动态灰色预测模型在崩落法矿山地表沉降预测中的应用

Application of dynamic gray forecasting models in forecasting of surface subsidence for mines using caving method

针对灰色预测模型GM(1,1)更适用于短期预测的问题,以其为基础采用等维新息和等维灰数递补两种动态灰色预测方法对矿山地表的沉降量进行了中长期预测研究。探讨了等维新息模型和等维灰数递补模型预测的实现机理,对不同模型的预测结果和精度进行了比较研究。研究发现等维新息模型可以利用新息调整灰色模型,等维灰数递补模型可以利用灰度约束灰平面的大小,两者相结合能够提高模型的预测精度。工程实践表明:将等维新息模型和等维灰数递补模型结合运用,适用于中长期的沉降预测,能够获得更为合理和精确的预测结果。

According to problems of the GM(1,1)forecasting model is more suitable for short-term forecasting,dynamic forecasting models by incorporating the equal-dimension gray filling technique and the equal-dimension new information technique with the GM(1,1)forecasting model were established and applied to forecasting the medium and long-term subsidence of a mine.The study of the forecasting mechanism of the equal-dimension new information forecasting model and the equal-dimension gray filling forecasting model and a comparison of the forecasting accuracy of different forecasting models were carried out.The results indicated that the equal-dimension new information model can adjust the gray model by using new information.The equal-dimension gray filling model can adjust the gray model by restraining the gray plane size with gray number.It showed that the forecasting accuracy of the combination forecasting modelwas better than of any individual forecasting model.The industrial test results indicated that the combination forecasting model of the equal-dimension gray filling forecasting model and the equaldimension new information forecasting model were suitable for forecasting the medium and long-term subsidence of a mine.The combination forecasting model can get more reasonable and accurate forecasting results.