论文部分内容阅读
RBF网络是一个三层的前馈型神经网络,它隐含层的转换函数是局部响应的非线性函数,所以它能够以任何精度逼近任意连续函数,这为复杂的变形系统的解释和模型化提供了可能,因而利用RBF网络对混沌时序的分析和预测是变形分析的一种新的途径。本文首先介绍RBF网络,对其变形监测数据的混沌现象进行分析和对RBF网络的混沌时间序列作出分析、预测,最后,总结出运用RBF网络对变形分析和预测对数据拟合模型的精度和预测能力都有很大的提高作用。
The RBF network is a three-layer feedforward neural network whose implicit layer transfer function is a non-linear function of local response, so it can approximate any continuous function with any precision, which explains and models the complex deformation system So it is possible to use the RBF network analysis and prediction of chaotic time series is a new way of deformation analysis. This paper first introduces the RBF network, analyzes the chaos phenomenon of its deformation monitoring data and analyzes and predicts the chaotic time series of the RBF network. Finally, the accuracy and forecasting of the data fitting model based on deformation analysis and forecasting using RBF network are summarized Ability has greatly improved.