基于径向基函数神经网络的舰艇声呐部位自噪声预报

Forecast of naval vessel sonar self-noise based on radial basis function neural network

径向基函数神经网络具有学习速度较快,函数逼近能力强的特点.文章分析了影响声呐部位自噪声的各种声源参数,以舰艇声呐部位自噪声作为目标函数,将径向基函数神经网络用于舰艇声纳部位自噪声预报.利用舰艇声呐实测数据进行网络训练,训练好的神经网络可以对舰艇声呐部位自噪声进行精确预报.

A radial basis function neural network has excellent abili（y of rapid learning and approximation function. In this paper, all the acoustic source factors influencing the naval vessel sonar self-noise are analyzed. By taking naval vessel sonar self-noise as the object function, the radial basis function neural network is used to forecast naval vessel sonar self-noise. The actual data of naval vessel sonar are used to train the network, and then the trained neural network can forecast the naval vessel sonar self-noise accurately.