Chaotic Time Series Prediction Based on Improved Local Phase Space Weight Method

The phase space reconstruction is the basis of the chaotic time series analysis and prediction.On the view of Embedding dimension and delay time is associated,the C-C algorithm is applied to calculate the embedding dimension and delay time,thus the chaotic time series phase space reconstruction,and then use the improved weighted one-rank local method prediction model to predict.Lorenz chaotic system and Rossler chaotic system simulation predicted that the embedding dimension and delay time with the C-C algorithm has the advantages of simple,fast,and the proposed prediction model simulation predictions further shows that the proposed prediction the method is feasible and has a good effect on the short-term prediction of chaotic systems.