广义随机模糊神经网络及在随机混沌时间序列预测中的应用  

General Stochastic Neural Network and Its Application to Prediction of Stochastic Chaotic Time Series

作　　者：张静[1] 

机构地区：[1]襄樊学院物理系,湖北襄樊441000

出　　处：《噪声与振动控制》2006年第2期11-13,共3页Noise and Vibration Control

基　　金：湖北省教育厅科研项目(2001D69001)

摘　　要：针对随机模糊神经网络缺乏自适应性,引入广义高斯函数和广义随机模糊神经网络,使系统中隶属函数具有自适应性;并对参数进行遗传退火算法优化,使系统具有最佳结构和参数。以随机混沌时间序列为例进行仿真预测分析,结果表明广义随机模糊神经网络能够更好地预测原随机混沌时间序列,精度良好,具有抗噪声干扰能力。A general stochastic neural network （GFSNN）, which membership functions are general Gaussian functions and are adaptable, is proposed to predict chaotic time series, and the model＇ s structure and parameters are optimized by the algorithms of GA-Annealing strategy and are applied to forecast stochastic chaotic time series. The simulation results show that the GFSNN is superior in modeling stochastic chaotic systems and have advantage of good precision. This is the basic of controlling chaotic systems using fuzzy neural network controllers.

关 键 词：声学 广义随机模糊神经网络 随机混沌时间序列 预测 遗传退火算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]