航行速度的非线性建模与仿真研究
|
摘要
舰船航行速度具有十分强烈的非线性变化特点,而当前舰船航行速度建模方法均只考虑其线性特性,使得舰船航行速度预测错误很大。为了提高舰船航行速度预测精度,提出一种航行速度预测的非线性建模方法。首先采用舰船航行速度的历史样本数据,建立舰船航行速度预测的训练样本集合,然后引入回声状态网络对舰船航行速度训练样本的变化规律进行描述,建立舰船航行速度预测模型,最后采用具体舰船航行速度数据对非线性建模性能进行测试。本文建模方法可以捕捉舰船航行速度强烈的非线性变化特点,舰船航行速度预测错误小,舰船航行速度预测精度要小于当前线性建模方法,而且降低了舰船航行速度建模的时间,具有比较显著的优越性。
Velocity of a ship has very strong non-linear characteristics. Currently, the modeling methods of navigational speed only consider its linear characteristics, which makes the prediction errors of navigational speed of a ship very large.In order to improve the prediction accuracy of navigational speed of a ship, a non-linear modeling method of navigational speed prediction is proposed. Firstly, using the historical sample data of navigational speed, the training sample set of navigational speed prediction is established. Then, the echo state network is introduced to describe the change rule of navigational speed training sample, and the prediction model of navigational speed is established. Finally, the performance of non-linear modeling is tested with specific naval speed data. The modeling method in this paper can capture the strong non-linear characteristics of the ship's navigation speed. The prediction error of the ship's navigation speed is small, and the prediction accuracy of the ship's navigation speed is less than that of the current linear modeling method. It also reduces the time complexity of the ship's navigation speed modeling and has obvious advantages.
Velocity of a ship has very strong non-linear characteristics. Currently, the modeling methods of navigational speed only consider its linear characteristics, which makes the prediction errors of navigational speed of a ship very large.In order to improve the prediction accuracy of navigational speed of a ship, a non-linear modeling method of navigational speed prediction is proposed. Firstly, using the historical sample data of navigational speed, the training sample set of navigational speed prediction is established. Then, the echo state network is introduced to describe the change rule of navigational speed training sample, and the prediction model of navigational speed is established. Finally, the performance of non-linear modeling is tested with specific naval speed data. The modeling method in this paper can capture the strong non-linear characteristics of the ship's navigation speed. The prediction error of the ship's navigation speed is small, and the prediction accuracy of the ship's navigation speed is less than that of the current linear modeling method. It also reduces the time complexity of the ship's navigation speed modeling and has obvious advantages.
引文
[1]李俊,赵豪,赵涛.水下航行器速度和姿态匹配的动基座快速对准滤波方法[J].鱼雷技术, 2008, 12(2):13–16.
[2]镇璐,孙晓凡,王帅安.排放控制区限制下邮轮航线及速度优化[J].运筹与管理, 2019, 28(3):31–38.
[3]段立金.水流阻力对浅水狭航道船舶航行速度的影响分析[J].舰船科学技术, 2018, 40(6):13–15.
[4]李智生,赵欣.航行体不同速度出筒尾部流场及受载特性分析[J].舰船电子工程, 2017, 37(8):66–69, 156.
[2]镇璐,孙晓凡,王帅安.排放控制区限制下邮轮航线及速度优化[J].运筹与管理, 2019, 28(3):31–38.
[3]段立金.水流阻力对浅水狭航道船舶航行速度的影响分析[J].舰船科学技术, 2018, 40(6):13–15.
[4]李智生,赵欣.航行体不同速度出筒尾部流场及受载特性分析[J].舰船电子工程, 2017, 37(8):66–69, 156.