一种基于改进回声状态网络的船舶摇荡预报

回声状态网络在船舶摇荡运动姿态的预报中得到了广泛的应用,由于远洋风浪环境的复杂性,使得船舶摇荡的非线性程度进一步增加,因此传统回声状态网络的预测精度明显不足。本文利用改进的回声状态网络方法,引入小波函数对数据进行预处理,对船舶的实测摇荡时历进行处理,提高了预报的精度。实验结果表明,与传统回声状态网络方法相比,本方法的预报精度和可靠性得到了提高。

An approach of ship sway forecasting based on improved ESNs

Echo state network has been widely used in the prediction of ship motion attitude. Because of the complexity of the ocean waves environment, the nonlinear characteristics of ship swing further increase, therefore the forecasting accuracy of tradition ESNs cannot meet the requirements. Improved ESN by wavelet function, processed the ship sway data, therefore miproved the accuracy of motion forecasting. The result shows that compared with the traditional echo state network method, this method could improve the accuracy and reliability of forecasting.