Alpha-beta滤波在船舶动力定位状态估计中的应用

Application of alpha-beta filter in state estimation of ship dynamic position

船舶动力定位系统已广泛应用于海洋工程中,状态估计是动力定位系统的重要组成部分。在工程应用中,状态估计方法主要采用基于卡尔曼滤波的算法,但是这些算法对于船舶首向的滤波效果并不理想。alphabeta滤波是一种不基于模型的稳定常增益滤波器,其结构与卡尔曼滤波类似。本文设计了一种混合滤波器,采用alpha-beta滤波对船舶首向进行滤波,扩展卡尔曼滤波对船舶横向和纵向进行滤波,以改善船舶首向的滤波效果。通过将混合滤波器的滤波效果与扩展卡尔曼滤波器进行对比,验证了alpha-beta滤波用于船舶首向滤波的可行性和有效性。

Ship dynamic positioning system has been widely used in ocean engineering, the state estimation is an important part of dynamic positioning system. In engineering applications, the state estimation mainly uses the algorithms based on Kalman filter, but the heading filtering effect of these algorithms is not ideal.Alpha-betafilter, whose structure is similar to Kalman filter, is a kind of filter which is not based exact model and has fixed gain. We design a hybrid filter, which uses alpha-beta filter for ship＇s heading and employs extend Kalman filter for the estimation of horizontal and vertical, for the purpose of improving the ship＇s heading filtering effect. By comparing the filtering effect between the hybrid filter and extend Kalman filter, it can prove the feasibility and effectiveness of thealpha-beta filter for ship＇s heading filtering.