论文部分内容阅读
将灰色系统理论与离散状态的马尔可夫链理论相结合,提出了灰色马尔可夫预报模型以预报矿井涌水量.该模型针对灰色数据系列首先建立GM(1,1)模型进行趋势预测,然后利用马尔可夫状态概率转移矩阵预报方法对其预测值进行二次拟合,可提高波动性较大的随机变量的预报精度.实例计算表明:灰色马尔可夫预报模型精度明显高于GM(1,1)模型及GM(1,1)残差修正模型.该结论拓宽了灰色预报模型的应用范围,为矿井涌水量的科学预报提供了一种新方法.
Combining the gray system theory and discrete state Markov chain theory, a gray Markov forecasting model is proposed to predict the mine water inflow. This model firstly establishes the GM (1,1) model for the gray data series and then forecasts the trend. Then, The quadratic fitting of the forecasting value by the Markov state probability transfer matrix forecasting method can improve the prediction accuracy of the random variables with large volatility.Examples show that the accuracy of the gray Markov forecasting model is obviously higher than that of GM (1 , 1) model and GM (1, 1) residual error correction model. This conclusion broadens the application range of gray forecast model and provides a new method for scientific prediction of mine water inflow.