[1]张彦峰,杨震,王立鹏,等.基于LSTM的船舶运动多模态预测方法[J].智能系统学报,2026,21(1):201-213.[doi:10.11992/tis.202512005]
ZHANG Yanfeng,YANG Zhen,WANG Lipeng,et al.Multi-modal prediction method for ship motion based on LSTM[J].CAAI Transactions on Intelligent Systems,2026,21(1):201-213.[doi:10.11992/tis.202512005]
点击复制
ZHANG Yanfeng,YANG Zhen,WANG Lipeng,et al.Multi-modal prediction method for ship motion based on LSTM[J].CAAI Transactions on Intelligent Systems,2026,21(1):201-213.[doi:10.11992/tis.202512005]
基于LSTM的船舶运动多模态预测方法
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
21
期数:
2026年第1期
页码:
201-213
栏目:
学术论文—水下智能与海洋具身智能
出版日期:
2026-03-05
- Title:
- Multi-modal prediction method for ship motion based on LSTM
- Keywords:
- ship; roll; pitch; time series; short-term prediction; long short-term memory; double-layer modal decomposition; Kepler optimization
- 分类号:
- TP183
- 摘要:
- 为提升长短期记忆网络(long short-term memory, LSTM)在船舶运动姿态领域的预测精度，同时简化参数调优的过程。针对船舶运动数据设计了双层模态分解策略，结合开普勒优化算法(Kepler optimization algorithm, KOA)，提出一种船舶运动姿态多模态预测模型。采用改进的完全自适应噪声集合经验模态分解(improved complete ensemble empirical mode decomposition with adaptive noise, ICEEMDAN)和变分模态分解(variational mode decomposition, VMD)的双层模态分解(double-layer mode decomposition, DLMD)模式，对原始船舶运动姿态时序数据进行双层分解，解决了第一层分解存在的高频分量过度平滑问题，同时消除了高频噪声与有效信号在相同频带下的耦合干扰，提升了模型的预测效果；进一步在模型训练中引入KOA优化LSTM的超参数，解决了LSTM调参效率低、易陷入局部最优的问题。基于实船运动数据集开展消融实验和算法整体验证实验，消融实验验证了DLMD和KOA模块的独立贡献；算法整体验证实验结果验证了两个模块的共同作用，并表明了该模型能够对船舶横摇和纵摇姿态实现较高精度的预测及超参数组合的自动优化。
- Abstract:
- To improve the prediction accuracy of the long short-term memory (LSTM) model in the field of ship motion attitude and simplify the parameter tuning process, a double-layer mode decomposition strategy is designed for ship motion data. Combined with the Kepler optimization algorithm (KOA), a multi-modal prediction model for ship motion attitude is proposed. This model adopts a double-layer mode decomposition (DLMD) scheme integrating improved complete ensemble empirical mode decomposition with adaptive noise (ICEEMDAN) and variational mode decomposition (VMD) to perform double-layer decomposition on the original time-series data of ship motion attitude. This decomposition resolves the over-smoothing issue of high-frequency components in the first-layer decomposition, while eliminating the coupling interference between high-frequency noise and effective signals in the same frequency band, thus improving the prediction performance of the model. Furthermore, KOA is introduced to optimize the hyperparameters of LSTM during model training, which addresses the problems of low efficiency and easy trapping in local optima in LSTM parameter tuning. Ablation experiments and overall algorithm verification experiments are conducted based on a real ship motion dataset. The ablation experiments verify the independent contributions of the DLMD and KOA modules; the results of the algorithm verification experiments confirm the synergistic effect of the two modules, and demonstrate that the proposed model can achieve high-precision prediction of ship roll and pitch attitudes as well as automatic optimization of hyperparameter combinations.
- 参考文献/References:
-
[1] 陈航宇. 波浪中船舶运动极短期预报深度学习方法泛化性研究[D]. 哈尔滨工程大学, 2023. CHEN Hangyu. Research on the generalization of deep learning methods for ultra-short-term prediction of ship motion in waves[D]. Harbin Engineering University, 2023.
[2] NAAIJEN P, Van OOSTEN K, ROOZEN K, et al. Validation of a deterministic wave and ship motion prediction system[C]//International Conference on Offshore Mechanics and Arctic Engineering. San Diego: American Society of Mechanical Engineers, 2018: 51272.
[3] ZHANG Daiyong, CHU Xiumin, LIU Chenguang, et al. A review on motion prediction for intelligent ship navigation[J]. Journal of marine science and engineering, 2024, 12(1): 107
[4] ZHU Dexiang, KATORY M. A time-domain prediction method of ship motions[J]. Ocean engineering, 1998, 25(9): 781-791
[5] 安瑾. 基于LSTM的船舶轨迹预测[J]. 中国储运, 2025(8): 121 AN Jin. Ship trajectory prediction based on LSTM[J]. China storage & transport, 2025(8): 121
[6] LI Guoyuan, ZHANG Houxiang, KAWAN B, et al. Analysis and modeling of sensor data for ship motion prediction[C]//OCEANS 2016. Shanghai: IEEE, 2016: 1-7.
[7] CHENG Xu, LI Guoyuan, SKULSTAD R, et al. Data-driven uncertainty and sensitivity analysis for ship motion modeling in offshore operations[J]. Ocean engineering, 2019, 179: 261-272
[8] LI Mingwei, GENG Jing, HAN Duanfeng, et al. Ship motion prediction using dynamic seasonal RvSVR with phase space reconstruction and the chaos adaptive efficient FOA[J]. Neurocomputing, 2016, 174: 661-680
[9] SHEN Wenhe, HU Xinjue, LIU Jialun, et al. A pre-trained multi-step prediction informer for ship motion prediction with a mechanism-data dual-driven framework[J]. Engineering applications of artificial intelligence, 2025, 139: 109523
[10] 谷达京, 施哲源, 陈根良, 等. 基于神经网络的船体运动位姿预测方法[J]. 舰船科学技术, 2022, 44(15): 55-59 GU Dajing, SHI Zheyuan, CHEN Genliang, et al. A pose prediction method for ship hull motion based on neural network[J]. Ship science and technology, 2022, 44(15): 55-59
[11] ROBERT S, LI Guoyuan, THOR I, et al. A hybrid approach to motion prediction for ship docking-integration of a neural network model into the ship dynamic model[J]. IEEE transactions on instrumentation and measurement, 2020, 70: 1-11
[12] HU Xiong, ZHANG Boyi, TANG Gang. Research on ship motion prediction algorithm based on dual-pass long short-term memory neural network[J]. IEEE access, 2021, 9: 28429-28438
[13] 张宇浩. 基于长短期记忆网络的船舶姿态预测研究[D]. 南京航空航天大学, 2023. ZHANG Yuhao. Research on ship attitude prediction based on long short-term memory network [D]. Nanjing University of Aeronautics and Astronautics, 2023.
[14] ZHANG Tao, ZHENG Xiaoqing, LIU Mingxin. Multiscale attention-based LSTM for ship motion prediction[J]. Ocean engineering, 2021, 230: 109066
[15] SUN Qian, TANG Zhong, GAO Jingpeng, et al. Short-term ship motion attitude prediction based on LSTM and GPR[J]. Applied ocean research, 2022, 118: 102927
[16] GENG Xiaoyu, LI Yibing, SUN Qian. A novel short-term ship motion prediction algorithm based on EMD and adaptive PSO-LSTM with the sliding window approach[J]. Journal of marine science and engineering, 2023, 11(3): 466
[17] LIU Yucheng, DUAN Wenyang, HUANG Limin, et al. The input vector space optimization for LSTM deep learning model in real-time prediction of ship motions[J]. Ocean engineering, 2020, 213: 107681
[18] YAO Yuxin, HAN Liang, WANG Jiangyun. LSTM-PSO: long short-term memory ship motion prediction based on particle swarm optimization[C]//2018 IEEE CSAA Guidance, Navigation and Control Conference. Xia’men: IEEE, 2018.
[19] JIANG Yan, HOU Xianrui, WANG Xuegang, et al. Identification modeling and prediction of ship maneuvering motion based on LSTM deep neural network[J]. Journal of marine science and technology, 2022, 27(1): 125-137
[20] ZHANG Guoyin, TAN Feng, WU Yanxia. Ship motion attitude prediction based on an adaptive dynamic particle swarm optimization algorithm and bidirectional LSTM neural network[J]. IEEE Access, 2020, 8: 90087-90098
[21] HAN Chongyang, HU Xiong. A prediction method of ship motion based on LSTM neural network with variable step-variable sampling frequency characteristics[J]. Journal of marine science and engineering, 2023, 11(5): 919
[22] WANG Ju, ZHANG Leifeng, YANG Sanqiang, et al. Optimized LSTM based on improved whale algorithm for surface subsidence deformation prediction[J]. Electronic research archive, 2023, 31(6): 3435-3452
[23] CHOE T H, HO C S. An improvement of PM2.5 concentration prediction using optimised deep LSTM[J]. International journal of environment and pollution, 2021, 69(3-4): 249-260
[24] SHAO Bilin, SONG Dan, BIAN Genqing, et al. Wind speed forecast based on the LSTM neural network optimized by the firework algorithm[J]. Advances in materials science and engineering, 2021, 2021(1): 4874757
[25] MA Zhiqiang, CAI Hongyang. Research on electricity load forecasting based on EEMD-GA-LSTM[C]//2024 36th Chinese Control and Decision Conference. Xi’an: IEEE, 2024: 4434-4439.
[26] GUNASEKAR S, KUMAR G J R, KUMAR Y D. Sustainable optimized LSTM-based intelligent system for air quality prediction in Chennai[J]. Acta geophysica, 2022, 70(6): 2889-2899
[27] 鲜思渔, 赵泽田, 吴轩宇, 等. 基于改进红鸢优化算法与LSTM的核电换热器寿命预测方法[J]. 图学学报, 2025, 46(5):1085-1093. XIAN Siyu, ZHAO Zetian, WU Xuanyu, et al. A life prediction method for nuclear power heat exchangers based on improved red kite optimization algorithm and LSTM [J]. Journal of graphics, 2025, 46(5):1085-1093.
[28] WANG Wenbo, TONG Meng, YU Min. Blood glucose prediction with VMD and LSTM optimized by improved particle swarm optimization[J]. IEEE Access, 2020, 8: 217908-217916
[29] SUN Hongbin, CUI Qing, WEN Jingya, et al. Short-term wind power prediction method based on CEEMDAN-GWO-Bi-LSTM[J]. Energy reports, 2024, 11: 1487-1502
[30] SEPP H, JURGEN S. Long short-term memory[J]. Neural computation, 1997, 9(8): 1735-1780
[31] 王玲桃, 任宏伟, 王紫瑜, 等. 基于ICEEMDAN和改进小波阈值的输电线路故障行波信号降噪[J]. 电子设计工程, 2025, 33(12): 56-61 WANG Lingtao, REN Hongwei, WANG Ziyu, et al. Denoising of fault traveling wave signals of transmission lines based on ICEEMDAN and improved wavelet threshold[J]. Electronic design engineering, 2025, 33(12): 56-61
[32] 刘昕明, 吉建光, 李玮, 等. 基于双模态分解的发电站母线短期负荷预测[J]. 电气工程学报, 2024, 19(1): 124-132 LIU Xinming, JI Jianguang, LI Wei, et al. Short-term load prediction of power station busbars based on double-modal decomposition[J]. Transactions of China electrotechnical society, 2024, 19(1): 124-132
[33] MOHAMED A B, REDA M, SHAIMAA A. Kepler optimization algorithm: a new metaheuristic algorithm inspired by Kepler’s laws of planetary motion[J]. Knowledge-based systems, 2023, 268: 110454
备注/Memo
收稿日期:2025-12-2。
基金项目:国家自然科学基金面上项目(62173103，62573151，52171332)；黑龙江省自然科学基金项目(LH2024F037)；中央高校基本科研业务费专项(3072024XX0403).
作者简介:张彦峰，硕士研究生，主要研究方向为船舶状态预报。E-mail：zhangyf_gl@hrbeu.edu.cn。;杨震，讲师，主要研究方向为船舶运动控制和智能预报，获省部级科技进步一等奖。发表学术论文20余篇。E-mail：yzhen@hrbeu.edu.cn。;王立鹏，副教授，博士生导师，主要研究方向为复杂系统建模与控制，主持国家自然科学基金面上项目、青年项目、民品横向等，获省部级科技进步特等奖、一等奖，获发明专利授权9项。发表学术论文30余篇。E-mail：wlp_heu@163.com。
通讯作者:杨震. E-mail：yzhen@hrbeu.edu.cn
更新日期/Last Update:
2026-01-05