混沌时间序列预测的局域法在边坡变形分析中的应用

边坡作为一个复杂系统,其本身的各种参量是不确定的和随机的,在其演化过程中,表现出复杂的非线性行为,发生一系列的混沌现象。本文运用现代混沌理论,对边坡变形的预测问题进行探索性研究,把混沌时间序列理论引入到边坡工程研究中,对该理论的建立及预测方法进行系统地讨论,为该领域的研究提供完整的技术方法。通过对新滩滑坡的研究结果表明,混沌时间序列方法对混沌序列的预测较线性时间序列具有较高的精度。

A PREDICTION METHOD OF CHAOTIC TIME SERIES FOR SLOPE DEFORMATION

The parameters of the slope system are uncertain and random due to slope complexity, therefore, a series of chaos can occur in the course of slope evolution. The chaotic time series method is developed based on the chaotic theory and adopted in predicting slope deformation. The development of the prediction model and its analysis are demonstrated using the case study of Xintan landslide. The results indicate that this prediction method can get a better prediction and analysis accuracy than conventional time series method.