船舶动力定位系统多模型自适应融合滤波

Multiple model adaptive fusion filtering for dynamic positioning of marine vessels

为了减小波浪高频和环境噪声对船舶运动的影响,解决动力定位系统中的滤波问题,采用一种改进的多模型自适应融合(MMAF)估计算法。MMAF算法运用DGPS和平台罗经建立对地速度基准的研究方法,采用多模型描述系统结构,且每个模型对应不同滤波算法,利用不同滤波器所提供信息的融合,得到有效的船舶位置和首向估计信息。仿真实验表明,与其他方法相比,MMAF的应用使得滤波效果得到提升,性能得到改进。

Amultiplemodeladaptivefusion（MMAF）estimationalgorithmisadoptedtosolve problems of reducing the wave frequency and the influence of environmental noise on the ship movement. In the MMAF algorithm, the method of establishing speed reference with DGPS and platform gyrocompass is used and the system structure is described by multiple models and each model corresponds to different filter algorithm, the fusion of information provided by different filter, yielding effective position and heading information estimates of the marine vessel. Simulations illustrate the efficacy of the MMAF techniques proposed and the improvement in performance that is obtained when compared with other approaches.