Identification modeling of underwater vehicles' nonlinear dynamics based on support vector machines

An identification method based on support vector machines (SVM) is proposed for modeling nonlinear dynamics of underwater vehicles (UVs), and a typical torpedo-shaped autonomous underwater vehicle (AUV) is employed for the purpose of validation. To obtain the hydrodynamic derivatives of the vehicle and the dynamical models of the thruster and fins, a series of hydrodynamic experiments are conducted by using vertical planar motion mechanism (PMM) and circulating water channel (CWC). Maneuvering simulation is carried out by using the hydrodynamic model obtained from experiments, and SVM is applied to identify the damping terms together with Coriolis and centripetal terms by analyzing the simulation data. By using the identified nonlinear model and experiment-based hydrodynamic model respectively, maneuvering simulations and control applications are implemented. The results are compared to verify the proposed method, and the effectiveness and good generalization performance of SVM in modeling the nonlinear dynamics of UVs are demonstrated.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01489737
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jul 24 2013 2:07PM