Chaotic Time Series Prediction Method Based on BP Neural Network and Extended Kalman Filter

Abstract

For neural networks, there are local minimum problems and slow convergence speeds. In order to improve the prediction accuracy of the BP neural network prediction model for chaotic time series, the EKF algorithm with BP neural network is used in the field of chaotic time series prediction. Namely, the use of the weight of its output of BP neural network is suitable for the state equation and observation equation of the Kalman filter, which gives the evolution of the Kalman filter algorithm suitable for nonlinear systems. Extended Kalman filter (EKF) algorithmtypical and Mackey-Glass chaotic time series were simulated. The simulation results show that the method of chaotic time series with nonlinear fitting better and higher prediction accuracy.