基于径向基神经网络的水下振动物体辐射噪声级别分类研究

运用径向基神经网络，利用水下振动物体内表面加速度信号对其辐射噪声级别进行分类．达到判断其声隐身性的目的，该方法的运算量较传统方法大大降低，极大地提高了计算速度。实例表明，该方法能较准确地对水下振动物体辐射声场声压级别进行分类。进而对其推广应用于潜艇提供了较好的依据。

On classification of underwater vibrating object sound radiation level based on RBF neural network

By the use of RBF neural network, acceleration signals on the inside surface of the underwater vibrating object are used to classify its sound radiation level in order to measure its acoustic stealth. The method has much less complexity of computation than the traditional method, and the computing speed is improved a lot. The example shows that this method has good ability of classifying the underwater vibrating object sound radiation level, and also provides a nice basis on the possibility to use this method on submarine.