基于径向基函数网络的船舶非线性参数模型建模  

Modeling of ship's nonlinear parameters with radial basis function

作　　者：杨雪晶 赵希人[1] 

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001

出　　处：《哈尔滨工程大学学报》2006年第3期391-394,399,共5页Journal of Harbin Engineering University

基　　金：国防科学技术工业委员会基础研究基金资助项目(41314020201)

摘　　要：船舶在航行中受到各种随机扰动,为了实时控制船舶运动姿态,需建立一个基于航速、海情和航向角实时变化的三维智能化运动模型.文中以某水面舰艇为研究对象,利用典型航速、航向角、海情下水池实验测量的数据算得水动力参数,基于径向基函数神经网络(RBF)算法,逼近三维空间各水动力参数的非线性函数,从而得到随航速、海情和航向角自适应变化的船舶运动非线性参数模型.仿真结果表明,径向基函数神经网络学习算法可以快速、精确地建立非线性参数模型.所建模型误差小于2%,该模型可用于船舶控制,使LQG减横摇控制效果提高41.6%.A ship is affected by all kinds of random disturbances when navigating the ocean. To provide real-time control for a ship＇s motion positioning, a model is needed which responds in real-time to changes in navigating speed, ocean condition, and course. Using experimental data of these factors, research based on radial basis function was done on hydrodynamic parameters＂ nonlinear function of three-dimensional space, resulting in a nonlinear parameter model which can self-adapt to changes in navigating speed, ocean condition, and course. Simulation results indicate that the modeling method with radial basis function is quick and accurate, with an error rate below 2%. This model can be used in ship＇s motion control. It is indicated that the control effect of a linear quadratic Gaussian （LQG） method, which is used to reduce roll motion, is increased by 41.6%.

关 键 词：径向基函数网络 水动力参数 非线性参数模型 船舶横向运动 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]