy of prediction precision for gas emission in the mining working face，the grey prediction method was proposed and applied to dynamically predicting gas emission.The measurement data，acquired from No.2316 mining working face of Juji Coal Mine，were reorganized to establish GM（1，3） prediction model.Then，the later examination difference ratio c and little probability coefficient p are proposed to evaluate the accuracy of model.The research results shows that the mean relative error of dynamic prediction value，obtained by GM（1，3）model after reorganized data，is 5.65%.The later examination difference ratio c is less than 0.35，and the little probability coefficient p is more than 0.95.The accuracy of prediction model reaches Grade 1.The prediction relative error is merely 2.26%，what performs that the prediction value is similar to actual value.GM（1，N） prediction model could predict gas emission of working face dynamic and accurately.