基于神经网络的围网渔船横摇运动研究

为了快速预报一艘65 m的围网渔船在任意载况规则波中横摇运动频率响应函数及不规则波中各海况下的横摇角有义值,应用神经网络的原理和算法,以该船7种实际航运载况的耐波性计算数据为训练样本,建立满足精度要求的围网渔船横摇运动神经网络预估模型。并分析吃水、重心高度、航速和浪向对横摇运动的影响。该模型可应用于风浪中船舶航行安全性评估。

Research of Tuna Seiner's Roll Motion Based on Neural Network

In order to quickly predict the frequency response function of roll motion for a 65m tuna seiner＇s in regular waves and the significant values of roll motion in irregular waves under different sea conditions,making the sea-keeping calculation database of the tuna seiner＇s under 7 real load condition as training samples,a neural network predict model was established based on neural network theory and algorithms which can meet the accuracy requirements for tuna seiner＇s roll motion prediction.Then the influences of the draught,weight center height,speed and wave direction on roll motion were analyzed.It can be applied to ship sailing safety assessment in real sea.