基于多层感知机的船舶运动及作业率预报
The Prediction of the Motion and Operation Capability of Ship Based on Multilayer Perceptron
戈亮 1张文旭 2李亚军3
作者信息
- 1. 海军装备部装备项目管理中心,北京 100071
- 2. 中国舰船研究设计中心,武汉 430060
- 3. 江苏科技大学船舶与海洋工程学院,镇江 212100
- 折叠
摘要
以传统耐波性方法计算的典型航速和浪向对应的船舶运动结果为训练集,以全航速和浪向的舰船运动数据为测试集,以多层感知机为训练模型构建了一种船舶运动和作业率预报方法.采用基于海况和运动模态的归一化对数据进行预处理,采用正则化、学习率调度器及Adam法对模型进行训练.将预报结果与耐波性软件计算结果进行比较,验证了预报方法的有效性.研究表明:典型浪向个数对预报舰船运动及舰船作业率的精度有重要影响,典型航速个数的影响则相对较小.
Abstract
A method based on the multilayer perceptron to predict ship motion and operation capability is proposed in this paper.The training dataset is constructed by ship motion data of typical speeds and wave angles,and test dataset con-sists of ship motion data of full speeds and wave angles.The normalization based on sea state and motion mode is applied to dataset,and regularization,learning rate scheduler is used to train the model.The method is validated by the comparison between the predicted ship motion and the results calculated by seakeeping software.It is found that the number of typical wave angle has import effect on the prediction precision of ship motion and operation capability,while the number of typi-cal speeds has less effect on this.
关键词
舰船作业率/多层感知机/舰船运动Key words
ship operation capability/multilayer perceptron/ship motion引用本文复制引用
出版年
2024
NETL
NSTL
万方数据