摘要
暂态混沌神经网络是一种基于Hopfield网络提出的混沌神经网络,具有收敛速度快、不易陷入局部极小等优点.许多低阶的混沌系统都可以展成二阶volterra级数,因此提出一种基于暂态混沌神经网络和volterra级数的低阶混沌时间序列预测方法.该方法利用暂态混沌神经网络计算系统的volterra级数系数,确定系统的动力学模型,从而实现混沌时间序列预测.利用Logistic模型对该方法进行测试,结果表明,预测相对误差小于0.5%,预测可达到较高的速度和精度.
The transiently chaotic neural network （TCNN） is proposed based on the Hopfield neural network （HNN）, but it can escape from local minima more efficiently than HNN. Since a lot of simple chaotic time series can be expanded as second-order volterra series, a new prediction method based on TCNN and Volterra series is proposed in this paper. The method employed a TCNN to compute the coefficients of Volterra series, and then to decide the system＇s model. By using this model, a chaotic time series prediction method is realized. A logistic map is used to test the proposed model. The results show that the relative error is lower than 0.5%. The test results show that not only the convergence speed but also the accuracy are very high.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2007年第2期165-168,共4页
Journal of Harbin Engineering University
