A Prediction Method for Underwater Acoustic Chaotic Signal Based on RBF Neural Network

摘要

In this paper, the chaotic time series RBF neuralnetwork model was designed. A prediction method forunderwater acoustic chaotic signal based on RBF neuralnetwork is proposed in this paper according to thecharacteristics of chaotic signal with the short-termprediction. Typical Henon chaotic signal and the actualunderwater acoustic chaotic signal are respectivelypredicted by the RBF neural network. Then the predictionresults are analyzed. The results show that the proposedprediction method increases at least two orders ofmagnitude in the mean square error terms compared withexisting prediction method, and that the RBF neuralnetwork can be used to predict the chaotic signal effectively.