本文为在线建模与预报随机波浪中的船舶运动，将时序卷积神经网络和采用经验回放功能的增量学习机制结合，并对训练及回放数据进行稀疏化，建立能够挖掘时序特征并快速更新的网络结构。以4～6级海况下横浪中船舶横摇运动、迎浪中船舶纵摇及垂荡运动为测试算例，验证模型的有效性。相对于时序卷积神经网络、增量时序卷积神经网络、稀疏增量长短期记忆神经网络与稀疏增量Transformer模型，本模型在相同数据量上有更高的精度和效率，并随着增量阶段推进预报精度更稳定，在不同海况下具有泛用性。
To enable online modeling and prediction of ship motions in stochastic waves, a temporal convolutional neural network is combined with an incremental learning mechanism incorporating experience replay. Sparse selection is applied to both training and replay data, so as to establish a network structure capable of mining temporal features while enabling rapid updates. The effectiveness of the model is validated using test cases of ship roll motion in beam waves under sea states 4, 5, and 6, as well as ship pitch and heave motions in head waves. Compared to temporal convolutional network, incremental temporal convolutional network, sparse incremental long short-term memory networks and sparse incremental Transformer model, the proposed model achieves higher accuracy and efficiency with the same data volume and maintains more stable prediction accuracy as the incremental phase progresses. Furthermore, the method is effective for ship under varying sea states.
2026,48(8): 36-43 收稿日期：2025-9-9
DOI：10.3404/j.issn.1672-7649.2026.08.006
分类号：U675.5
基金项目：国家自然科学基金资助项目（52571363, U2441289）
作者简介：王超(2001-),男,硕士研究生,研究方向为船舶运动预报
参考文献：
[1] LI X, LV X, YU J, et al. Neural network application on ship motion prediction[C]// Proceedings of the 9th International Conference on Intelligent Human-Machine Systems and Cybernetics, 2017: 414-417.
[2] HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780
[3] 易文海, 高志亮. 基于LSTM神经网络的随机横浪中船舶横摇运动极短期预报[J]. 武汉理工大学学报(交通科学与工程版), 2021, 45(6): 1113-1117 YI W H, GAO Z L. Very short-term prediction of ship rolling motion in random transverse waves based on LSTM neural network[J]. Journal of Wuhan University of Technology(Transportation Science & Engineering), 2021, 45(6): 1113-1117
[4] WANG Y, WANG H, ZHOU B, et al. Multi-dimensional prediction method based on Bi-LSTMC for ship roll[J]. Ocean Engineering, 2021, 242: 110106
[5] WANG Y, WANG H, ZOU D, et al. Ship roll prediction algorithm based on Bi-LSTM-TPA combined model[J]. Journal of Marine Science and Engineering, 2021, 9(4): 387
[6] XUE J, GAO Z, YI W. An APSO-based TPA-BLSTM model for predicting ship motion in irregular waves using wave-series Input[J]. Ships and Offshore Structures, 2025, 20(3): 372-383
[7] LEA C, VIDAL R, REITER A, et al. Temporal convolutional networks: a unified approach to action segmentation[C]// Proceedings of the 14th European Conference on Computer Vision, 2016: 47–54.
[8] ZHANG B, WANG S, DENG L, et al. Ship motion attitude prediction model based on IWOA-TCN-Attention[J]. Ocean Engineering, 2023, 272: 113911
[9] 张东阳, 陆子轩, 刘军民, 等. 深度模型的持续学习综述: 理论、方法和应用[J]. 电子与信息学报, 2024, 46(10): 3849-3878 ZHANG D Y, LU Z X, LIU J M. A survey of continual learning with deep networks: theory, method and application[J]. Journal of Electronics & Information Technology, 2024, 46(10): 3849-3878
[10] KIRKPATRICK J, PASCANU R, RABINOWITZ N, et al. Overcoming catastrophic forgetting in neural networks[J]. Proceedings of the National Academy of Sciences, 2017, 114(13): 3521-3526
[11] ROLNICK D, AHUJA A, SCHWARZ J, et al. Experience replay for continual learning[C]// Proceedings of the 33rd International Conference on Neural Information Processing Systems, 2019: 350-360.
[12] CHAUDHRY A, RANZATO M A, ROHRBACH M, et al. Efficient lifelong learning with A-GEM[C]// Proceedings of the 7th International Conference on Learning Representations, 2019.
[13] YAN S, XIE J, HE X. Der: dynamically expandable representation for class incremental learning[C]// Proceedings of the Conference on Computer Vision and Pattern Recognition, 2021: 3014-3023.
[14] WEHBE B, FABISCH A, KRELL M M. Online model identification for underwater vehicles through incremental support vector regression[C]// Proceedings of the International Conference on Intelligent Robots and Systems, 2017: 4173-4180.
[15] 陈东洋. 面向船舶性能预报的增量学习方法研究[D]. 无锡: 江南大学, 2024.
[16] DANG Y, HUANG Y, SHEN X, et al. Incremental sparse gaussian process-based model predictive control for trajectory tracking of unmanned underwater vehicles[J]. IEEE Robotics and Automation Letters, 2025, 10(3): 2327-2334
[17] 易文海. 基于LSTM神经网络的随机海浪中船舶运动极短期预报方法研究[D]. 武汉: 武汉理工大学, 2021.
