Simulation and application research of chaotic time series prediction based on RBF neural network

This paper discusses a method for chaotic time series prediction based on radial basis function (RBF) neural network. The number of input nodes for RBF is determined by embedding dimension based on chaotic phase-space reconstruction. Both Grassberger-Procaccia algorithm and Takens' method are employed to calculate minimal embedding dimension of chaotic time series. Finally, the prediction accuracy was evaluated by Mean Square Error (MSE). The chaotic time series data from Lorenz simulation signal and rolling bearing vibration signal was used to verify the proposed method. It was found from the experimental result that, this method is effective and feasible for the prediction of chaotic time series.