论文部分内容阅读
边坡系统的复杂性使得部分位移监测序列呈现出非线性分布特征,GM(1,1)模型、Verhulst模型有时不能解决上述时间位移序列的模拟问题,而DMG模型适合解决具有摆动发展趋势的序列问题,此时可以考虑采用DMG模型。边坡变形监测时间序列一般为非等时距序列,可采用Lagrange线性插值法进行等时距化处理。建立了DMG模型,通过联合使用残差修正模型的方法,使模拟的结果达到较高的精度。运用DMG、GM(1,1)和Verhulst模型对某高速公路边坡变形序列进行了模拟,通过对比分析验证了这种模型在边坡变形模拟预报方面应用的可行性。
The complexity of the slope system makes partial displacement monitoring sequence appear non-linear distribution characteristics. The GM (1,1) model and the Verhulst model can not sometimes solve the above-mentioned simulation problem of time-series displacement. However, the DMG model is suitable for solving the sequence with the tendency of oscillation Problem, at this time can consider the use of DMG model. The time series of slope deformation monitoring is generally non-isochronous sequence, and Lagrange linear interpolation can be used to treat the time interval. The DMG model is established. Through the joint use of the residual correction model, the simulation results achieve higher accuracy. By using DMG, GM (1,1) and Verhulst model, the slope deformation sequence of a freeway is simulated, and the feasibility of this model in slope deformation simulation is verified by comparative analysis.