基于IMM算法的组合导航系统故障诊断研究-通信与信息系统专业论文.docxVIP

摘 要 惯性导航、卫星导航与天文导航这三种具有代表性的导航方式各有优缺点，充 分发挥它们各自的优势，互相取长补短，组成组合导航系统是实现精确制导的重要 发展方向，也是组合导航研究的典型对象。为进一步提高组合导航系统的可靠性和 精度，本文对组合导航系统的故障检测与诊断算法进行了研究。 针对上述问题，本文详细介绍了组合导航系统总体方案的设计，包括组合导航 系统的四种工作模式、状态估计方法和组合导航系统的容错解决方案。在组合导航 故障检测算法方面，主要讨论了目前较为成熟的几种滤波算法，包括集中式卡尔曼 滤波、联邦卡尔曼滤波、自适应卡尔曼滤波和交互式多模型算法，其中交互式多模 型算法的精度稍高于其它几种卡尔曼滤波算法，对于飞行器等重要运载体来说，导 航系统的可靠性比精度更为重要。 为进一步验证交互式多模型算法在组合导航系统故障检测与诊断中的优越性， 本文以 GPS/INS 组合导航系统为例，建立了组合导航系统的数学模型和故障模型， 在 MATLAB 仿真平台下将交互式多模型算法与传统卡尔曼滤波算法进行了对比，结 果表明在组合导航系统的传感器或执行器发生故障时，交互式多模型算法的精度比 传统卡尔曼滤波算法更高。 关键词： 组合导航，卫星/惯性导航，卡尔曼滤波，交互式多模型，故障检测 Abstract Inertial navigation, satellite navigation, and celestial navigation for the navigation of these three methods have advantages and disadvantages of representative, give full play their respective advantages to complement each other to form a precise integrated navigation system is an important development direction of guidance, navigation research is Typical objects. Integrated navigation systems to further improve the reliability and accuracy, this paper integrated navigation systems fault detection and diagnosis algorithm is studied. In response to these problems, this paper introduces the overall scheme of the integrated navigation system design, including the four integrated navigation system operating mode, state estimation and fault-tolerant integrated navigation system solutions. Fault detection algorithm in Integrated Navigation, the main discussions of the current mature several filtering algorithms, including the centralized Kalman filter, the Federal Kalman filter, adaptive Kalman filter, and interactive multiple model algorithm, in which the interactive multiple model algorithm Slightly higher accuracy of several other Kalman filter, for aircraft, missiles and other important transporters, the navigation systems reliability is more important than accuracy. To further verify the interactive multiple model algorithm in integrated navigation system fault detection and diagnosis of the superiority of this paper, GP