School of Mining, Liaoning Technical University

为提高煤矿对矿井涌水量预测的准确性,基于ARIMA季节乘积模型,提出一种新的矿井涌水量的预测方法,通过普通差分和季节差分保证矿井涌水量时间序列的平稳化,以模型定阶、参数估计和假设检验等过程建立合适的乘积季节模型ARIMA（2,1,1）（1,1,1）12。利用该模型对某煤矿2015年各月的涌水量进行预测,得出预测结果,并与实测数据进行了对比分析。研究结果表明:预测结果与实际数据最大误差为3.43%,最小误差仅为0.77%,与实测数据有较好的拟合,预测效果较好,能够很好地满足煤矿实际需求,验证了乘积季节模型可以对矿井涌水量的能做出准确预测,为煤矿生产中涌水量预报和水害防治工作提供了新的思路。

In order to improve a prediction accuracy of the mine water inflow volume in a coal mine, based on the ARIMA seasonal product model, a new prediction method of the mine water inflow volume was provided. The ordinary difference and the seasonal difference were applied to ensure the mine water inflow volume of the ti me series stabilized method. A suitable product seasonal model ARIMA( 2, 1, 1)( 1, 1, 1) 12 was established with the model order, parameter estimation, hypothesis te st and other process. The model was applied to predict each month water inflow volume in the year of 2015 in a mine and the predicted results obtained were compared and analyzed with the actual measured data. The study results showed that there a max error of 3. 43% between the predicted results and the actual data. The min error was only 0.77%. The predicted results could be well fitted with actual measured data. The predicted effect was good and could well meet the actual requirements of the mine. The product seasonal model verified could make the accurate prediction of the mine water inflow volume and could provide a new idea to predict the mine water i nflow volume from the coal mine production and to prevent and control of the water disaster.

国家自然科学基金资助项目(51604139)；