船舶航向的神经网络并行鲁棒模型参考控制(英文)  

Neural Network-based Parallel Robust Model Reference Control Method of Ship Course

作　　者：吕进[1] 郭晨[1] 

机构地区：[1]大连海事大学自动化与电气工程学院,辽宁大连116026

出　　处：《系统仿真学报》2007年第15期3489-3493,共5页Journal of System Simulation

基　　金：This work was supported in part by the National Natural Science Foundation of China (60474014);in part by the Specialized Research Fund for the Doctoral Program of Higher Education of P.R.C. under grant (20040151007);in part by the Application foundation Research Project of Ministry of Communications of P.R.C. under grant (200432922504)

摘　　要：针对大型船舶的航向控制特性,提出一种神经网络并行自学习鲁棒模型参考控制方法。这种复合控制结构利用神经网络的学习能力和非线性映射能力,解决传统自适应控制中模型的在线辨识和控制器的在线设计问题,以达到对不确定非线性船舶的高精度输出跟踪控制;通过引入运行监控器,克服神经网络控制方法实时性差的问题;利用一个鲁棒反馈控制器,来保证神经网络模型学习初期闭环系统的稳定性。仿真结果表明这一方法对设定航向具有精确的跟踪控制效果。Aiming at the control feature of large ship, a neural network based parallel self-learning robust model-reference control method of ship＇s course was proposed. This compounded control structure using the self-learning and nonlinear map-capability of neural networks, solved problems of online model-identification and online design of the controller in traditional adaptive control, so that the high precise output-track control of uncertain nonlinear large ship could be realized. A robust feedback-controller was imported to ensure closed-loop stability in the initial learning stages of NN model and improve the NN control＇s real-time ability. Simulation results showed that the method had perfect control effect.

关 键 词：神经网络 船舶运动控制 模型参考控制 非线性控制 仿真 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] U664[自动化与计算机技术—控制科学与工程]