基于VMD和GRNN的混沌时间序列预测  

Chaotic Time Series Prediction Using Variational Mode Decomposition and Generalized Regression Neural Network

作　　者：杨洪军[1] 徐娟娟[1] 刘杰[1] YANG Hong-jun;XU Juan-juan;LIU Jie(College of Manzhouli,Inner Mongolia University,Manzhouli Inner Mongolia 021400,China)

机构地区：[1]内蒙古大学满洲里学院,内蒙古满洲里021400

出　　处：《计算机仿真》2019年第3期448-452,共5页Computer Simulation

基　　金：内蒙古高等学校科学研究基金资助项目(NJZY6558)

摘　　要：随着非线性混沌动力学的发展,混沌时间序列的预测已经成为一个非常重要的研究方向。针对混沌时间序列的非线性和非平稳性的特点,提出一种变模态分解(VMD)和广义神经网络(GRNN)相结合的混沌时间序列预测方法,首先将混沌时间序列分解为多个固有模态函数(IMF)和余量(RF),然后对每个分量建立GRNN预测模型,最后将各分量的预测结果之和作为混沌时间序列的预测结果。采用Mackey-Glass混沌时间序列作为仿真实例,实验结果表明VMD-GRNN模型的预测精度相对于BP、ARMA和EMD-GRNN均有提高,证明了上述方法的有效性。With the development of nonlinear chaotic dynamics,the prediction of chaotic time series has become a very important research direction.In view of the complexity,nonlinearity and nonstationarity of chaotic time series,a chaotic time series prediction model is proposed based on VMD and generalized neural network(GRNN).Firstly,the sequence was decomposed into several intrinsic mode functions(IMF)and residual(RF);then a GRNN prediction model for each component was established.Finally,the prediction results of each component were taken as the prediction results of nonlinear time series.The method was applied to Mackey-Glass chaotic time series prediction.The simulation results show that the VMD-GRNN models have higher predication accuracy compared with BP,ARMA and EMD-GRNN,thus illustrate the effectiveness of the method.

关 键 词：变分模态分解 广义回归神经网络 混沌时间序列 预测 

分 类 号：TP391.2[自动化与计算机技术—计算机应用技术]