准确地测量、预测围岩变形是保证隧道施工安全的重要措施之一.围岩变形具有时间序列的单调增长性和随机性等特点,针对其特点,文章采用常规GM(1,1)模型、齐次指数函数灰色模型、非齐次指数函数灰色模型,对隧道的围岩变形进行了预测与对比分析.结果表明,常规GM(1,1)模型、齐次指数函数灰色模型、非齐次指数函数灰色模型三种模型的中误差分别为7.16%、5.01%、2.99%,最终沉降预测值相对误差分别为-8%、-5.52%、-3.05%.非齐次指数函数灰色模型具有较高的精度,在隧道围岩变形预测中具有广泛推广的应用价值.%Considering that accuracy in measuring and predicting the rock deformation featured by time-varying in?crease and randomness is a key measure for ensuring the safety of tunnel construction, this paper predicts the rock deformations under three models, i.e., a conventional GM (1,1) model, a homogeneous exponential function GM (1,1) model and a non-homogeneous exponential function GM (1,1) model, and makes a relevant comparison. The results show that:the errors under the three models are 7.16%, 5.01%and 2.99%, respectively, while the relative errors are-8%,-5.52%and-3.05%, respectively, indicating that the non-homogeneous exponential function GM (1, 1) model is highly accurate for deformation prediction.
展开▼
