Research On Design And Control Of Fin Stabilizer System With Dynamic Lift Feedback

When the ship is sailing on the sea,due to the disturbance of the marine environment,six degrees of freedom random motion is generated,especially rolling.Severe rolling seriously affects the seaworthiness of ships,the safety of cargo and the comfort of passengers,and reduces the launching accuracy of ship borne weapons.Thus,researchers have developed a variety of roll reduction devices.Fin stabilizer is the most successful and active anti-rolling device which has been widely used so far.The theoretical anti-rolling effect can reach more than 90%.However,due to the constraints of its own and external factors,it is difficult to achieve the desired goal in practical engineering applications.Firstly,the characteristics of fin stabilizer system are studied,and the key factors are analyzed from the principle of roll stabilization.The torque required for roll stabilization depends mainly on the dynamic lift,which can be divided into lift generation,lift detection and lift control.Because of the improvement of conventional fin stabilizer system,the lift detection and lift control are mainly studied in this paper.By analyzing the external interference model and the simplified application of the theoretical roll model in practical engineering,the main problems of the system are pointed out: the system feedback isn't accurate and the control strategy isn't suitable.Aiming at the feedback problem,the reason of excessive lift deviation caused by conventional fin angle feedback is analyzed.Through hydrodynamic simulation,the mechanical angle of fin is substituted for the effective angle of attack in engineering,and the influence of theoretical angle of attack,velocity angle and additional angle is ignored.For the control problem,it is analyzed that the conventional PID control can not meet the real-time control requirements of lift,and other control laws ignore the premise that objective constraints exist in the actual system.Therefore,it is necessary to use dynamic lift detection to improve the feedback mode of the system.On this basis,the appropriate control law is designed to improve the overall performance of fin stabilizer.Due to the complicated marine environment and the installation,maintenance and overhaul of sensors in practical applications,it is difficult to detect the dynamic lift of fins directly.Therefore,lift detection technology is the difficult point and innovation point to be solved.Based on the analysis of the structure and stress characteristics of the fin shaft,the two point simply supported composite beam model with an external cantilever is established.On the premise of inheriting the reliability of the conventional fin stabilizer,the lift detection method of the upper bearing load is given.The special bearing mounting seat and sensor are designed,and lift is directly decomposed from dynamic hydrodynamic force to avoid tedious operation and deviation.Subsequently,the hollow fin shaft structure of American Sperry is improved,and new sensor installation methods and measurement points are proposed.In order to reduce the difficulty of processing and assembling the fin shaft,a laser ranging lift detection method with the shaft-core removed is designed.The deflection and rotation angle of fin shaft deformation are analyzed by using Timoshenko beam theory,and then the quantitative relationship between lift on the fin and displacement of the fin shaft is established.In order to reduce the difficulty of lift detection,the factors that affect the deformation of the fin shaft are analyzed,and the feasibility of the design is verified by numerical calculation and finite element simulation with some actual fin stabilizer parameters as an example.For single sensor detection,the signal is prone to interference,so it needs to be filtered.Because the sensors in the lift detection don't influence each other,the fusion and filtering technology can be used to process the multi-sensor signals so as to complement each other and avoid the shortcomings of the single sensor.In order to prevent the fusion and filtering process from delay,the sequential structure fusion and filtering algorithm is designed to accelerate the speed.Firstly,the lift signal of the same nature is fused and filtered,but the state equation and performance index of the sequential structure are inconsistent,which will lead to the problem that the Kalman filter is difficult to be directly applied to the lift signal fusion and filtering problem.Therefore,the equivalence between sequential structure and centralized structure is proved,so as to ensure the flexibility and real-time of filtering process.Through the calculation and comparison of different signals,the fault or abnormal signals are removed.In order to improve the reliability of fusion,unbiased estimation method is usually used.Since the reliability of the fusion is inversely proportional to the variance,the reliability can be improved if the reliability of the fusion is reduced.For this reason,the reliability of the ridge estimate is determined by the variance and deviation,and the variance is adjusted by adjusting the partial parameter.Compared with unbiased estimation,the fusion reliability can be further improved when the same number of sensors is used.On the basis of lift detection and multi-sensor signal fusion,the improvement principle of fin stabilizer control system is given.And the application of dynamic lift feedback and fin angle feedback is analyzed.In view of practical engineering applications,no matter what kind of feedback is insufficient,it is difficult to meet the requirements of various situations and system reliability.Therefore,the structure of fin stabilizer control system is improved,which is designed with dynamic lift feedback.This paper analyzes the switching conditions suitable for different feedback,and forms redundant feedback to improve the reliability of fin stabilizer operation.In practical engineering,in order to prevent the dynamic stall of fin stabilizer and ensure the safety of the actuator in high sea state,it is necessary to limit the fin angle and lift.It is difficult for the conventional PID controller to control dynamic lift effectively due to the change of the external disturbance,the nonlinear model and the limit of the actuator.Therefore,the modular controller is designed.For the control range of lift input is limited,the control module based on RBF neural network is designed for dynamic compensation.In view of the nonlinear factors and the influence of external disturbances,a nonlinear disturbance observer module is designed to weaken the generalized disturbance.And the fuzzy sliding mode control module is designed to improve the efficiency of the conventional control method.Through the simulation comparison of different feedback methods,the feasibility of improving control system structure is verified,and the simulation of different encounter angle,speed and sea state is carried out to verify the effectiveness of the control strategy design.