J Shanghai Jiaotong Univ Sci ›› 2022, Vol. 27 ›› Issue (3): 402-410.doi: 10.1007/s12204-021-2284-6

• • 上一篇    下一篇

  

  • 收稿日期:2019-04-23 出版日期:2022-05-28 发布日期:2022-06-23

Submarine Multi-Model Switching Control Under Full Working Condition Based on Machine Learning

LIANG Liang1 (梁 良), SHI Ying1∗ (石 英), MOU Junmin2∗ (牟军敏)   

  1. (1. School of Automation, Wuhan University of Technology, Wuhan 430070, China; 2. Hubei Key Laboratory of Inland Shipping Technology; School of Navigation, Wuhan University of Technology, Wuhan 430063, China)
  • Received:2019-04-23 Online:2022-05-28 Published:2022-06-23

Abstract: A continuous submarine depth control strategy based on multi-model and machine learning switching method under full working condition is proposed in this paper. A submarine motion model with six-degree-of freedom is first built and decoupled according to the force analysis. The control set with corresponding precise model set is then optimized according to different working conditions. The multi-model switching strategy is studied using machine learning algorithm. The simulation experiments indicate that a multi-model controller comprised of the proportional-integral-derivative (PID), fuzzy PID (FPID) and model predictive controllers with support vector machine (SVM) switching strategy can realize the continuous submarine depth control under full working condition, showing a good control performance compared with a single PID controller.

Key words: multi-model switching control, machine learning, model decoupling, submarine

中图分类号: