基于人工神经网络的自航浮标测波方法可行性

Feasibility of Wave Measurement by Using a Sailing Buoy and the Artificial Neural Network Technique

海洋波浪会对船舶营运和安全产生不利影响,实时准确的随船海浪监测对绿色智能船舶及航行安全至关重要.利用船舶运动反演遭遇海浪信息是一种重要的随船海浪监测手段,具有成本低、时空分辨率高等诸多优点,受到广泛关注.针对现有船舶运动反演海浪模型存在的不足,提出了一种基于人工神经网络(Artificial Neural Network, ANN)的反演模型.以船舶摇荡运动时历作为输入,以波浪时历作为输出,利用人工神经网络进行船舶运动特征提取,输入线性函数进行波面时历反演.为验证反演模型的可行性和精度,开展了船模水池试验.结果表明,提出的基于人工神经网络的测波方法能够很好地实现规则波和不规则波浪时历反演,规则波反演统计误差大多小于10%,不规则波反演误差在10%左右.提出的方法为船舶运动反演海浪时域信息提供了一种有效可行的手段.

Ocean waves have adverse effects on the operation and safety of ships. Real-time and accurate onboard wave monitoring is essential for green smart ships and navigational safety. It is an important method to use ship motion to perform inversion of the information of encountering waves, which has many advantages such as low cost and high spatio-temporal resolution, and has attracted wide attention. An inversion model based on artificial neural network(ANN) is proposed to deal with the shortcomings of existing ship motion inversion ocean wave models. Taking the motion time history of the ship as the input and the wave elevation time history as the output, the artificial neural network is used to extract the motion features of the ship, and the linear function is input to perform wave surface time history inversion. To verify the feasibility and the accuracy of the inversion model, a ship model tank test is conducted. The results show that the proposed wave measurement method based on the artificial neural network can well achieve regular wave and irregular wave elevation time history inversion. Most of the statistical errors of regular wave inversion are less than 10% and the error of irregular wave inversion is about 10%. The method proposed provides an effective and feasible means for inversion of wave time information in ship motion.