【作者】 舒赛刚； 任雪梅； 陈杰；

【Author】 Shu Saigang, Ren Xuemei, Chen Jie (Department of Automatic Control, Beijing Institute Technology, Beijing 100081, China) (Department of Automatic Control, Beijing Institute Technology, Beijing 100081, China ) (Department of Automatic Control, Beijing Institute Technology, Beijing 100081, China )

【机构】 北京理工大学自动控制系206教研室；

【摘要】 针对混沌系统的特点,本文提出一种新的模糊神经网络用于混沌系统的预测,同时给出有效的算法,该算法训练速度快且预测精度高,并且用仿真实验来验证该模糊神经网络对混沌系统具有很好的预测性能。本文利用该模糊神经网络对一维Logistic map标准混沌系统进行预测仿真实验,同时与一般的神经网络进行对比,最后还讨论了该模型对被噪音污染的混沌数据的处理能力。更多还原

【Abstract】 According to the characters of the chaotic systems, this paper presents a new fuzzy neural network to predict the output of the chaotic systems, which has a higher training speed and is more accurate in predicting chaotic systems than the general neural network. An effective algorithm to train the fuzzy neural network is also presented. The effectiveness of the proposed fuzzy neural network is demonstrated by simulating experiments both on prediction of the one dimension Logistic map standard chaotic system and on dealing with the chaotic time series corrupted by noise signal.更多还原