控制

基于模型参考和随机森林算法的船舶操纵运动辨识建模

  • 梅斌 ,
  • 孙立成 ,
  • 史国友
展开
  • (1.大连海事大学  a.航海学院; b.航海安全保障研究所,辽宁 大连 116026;2.中国船级社,北京 100007)
梅 斌(1991 - ),男,博士生,E-mail:meibindmu@163.com.

收稿日期: 2017-10-18

  修回日期: 2017-11-02

  网络出版日期: 2017-11-28

基金资助

国家自然科学基金资助项目(51579025);辽宁省自然科学基金资助项目(20170540090).

Modeling and identification of ship maneuvering motion based on model reference and stochastic forest algorithm

  • MEI Bin ,
  • SUN Li-cheng ,
  • SHI Guo-you
Expand
  • ( 1a.Navigation College;1b.Key Laboratory of Navigation Safety Guarantee of Liaoning Province, Dalian 116026, China; 2. China Classification Society, Beijing  100007, China )

Received date: 2017-10-18

  Revised date: 2017-11-02

  Online published: 2017-11-28

Supported by

 

摘要

采用模型参考和机器学习相结合的辨识建模结构对船舶操纵运动建模.首先,选择已公开模型作为参考模型;其次,使用相似准则把被辨识船舶速度转移到参考模型;最后,使用随机森林模型构建被辨识船舶加速度和参考模型加速度的映射关系.随机森林模型具有训练快、避免过拟合的优点.使用船模试验数据进行辨识建模和模型验证,并与MMG模型和BPNN进行对比.结果表明,该辨识方法具有较强的可行性、预报能力和泛化能力.

本文引用格式

梅斌 , 孙立成 , 史国友 . 基于模型参考和随机森林算法的船舶操纵运动辨识建模[J]. 大连海事大学学报, 2018 , 44(2) : 15 -21 . DOI: 10.16411/j.cnki.issn1006-7736.2018.02.003

Abstract

The modeling structure of model reference and machine learning was used for the ship maneuvering motion modeling. Firstly, the open model was selected as a reference model. Secondly, the similarity criterion was used to transfer the speed of the identified ship to the reference model. Finally, the random forest model was used to construct the mapping relationship between identified acceleration and reference model acceleration. The random forest model has the advantages of fast training and avoiding overfitting. The ship model test data was used for identification modeling and model verification, and compared with MMG model and BPNN. Results show that the proposed method has strong feasibility, prediction ability and generalization ability.

参考文献

[1]郭春雨, 李夏炎, 王帅, 等.冰区航行船舶碎冰阻力预报数值模拟方法[J].哈尔滨工程大学学报, 2016, 37(2):145-150
GUO Chun-yu, LI Xia-yan, WANG Shuai, et al.A numerical simulation method for resistance prediction of ship inpack ice[J].Journal of Harbin Enginering University, 2016, 37(2):145-150
[3]Abkowitz M A.Measurement of hydrodynamic characteristics from ship maneuvering trials by system identification[J]. Transactions of Society of Naval Architects and Marine Engineers[J].Transactions of Society of Naval Architects and Marine Engineers, 1980, 88():283-318
[4]Luo W L, Zou Z J.Identification of response models of ship maneuvering motion using support vector machines[J].Journal of Ship Mechanics, 2007, 11(6):832-838
[5].C K?llstr?m,?str?m, Karl Johan, Identification and Modelling of Ship Dynamics, volume 7015[R]:Lund(Sweden):Identification and Modelling of Ship Dynamics, 1972
[6]Yoon H K, Rhee K P.Identification of hydrodynamic coefficients in ship maneuvering equations of motion by Estimation-Before-Modeling technique[J].Ocean Engineering, 2003, 30(18):2379-2404
[7]Haddara M R, Wang Y.Parametric identification of manoeuvring models for ships[J].International Shipbuilding Progress, 1999, 46(445):5-27
[8]Wang N, Er M J, Han M.Large Tanker Motion Model Identification Using Generalized Ellipsoidal Basis Function-Based Fuzzy Neural Networks[C].// Design Automation Conference, 1988. Proceedings. ACM/IEEE. IEEE, 1988:205-210.
[9]Hess, D,Faller, W.Simulation of Ship Maneuvers Using Recursive Neural Networks[C].Twenty-Third Symposium on Naval Hydrodynamics, National Academies Press, 2000.
[10]周志华.机器学习(Machine learning)[M]. 清华大学出版社, 2016.
ZHOU Zhi-hua.Machine Learning[M]. Tsinghua University Press.2016. (in Chinese)
[12]张心光, 邹早建.支持向量回归机不敏感因子ε对运动建模的影响[J].中国造船, 2011, 52(3):46-51
ZHANG Xin-guang, ZOU Zao-jian.The Effect of SVR Insensitive Factor on Modeling of Motion[J].SHIPBUILDING OF CHINA, 2011, 52(3):46-51
[14]Kwiatkowska M, Atkins A S, Ayas N T, et al.Integrating Knowledge-Driven and Data-Driven Approaches for the Derivation of Clinical Prediction Rules[C]// International Conference on Machine Learning and Applications, 2005. Proceedings. IEEE, 2006:171-176.
[15]Wang N, Zhu M, Li J, et al.Data-driven vs. model-driven: Fast face sketch synthesis[J][J].Neurocomputing, 2017, 257():214-221
[16]Hayes M N.Parametric identification of nonlinear stochastic systems applied to ocean vehicle dynamics[D]. Massachusetts Institute of Technology, 1971.
[17]Sutulo S, Soares C G.Offline system identification of ship manoeuvring mathematical models with a global optimization algorithm[C]// Marsim 2015: International Conference on Ship Manoeuvrability and Maritime Simulation, Newcastle University, United Kingdom, 8–11 September. 2015.
[18]朱新坚, 邵惠鹤, 张钟俊.离散化对非线性控制系统动力学行为的影响[J].信息与控制, 1996, 25(2):115-119
Zhu Xin-jian, SHAO Huihe, ZHANG Zhong-jun.The effect of discrete analysis on the dynamic behavior of nonlinear control systems[J].Information and control, 1996, 25(2):115-119
[20]Zhang Z, An D U, Kim H, et al.Comparative study of Matrix exponential and Taylor series discretization methods for nonlinear ODEs[J].Simulation Modelling Practice & Theory, 2009, 17(2):471-484
[21]Kazantzis N, Kravaris C.Time-discretization of nonlinear control systems via Taylor methods[J].Computers & Chemical Engineering, 1999, 23(6):763-784
[22]SNAME.Nomenclature for treating the motion of a submerged body through a fluid[S]. T & R Bulletin, 1950.
[23]M.S. Chislett and J. Stroem-Tejsen (1965). Planar Motion Mechanism Tests and Full-Scale Steering and Maneuvering Predictions for a Mariner Class Vessel[R], Technical Report Hy-5, Hydro- and Aerodynamics Laboratory, Lyngby, Denmark.
[24]Leo Breiman.Random Forests[J].Machine Learning, 2001, 45(1):5-32
[25]Hastie T, Tibshirani R, Friedman J.The element of statistical learning:Data Mining, Inference, and Prediction[M].Springer, 2009.
文章导航

/