基于RBF神经网络的船舶辐射噪声预测

To Predict the Radiation Noise of Ship Based on RBF Neural Network

本文根据船舶辐射噪声信号的混沌特性,基于重构相空间的轨迹演化规律,构建一种快速的RBF神经网络对船舶辐射噪声进行预测。实验证明,运用这种RBF神经网络对船舶辐射噪声的预测精度高于Volterra自适应滤波预测滤波器,而且更快的收敛速度。

According to the chaotic characteristic of ship radiation noise signal,based on the reconstruction evolutionary regularity of reconstruction phase space,this paper constructs a rapid RBF neural network to predict the radiation noise on the ship.Experiments have proved that using the RBF neural network to predict the ship＇s radiation noise is higher than that of Volterra adaptive prediction filter,and has faster convergence.