关键词: 水下拖曳体, 联合控制, 水动力学, LMBP 算法, 神经网络

Abstract:

In this paper , a self-stable controllable underwater towed vehicle was taken as the research object , and the 3D hydrodynamic response perlormances of the vehicle with a constant velocity under the joint manipulations of a depressing wing and a twin-stern thruster were discussed. In the investigation , a numerical hydrodynamic model was established based on the LMBP algorithm of neural network theory. The experimental data for training the pro-posed model were provided from those of 2D vehicle prototype towing experiments conducted in a towing tank under the manipulation of single control mechanism such as a depressing wing or a twin-stern thruster , respectively. The 3 D hydrodynamic perlormances of the vehicle under the joint manipulations of the two above-mentioned control mechanisms were then analyzed based on the proposed model. The results of numerical simulation indicate that , by using the proposed model , the 3 D hydrodynamic perlo口nances of the towed vehicle under the joint manipulations of multiple controls can be effectively simulated.

Key words: underwater towed vehicle, joint manipulation, hydrodynamics, LMBP algorithm, neural network