长基线导航系统滤波算法的研究与实现

发布时间：2018-01-16 08:20

本文关键词：长基线导航系统滤波算法的研究与实现　出处：《沈阳理工大学》2015年硕士论文　论文类型：学位论文

【摘要】：作为海洋探索与开发的重要工具,载人潜水器无论在军用领域、民用领域还是科研领域都具有非常广阔的应用前景,导航定位技术是其发展的关键。随着世界各国对海洋探索与开发的日渐深入,人们对水下导航定位系统提出了更高的要求,但由于硬件和成本的限制,使得通过高精度的导航传感器来获得高精度的导航定位系统变得愈加困难。此外,复杂未知的海洋环境,要求导航定位系统具有高度的自主性、可靠性和强抗干扰能力,由单一传感器构成的导航系统已难以满足要求。因此研究适应性强精度高的非线性滤波算法成为获得可靠的高精度的导航定位系统的主要途径之一。无迹卡尔曼滤波算法是目前广泛使用的非线性滤波算法,而平方根无迹卡尔曼滤波算法相对于前者具有显著的优势,因此本文以实际科研项目为研究背景,主要围绕平方根无迹卡尔曼滤波算法及其在LBL/DR组合导航系统中的应用开展相关研究工作。首先,简要介绍了LBL/DR组合导航系统的组成及工作原理。其次,对无迹卡尔曼滤波算法和平方根无迹卡尔曼滤波算法的应用前提和算法的实现过程进行了详细地阐述,并从理论上证明了平方根无迹卡尔曼滤波算法不仅能够解决无迹卡尔曼滤波算法应用过程中存在的滤波器的计算发散问题,而且还可以提高计算效率。再次,在平方根无迹卡尔曼滤波算法的基础上,围绕其测量更新方法的不足和不具有应对噪声统计变化的自适应能力问题开展研究工作。针对平方根无迹卡尔曼滤波算法测量更新方法的不足,前人已经做了大量的理论研究工作,本文仅对其中一种易于工程实现的迭代平方根无迹卡尔曼滤波算法进行介绍,该方法可使滤波估计输出具有更高的精度和更小的方差。针对平方根无迹卡尔曼滤波算法不具有应对噪声统计变化的自适应能力,其在噪声统计未知时变情况下易出现滤波精度下降甚至发散的问题,本文提出了一种带时变噪声统计估计器的自适应平方根无迹卡尔曼滤波器。在滤波过程中,自适应滤波一方面利用量测值修正预测值,另一方面也对未知的或不确切的噪声统计参数进行估计修正。最后,利用载人潜水器以往的海试数据对迭代平方根无迹卡尔曼滤波算法和自适应平方根无迹卡尔曼滤波算法进行了验证。
[Abstract]:As an important tool of ocean exploration and development, manned submersible has a very broad application prospect in military, civil and scientific research fields. Navigation and positioning technology is the key to its development. With the deepening of ocean exploration and development in the world, people put forward higher requirements for underwater navigation and positioning system, but due to hardware and cost constraints. It becomes more and more difficult to obtain high precision navigation and positioning system by high precision navigation sensor. In addition, the complex unknown marine environment requires the navigation and positioning system to have a high degree of autonomy. Reliability and strong anti-jamming ability. The navigation system composed of a single sensor is difficult to meet the requirements. Therefore, the study of nonlinear filtering algorithm with high adaptability and high precision has become one of the main ways to obtain reliable navigation and positioning system with high accuracy. Filtering algorithm is a widely used nonlinear filtering algorithm. The square root unscented Kalman filter algorithm has a significant advantage over the former, so this paper takes the actual research project as the research background. This paper mainly focuses on square root unscented Kalman filter algorithm and its application in LBL/DR integrated navigation system. First of all. The composition and working principle of LBL/DR integrated navigation system are introduced briefly. Secondly. The application premise and realization process of unscented Kalman filter algorithm and square root unscented Kalman filter algorithm are described in detail. It is proved theoretically that square root unscented Kalman filter algorithm can not only solve the problem of filter divergence in the application of unscented Kalman filter algorithm, but also improve the computational efficiency. Based on the square root unscented Kalman filtering algorithm. The research work is focused on the deficiency of the measurement updating method and the adaptive ability to deal with the statistical changes of noise, and the deficiency of the square root unscented Kalman filter algorithm. Previous researchers have done a lot of theoretical research. This paper only introduces one of the iterative square root unscented Kalman filtering algorithms which is easy to be implemented in engineering. This method can make the output of filter estimation have higher accuracy and smaller variance. The unscented Kalman filter algorithm for square root has no adaptive ability to deal with the noise statistical changes. In this paper, an adaptive square root unscented Kalman filter with time-varying noise estimator is proposed. On the one hand, the adaptive filter uses the measured value to correct the prediction value, on the other hand, it also estimates the unknown or inaccurate noise statistical parameters. Finally. The iterative square root unscented Kalman filter algorithm and the adaptive square root unscented Kalman filter algorithm are verified by the previous sea test data of the manned submersible.
【学位授予单位】：沈阳理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P715;TN713

【共引文献】