基于小波神经网络理论的边坡位移预测

研究边坡位移混沌时间序列的预测,利用混沌系统的相空间重构理论,提出基于小波神经网络的边坡位移预测方法.通过计算表明,该方法与其它方法相比可避免误差曲面局部最小,网络节点少,参数确定较为容易,学习效率高,收敛速度快,自适应性强,精度高等优点,为边坡位移预测提供了一种可行的、新的探索途经.

Slope displacement forecast based on wavelet neural network

This paper studies the chaotic time series of the slope displacement forecast. Using the theory of reconstructing phase space in the chaotic time series. It puts forth a method of the slope displacement forecast based on the wavelet neural network. The calculated results show that compared with the other method this method can avoid the local minimal error margin curved face and the network node little, the parameter is more easy to determine, the study efficiency is high, the convergent speed is quick, the oneself adaptability is strong, and the computation accuracy is very high etc. This method offers a viable and new way for the slope displacement forecast.