通过对边坡位移历史数据序列进行特征分析 ,计算出饱和嵌入维数和最大 Lyapunov指数 ,给出了边坡位移的最大可预报时间尺度。在此基础上 ,确定了神经网络的输入节点数 ,建立了基于自适应神经网络的边坡位移预报方法 .通过对三峡升船机高边坡和新滩滑坡实际位移数据进行预测 ,结果令人满意 .这对于建立边坡位移的实时监测 -预警系统有重要意义 By analyzing the characteristics of slope displacement historical data series, the maximum embedding dimension and maximum Lyapunov exponent are calculated, and the maximum predictable time scale of slope displacement is given. On this basis, the number of input nodes of the neural network is determined, and a method of slope displacement prediction based on adaptive neural network is established.By predicting the actual displacement data of the Three Gorges ship lift high slope and Xintan landslide, Satisfactory. This is of great importance to establishing a real-time monitoring and early warning system for slope displacement