Research On Rudder Fin Joint Anti-Rolling System Under Multi Disturbance And Strong Coupling

The ship will roll and yaw violently when encountering wind and waves,which will affect the stability,maneuverability,and safety of the ship.The rudder fin joint anti-rolling can overcome the shortcomings of a single antirolling device and improve the anti-rolling effect while ensuring heading accuracy.Therefore,the rudder fin joint anti-rolling system under the interference of sea waves is studied to solve the existing problems of model mismatch,mechanism constraints,time-varying interference,state coupling,and parameter tuning.So as the anti-interference performance and output tracking performance of the system are improved.The nonlinear mathematical model of the ship is built based on the mathematical model method.Newtonian dynamics and fluid mechanics are used to analyze the forces acting on the hull,fin stabilizer,rudder,and propeller.According to the complex and changeable characteristics of sea waves,a disturbance mathematical model including significant wave height,encounter angle,and angular frequency is established.To improve the simplicity of the model,Taylor expansion is used to simplify the nonlinear model into a three-degree of freedom state space model including sway,roll,and yaw.The ship maneuverability simulation under different wave levels is also carried out.Based on the model,the problems of model mismatch,system timevarying disturbance,and frequent steering and fin manipulation are solved.A disturbance model predictive control based on sliding mode observer is proposed.The model mismatch and wave disturbance are considered timevarying disturbances at the input.The sliding mode observer is designed to observe the time-varying disturbance and system output in real-time,and the stability of the observer is proved.The disturbance model predictive controller is designed according to the incremental model and observation information.The ship motion control problem is transformed into a quadratic programming problem.To avoid frequent maneuvers,constraints such as rudder fin angle and angular velocity are added,and the control law constraint saturation after disturbance compensation is discussed.Finally,to avoid the mechanism wear caused by high-frequency noise,a filter is used to protect the rudder fin control law.In addition,the problems of state coupling,output delay,and controller parameter maladjustment are deeply studied.A rudder fin joint anti-rolling system based on active disturbance rejection control and reinforcement learning is proposed.Aiming at the state coupling problem,the model is decoupled with the idea of total disturbance.The roll disturbance rejection controller and the roll yaw predictive observer are designed respectively.The predictive output is used instead of the actual output feedback controller to solve the output delay problem.The depth deterministic strategy gradient in reinforcement learning is introduced to modify the parameters of the prediction observer in real time to ensure the accuracy of the prediction output.To meet the requirements of both controllers,an objective function consisting of output expected error and a scale factor is proposed.By adjusting the size of the corresponding output scale factor,the output tracking of various sea conditions can be handled.Then the control laws of rudder angle and fin angle can be reasonably allocated.