应用神经网络建立水下拖曳体轨迹姿态水动力控制模型  

A Hydrodynamic and Control Model for an Underwater Towed Vehicle Based on Neural Network Theory

作　　者：熊小辉[1] 吴家鸣[1] 李志印[1] 

机构地区：[1]华南理工大学交通学院,广东广州510640

出　　处：《湛江海洋大学学报》2005年第3期67-72,共6页Journal of Zhanjiang Ocean University

基　　金：国家自然科学基金资助项目(40276034);教育部留学回国人员科研启动基金资助项目

摘　　要：以拖曳体的拖曳水池样机试验取得试验数据作为训练样本,采用LMBP算法,建立基于神经网络理论构建的可控制水下拖曳体轨迹与姿态水动力控制数值模型,并进行LMBP模型仿真检验。结果表明,利用所建全的神经网络模型对拖曳体在一定控制动作下的水动力响应预报是令人满意的。A hydrodynamic neural network model for a trajectory and attitude controllable underwater towed vehicle is established based on LMBP algorithm. The training samples are provided from testing data of the towed vehicle prototype towing experiments conducted in a towing tank and the model can then be established. The numerical simulation results in the paper indicate that forecast of the hydrodynamic and control performances of the underwater towed vehicle under a given control manipulation by the established neural network model is satisfactory.

关 键 词：水下拖曳体 神经网络 LMBP算法 水动力学 

分 类 号：TV131.2[水利工程—水力学及河流动力学] P715.5[天文地球—海洋科学]