基于神经网络的舰船水平总振动频率计算方法

Calculation method of ship's horizontal total vibration frequency based on neural network

舰船水平总振动频率具有较强的波动性,当前线性方法无法描述舰船水平总振动频率的变化特点,导致舰船水平总振动频率计算精度低,为了提高舰船水平总振动频率的计算精度,提出了基于神经网络舰船水平总振动频率计算方法。采用BP神经网络强大的非线性学习能力对舰船水平总振动频率变化特点进行高精度的逼近,并对BP神经网络的参数优化问题进行研究,最后进行了舰船水平总振动频率计算的模拟实验,结果表明,BP神经网络解决了当前舰船水平总振动频率计算误差大的难题,获得了高精度的舰船水平总振动频率计算结果。

The total vibration frequency of ship level has strong fluctuation. The current linear method can not describe the variation characteristic of the total vibration frequency of ship level, which leads to the low calculation precision of the total vibration frequency of the ship level. In order to improve the calculation precision of the total horizontal vibration frequency of the ship, the total vibration frequency of the ship level based on the neural network is proposed. Using the strong nonlinear learning ability of BP neural network, a high precision approach to the variation of the ship's horizontal total vibration frequency is made, and the parameter optimization of the BP neural network is studied. Finally, the simulation experiment of the total vibration frequency of the ship is carried out. The result shows that the current ship water is solved by the BP neural network. The calculation error of the total total vibration frequency is large, and the high accuracy calculation results of the ship's horizontal total vibration frequency are obtained.