Research On Wave Filtering Of Vessel Dynamic Positioning System

As fewer and fewer resources such as oil are available for extraction on land,mankind has turned its attention to the sea.Energy extraction at sea requires fixing the ship on the surface.For shallow sea areas,anchoring systems and self-lifting systems can generally be used,but the cost and difficulty of these two methods increase as the depth of seawater increases,thus giving rise to power positioning systems that rely solely on propulsion systems to maintain their position and attitude.The total motion of a ship equipped with a power positioning system at sea will be disturbed by environmental forces,and its total motion will be in the form of superimposed high and low frequency motions.The low frequency motions can be compensated by designing a controller to control the thrusters,while the high frequency motions,if also introduced into the controller,will accelerate the wear and tear of the thrusters and lead to the waste of fuel.To address this problem,this paper proposes an exogenous Kalman filter-based wave filter,and then proposes a wave filter switching system based on the wave filter designed in this paper to address the problem that the effect of the wave filter deteriorates when the sea state changes,and finally verifies the effect through simulation.The specific research contents of this paper are.First,a suitable coordinate system is established and based on the established coordinate system,a ship model,an environmental force model and a measurement model required for this paper.The reasonableness of the established model was verified by simulation.Secondly,the wave filter based on exogenous Kalman filter is proposed in this paper.The common wave filtering methods are extended Kalman filter and passive nonlinear observer,both of which have their own advantages and disadvantages.The wave filter based on the exogenous Kalman filter is divided into two parts,and the estimation of the first part of the passive nonlinear observer is used as the exogenous input of the second part of the linearized Kalman filter,which converges faster than the extended Kalman filter in theory because the linearization of the second part is about a global exponentially converging estimation.The convergence speed and effectiveness of the designed wave filter are verified by simulation.Finally,the wave filter switching system is proposed to address the problem that the effect of the wave filter designed with fixed model parameters decreases after the sea state changes.The wave filter switching system is a soft switching system,which consists of three exogenous Kalman filter-based sub-sea wave filters,a fuzzy supervisor and a wave frequency tracker.Before designing the sub-surface wave filter,we need to reclassify the common sea state into three different classes,select a specific meaningful wave height as a typical value for the reclassified sea state,and use the approximate fitting method to obtain the model parameters corresponding to the selected meaningful wave height and design the sub-surface wave filter based on them.The framework of the wave filter switching system is designed based on the soft switching system.Since the sea state and the meaningful wave height cannot be measured directly,a wave frequency tracker is used to estimate the wave frequency instead of the sea state and the meaningful wave height.The required supervisor is designed using T-S fuzzy system.The main function of the supervisor is to weight the estimates of the sub-sea wave filters according to the estimated wave frequencies,and the estimates after weighted averaging are the final estimates.The performance of the designed wave filter switching system is verified by simulation.