Verhulst反函数模型滑坡起始预测时刻的选择

 Verhulst反函数模型用于滑坡预测的关键在于起始预测时刻t0的选择。针对用定量计算方法选择t0，研究模型中对数函数自变量存在出现负值的问题。计算所有t0满足模型要求的累计位移和相对位移预测数据，形成预测曲线，通过分析预测曲线与实际监测曲线选择起始预测时刻t0值，得出其选择的主要原则：(1) 预测位移在观测序列点n和n +1处成立；(2) 预测累计位移曲线与实测曲线形态相似；(3) 预测相对位移与实测相对位移在点n处差值最小。多种已知滑坡试验表明：其临滑时刻的预测精度可以达到1个时间间隔以内，t0选择方法是有效的。

SELECTION OF INITIAL PREDICTION TIME FOR LANDSLIDE WITH VERHULST INVERSE FUNCTION MODEL

It is a key problem to select the initial prediction time with Verhulst inverse function model in landslide study. Generally，there is a problem of selecting by quantitative calculation method in that the independent variable of logarithm function in the model may be negative. The predicted data of accumulated displacement and relative displacement are calculated with all ，and all predicted -curves are formed. The initial prediction time can be selected with analyzing the predicted curves and monitoring curves. The selecting principles are as follows：(1) the predicted displacements can be calculated on points n and n+1 of the monitoring list；(2) the curve configuration of predicted accumulated displacements is similar to the monitoring curve；and (3) the dispersion between the monitoring relative displacement and its predicted value on the point n is the least. The experimental results of several groups of measured data show that the precision of critical sliding time can be obtained within one time interval；and the method to choose is effective.