Study On The Method Of State Estimation In Dynamic Positioning System

The dynamic positioning system is a complex nonlinear system.The measurement error of the sensor,the inaccuracy of the system model and the interference of the external environment such as wind,wave and flow have great influence on the accuracy of positioning.Knowing a system's state is essential when deciding upcoming control actions.The state estimation is a critical part of the dynamic positioning system,because it can estimate the state of the system from series data which including noises.In this paper,the methods of the state estimation in the dynamic positioning system are studied.In order to improve the accuracy of dynamic positioning,the ship equipped with a dynamic positioning system is usually equipped with multiple same and different types of sensors to measure its status.An improved adaptive weighted fusion algorithm is proposed for the occasion of different types of sensors.The algorithm can accurately calculate the variance of each sensor and improve the accuracy of data fusion.In addition,an improved weighted dynamic allocation algorithm is proposed for the occasion of the same types of sensors.Using the iterative method to solve the problem that the weighted dynamic allocation algorithm needs to preserve the historical data,and improve the real-time performance of the algorithm.The simulation results show that the accuracy of adaptive weighted fusion algorithm is significantly higher than that of the traditional adaptive weighted fusion algorithm when the type of sensors is different,and the accuracy and real-time performance of the improved weight dynamic allocation algorithm are higher than those of the traditional adaptive weighted fusion algorithm and the improved adaptive weighted fusion algorithm when the type of sensors is the same.To solve the problem that the conventional unscented Kalman filter becomes inaccurate and diverges by time caused by the system uncertainty or complex ocean occasion.This paper proposed an adaptive noise estimation algorithm which can adaptively estimate the process noise.Numerical simulations show that the adaptive noise estimation algorithm can estimate the process noise accurately and improve the result of filter.Unscented Kalman filter has been widely used in dynamic positioning system since it has many advantages in estimation of nonlinear system.This paper aims to solve the problem that unscented Kalman filter is unable to track sudden changes of the states of a vessel when it faces the extreme sea environment,while the dynamic positioning system requires to estimate these states accurately and instantaneously.By identifying the instant of a sudden change and appropriately adjusting the estimated covariance matrix,we proposed an adaptive unscented Kalman filter which is able to track the states of vessel and reduce the deviation of the low-frequency position of the vessel efficiently.Numerical simulations show the effectiveness of the proposed scheme.