基于时序分析的边坡变形预报与变形行为特征

Prediction and Behavior Characteristic Research of Slope Deformation based on Time Series Analysis

把数理统计软件Eviews引入岩土工程领域,对东明矿J2边坡监测点高程监测数据分别建立双对数模型、指数模型、线性模型、三阶多项式模型进行回归分析,找出物理量随时间变化的规律,然后通过相关统计学及力学假设,依托现有实测工程数据,建立时序模型并优化之,将经典ARMA模型推广到ARIMA模型,提高了拟合和预测精度;依据所建模型给出位移预报曲线,并探讨了边坡变形行为特征,对揭示边坡系统变形规律,选取最优防护体系具有指导意义,为工程数据分析和数值模拟提供了新思路.

Mathematical statistical software-Eviews is introduced into geotechnical engineering field, based on the data of J2 slope elevation monitoring point in Dong Ming ore, hi-logarithm model, exponential model, linear model, as well as third-order polynomial model are established respectively. With regression analysis, the regularity that how data changes as time goes along is found out. Then, according to relevant hypothesis of statistics and dynamics, with current measured engineering data, time-series model is established and optimized. What＇s more, the classical ARMA model is upgraded to ARIMA mode, which promotes the approximating and forecasting precision. Based on this model, displacement prediction curve is given and the deformation behavior characteristics are discussed . There is guiding significance in revealing the regularity of slope deformation, as well as in selecting the best protection system. New ideas are also provided for later engineering data analysis and numerical simulation.