船舶参数横摇抑制的神经网络逆/内模控制  

Neural Network Inverse/Internal Model Control for Ship Parametric Roll Suppression

作　　者：徐涛 李晖 祁昱豪 XU Tao;LI Hui;QI Yuhao(College of Marine Electrical Engineering,Dalian Maritime University,Dalian 116026,Liaoning,China)

机构地区：[1]大连海事大学船舶电气工程学院,辽宁大连116026

出　　处：《船舶工程》2022年第6期109-116,共8页Ship Engineering

基　　金：国家自然科学基金项目(51879027)。

摘　　要：参数横摇是船舶纵浪或接近纵浪航行时,遭受到海浪等外部激励作用而引发的横摇运动共振现象,严重时会导致船舶倾覆。文章提出鳍桨联合抑制参数横摇的方案,利用减摇鳍产生抵抗力矩增大横摇阻尼,同时通过控制螺旋桨转速改变航速,进而改变船舶遭遇频率,规避参数横摇发生的条件。针对参数横摇的非线性和耦合特性,提出利用神经网络构建参数横摇运动的逆模型;并将其与原系统串联,构成伪线性复合系统,实现线性化和解耦。另外,根据解耦后横摇和船速子系统的特点,分别设计附加二自由度内模控制器,进一步提高系统的鲁棒性。采用实船参数,对所设计的方案进行仿真验证,结果表明,该方案能够实现参数横摇快速抑制。Parametric rolling is the resonance phenomenon of rolling motion caused by external excitation such as waves when the ship is sailing in or near longitudinal waves,which can cause the ship to capsize in serious cases.A scheme is proposed to suppress parametric rolling by using fin and propeller.The fin is used to generate the resistance moment to increase the roll damping.At the same time,the propeller speed is controlled to change the ship’s speed,and then the encounter frequency of the ship changes to avoid the conditions of parametric roll.In view of the nonlinear and coupling characteristics of parametric rolling,an inverse model of parametric roll motion based on neural network is proposed,which is connected in series with the original system to form a pseudo linear composite system to realize linearization and decoupling.In addition,according to the characteristics of the decoupled rolling and ship speed subsystems,additional two degree of freedom internal model controllers are designed to furtherly improve the robustness of the system.Using the real ship parameters,the design scheme is simulated and verified,and the results show that it can realize the rapid roll reduction of parametric rolling.

关 键 词：参数横摇 鳍桨联合 神经网络逆控制 内模控制 

分 类 号：U664.7[交通运输工程—船舶及航道工程]