提出采用多种模型和方法进行有机集成和综合应用将对边坡变形进行预测研究更为有效的途径。边坡变形的时间序列实际上可分解为趋势项和随机项两个部分。把回归模型与ARMA模型进行有机集成,建立了边坡变形的组合预测模型。选取锦屏一级水电站左岸高边坡的一个外观测点,建立边坡变形的组合预测模型,并与单一的统计回归模型作对比。对比结果显示,组合预测模型有着更高的预测精度。
A hybrid prediction model is established by integrating Statistics Regression model(SR) with Auto Regressive Moving Average(ARMA) model.The model is applied to simulate and predict the displacement of the left-side high slope in the Jinping Hydraulic Power Station based on historical data.The comparison of the predicted result indicats that the relative errors between prediction value of hybrid model and the result of Statisties Regression model is less than 4%,which is much better than that of Stastistics Regression model.