中文摘要：

采用基于位移判据的强度折减法分析边坡的稳定性.在坡高为20 m、坡角为45°的均质土坡的临界滑移线内外,按照坡面上、中、下各自布置3个监测点共9个监测点,通过FLAC3D自带的FISH语言,开发数据记录工具,记录不同监测点的水平位移与折减系数,并分析它们的关系.为了确定最优监测点以减少工作量和经费,运用灰色理论软件,通过对安全系数序列和各个观测点的水平位移序列进行灰色关联度分析,得出各个观测点的水平位移观测序列与安全系数序列之间的灰色（邓氏）关联度和相对关联度,通过分析对比关联度确定最优监测点;利用GM（1,1）预测模型对安全系数和最优监测点的水平位移进行预测.研究结果表明：与安全系数序列的邓氏关联度较大的1,2,4,5监测点可作为最优监测点;当安全系数为1.120 47时,各个监测点的水平位移达14.910～40.842 m,临界安全系数为1.120.

英文摘要：

Stability of the slope was analysed using the method of the strength reduction based on the displacement criterion. The target was a homogeneous, 20 m high soil slope that its dip angle was 45°. 9 monitoring points were set at the upper, middle and lower part of the slope surface totally inside and outside of its critical landslip line. FISH language of the FLAC3D was used as a computer language tool to program, horizontal displacements and the reduced coefficients of all monitoring points were recorded and their relationships were studied. In order to determine optimum monitoring point and save expenses, grey system theory was applied to get the degrees of grey incidence and the relative incidence between the array of the safety coefficient and the array of monitoring points＇ horizontal displacements to determine the optimum monitoring points. The safety coefficient and the horizontal displacement of the optimum monitoring point was forecast using the GM（1,1） model, and the mutation characteristic of the forecast value was analyzed to determine its critical safety factor. The results show that the points 1, 2, 4, 5 can be the optimum monitoring points, because its Dengs＇ degree of grey incidence with the safety coefficient is bigger than others. When the safety coefficient is 1.120 47, the horizontal displacements of the monitoring points is 14.910-40.842 m. The critical safety factor is 1.120.