摘要
针对船舶姿态预报精度的难题,结合船舶姿态变化特点,提出基于改进神经网络方法的船舶姿态高精度预报模型,首先对船舶姿态的数据进行采集,并对船舶姿态数据进行去噪处理,然后采用神经网络对船舶姿态变化特点进行高精度逼近,并对神经网络存在的一些缺陷进行相应的改进,最后进行船舶姿态预报的仿真实验。实验结果表明,改进神经网络提高了船舶姿态预报精度,克服了当前其它船舶姿态预报模型存在误差大的弊端,船舶姿态效果优势十分明显。
In view of the difficult problem of ship attitude prediction accuracy and the characteristics of ship attitude change, a high precision prediction model of ship attitude based on improved neural network is proposed. First, the data of ship attitude is collected, and the ship attitude data is de-noised, and the neural network is used to change the ship attitude.The high accuracy approximation is carried out, and some defects of the neural network are improved. Finally, the simulation experiment of ship attitude prediction is carried out. The experimental results show that the improved neural network improves the accuracy of the ship's attitude prediction and overcomes the disadvantages of the large error in the other ship's attitude prediction models, and the advantages of the ship attitude effect are very obvious.
出处
《舰船科学技术》
北大核心
2018年第9X期37-39,共3页
Ship Science and Technology
基金
2015年民办教育专项资金高职院校专业建设项目项目(计算机应用技术重点专业建设)
关键词
船舶姿态
预报模型
神经网络
自适应遗传算法
ship attitude
prediction model
neural network
adaptive genetic algorithm