基于自适应神经网络的船舶航向保持预定义性能PI控制  

PI control of ship course keeping with predefined performance based on adaptive neural network

作　　者：刘训文 褚善东[1] 骆海洋 钟平 LIU Xunwen;CHU Shandong;LUO Haiyang;ZHONG Ping(School of Navigation Engineering,Zhejiang International Maritime College,Zhoushan 316021,Zhejiang,China;Navigation Department,Zhoushan Marine School,Zhoushan 316021,Zhejiang,China)

机构地区：[1]浙江国际海运职业技术学院航海工程学院,浙江舟山316021 [2]舟山航海学校航海技术系,浙江舟山316021

出　　处：《上海海事大学学报》2024年第1期10-15,共6页Journal of Shanghai Maritime University

基　　金：舟山市科技局公益项目(2022C31039)。

摘　　要：为解决模型动态不确定和外部扰动未知的船舶航向保持问题,提出一种基于自适应神经网络的船舶航向保持预定义性能PI控制方案。在PID控制设计框架下,引入自适应神经网络和预设性能控制技术,从不确定补偿和设计角度提高船舶航向保持的精度和控制性能。在控制设计中,结合自适应神经网络技术与单参数学习技术,使得整个船舶航向保持闭环控制系统仅需要在线更新一个未知参数,系统的复杂度降低,且可以实现离线确定船舶航向误差的功能。基于李雅普诺夫稳定性理论进行分析,结果表明所提出的控制方案能保证整个闭环控制系统所有信号均有界。通过数值仿真验证了所提出方案的有效性和优越性。For the problem of the ship course keeping with uncertain model dynamic and unknown external disturbance,a scheme of PI control for ship course keeping with predefined performance based on the adaptive neural network is proposed.Under the design framework of PID control,introducing the techniques of the adaptive neural network and the predefined performance control,the ship course keeping accuracy and control performances are guaranteed from the perspective of the uncertain compensation and design.In the control design,combining with the adaptive neural network and single-parameter learning techniques,it is achieved that the closed-loop control system for ship course keeping only needs to update one unknown parameter,the complexity of the system is decreased,and the function of determining ship course errors offline can be realized.The analysis is conducted by the Lyapunov stability theory,and the results show that the proposed control scheme can guarantee the boundedness of all signals of the closed-loop control system.The effectiveness and superiority of the proposed scheme are verified by the numercial simulation.

关 键 词：船舶航向 自适应神经网络 PI控制 预定义性能 智能航行 

分 类 号：U671.99[交通运输工程—船舶及航道工程]