摘要
针对欠驱动水面船舶在直线航迹跟踪中受到强烈的风、浪、流时变干扰影响时出现大幅度的摇荡运动的情况,采用自适应卡尔曼状态估计和鲁棒预测控制,提出一种具有横摇角约束的控制器综合设计方法。建立时变干扰作用下的仿射切换系统模型及测量模型;基于扩展状态的自适应卡尔曼滤波方法,对船舶模型的状态和随机干扰力矩进行估计,并对估计的干扰力矩进行前馈补偿。考虑船舶的真实状态与卡尔曼观测状态之间存在观测误差,将控制器与观测器综合考虑,提出一种基于状态观测器和H_2/H_∞混合性能指标的直接约束鲁棒预测控制,在性能指标与控制器设计中直接利用观测状态;将状态约束、鲁棒稳定性约束、性能指标转化为LMI(线性矩阵不等式)的凸优化问题。理论证明了所设计闭环系统的一致有界稳定性,并且通过仿真验证了控制器能实现直线航迹鲁棒跟踪,在保证横摇角在约束范围内,对干扰有着有效的抑制作用。
For conditions in which time-varying disturbances of wind,wave,and current can result in large sway motion,this study proposes a novel robust predictive control method combined with augmented-state adaptive Kalman filter.First,multiple affine switching system and measurement model are established to estimate the states and stochastic disturbance torque using augmented-state adaptive Kalman filter,and feedforward compensation is adopted for the estimated disturbance torque.Then,as an observation error exists between real and observed states,a robust predictive control method with direct constraints is proposed on the basis of state observer and mixed H 2/H∞approach.The observed states are adopted directly in the performance index and controller design.The state constraints,robust stability constraints,and performance index are transformed into linear matrix inequality convex optimization.Finally,the uniform boundedness of a closed-loop system is proven theoretically.Furthermore,with the constrained roll angle,the effectiveness of the designed controller for straight path following and disturbance resistance is demonstrated by simulation.
作者
刘志林
李国胜
张军
LIU Zhilin;LI Guosheng;ZHANG Jun(College of Automation,Harbin Engineering University,Harbin 150001,China;Electrical and Information Engineering College,Jiangsu University,Zhenjiang 212013,China)
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2019年第2期312-317,共6页
Journal of Harbin Engineering University
基金
国家自然科学基金项目(51379044)
中央高校基本科研业务费(HEUCF0417)