项目名称: 面向深海船舶动力定位系统的高精度实时状态估计方法研究
项目编号: No.61301279
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 无线电电子学、电信技术
项目作者: 冯辉
作者单位: 武汉理工大学
项目金额: 25万元
中文摘要: 船舶动力定位系统广泛的应用于深海钻井船、海底铺缆船等许多重要的深海作业船舶,已经成为国家海洋战略不可缺少的支持系统。信息处理对于船舶动力定位系统的精确定位起到至关重要的作用。状态估计滤波器作为信息处理的核心部分,在整个动力定位系统中扮演着不可或缺的角色。本课题利用数字信号处理、优化理论、概率论、船舶运动学等理论与方法,研究深海船舶动力定位系统状态估计中存在的主要挑战问题。主要研究工作包括:基于序贯期望最大算法研究高频线性运动模型中的噪声参数估计方法;利用平移不变小波变换,研究低频非线性运动模型中的实时噪声参数估计算法;在此基础上,基于边缘化粒子滤波研究和实现面向深海船舶动力定位系统的高精度实时状态估计算法。本课题的研究能为船舶动力定位系统的开发提供理论基础,也能丰富船舶动力定位系统相关的理论与技术,还能促进信息处理技术与船舶控制技术的交叉研究。
中文关键词: 船舶动力定位系统;状态估计;参数估计;数据融合;推力优化分配
英文摘要: Vessel dynamic positioning (DP) system is widely used in many important offshore operation ships, such as deep ocean drilling ship, submarine pave cable ship, and so on. Therefore, it becomes an indispensable support system in national marine strategy. Information processing is extremely important to guarantee high precision dynamic positioning. As a core component of information processing module, state estimation filter plays an essential role in DP. In this project, we intend to solve the main challenges existed in deep ocean DP by applying the theories and methods of digital signal processing, optimization theory, probability theory and vessel kinematics. Considering the difference between high frequency and low freqency motion in the mathematic model of vessel dynamic positioning system, we firstly propose a parameter estimation method for linear high frequency motion model based on the sequential expectation-maximization algorithm. Then, by utilizing the shift-invariant wavelet transform, we propose a real-time parameter estimation approach for nonlinear low frequency motion model. Furthermore, we study and implement a deep ocean vessel dynamic positioning system-oriented high-precision real-time state estimation algorithm based upon the marginalized particle filter. The research of this project will provi
英文关键词: vessel dynamic positioning system;state estimation;parameter estimation;data fusion;optimized thrust allcation