论文部分内容阅读
本文根据需要,提出了不等时距的灰色预测GM(1,1)模型,它克服了等时距GM(1,1)预测模型要求原始数据列连续的缺陷。用提出的模型对碧口水电站右坝肩滑坡体从1982年到1989年(1987年空缺)的实际观测资料进行预测,所得结果与实测值误差均在5 mm以下,并用此模型预测了1990年到2005年三个监测点的水平位移年平均累计值及相应的年均速率,指出了该滑坡体可能滑动的年限,为工程实际提供依据,具有重要的意义。
In this paper, we propose a gray prediction GM (1,1) model with unequal time-distance, which overcomes the problem that the original data sequence is continuous when the isochronous GM (1,1) prediction model is used. The actual observation data of the right abutment landslide body of Bikou hydropower station from 1982 to 1989 (vacancy in 1987) are predicted by the proposed model. The errors between the measured results and the results obtained are below 5 mm. By 2005, the annual average cumulative horizontal displacements at three monitoring points and the corresponding annual average velocities indicate the possible sliding years of the landslide body, which provides the basis for engineering practice and is of great significance.