英文演讲
2020-03-03 21:36:15 来源：范文大全收藏下载本文
The integrated navigation system using Kalman filtering data proceing general technical.In the late 1960\'s Kalman filtering technology in the aerospace fields began after the application of real-time dynamic navigation algorithms, rapid development, the research Kalman filtering appears the nonlinear, square root Kalman filtering of integrated filtering algorithm, but then the Kalman filtering precision confined to solve and calculated stability, from the late 1980s so far, due to the development of computer technology level by adaptive Kalman filter, and centralized filter and federal filtering for represent, comprehensive consideration of the accuracy and reliability of the new generation of prudence and algorithm to navigation algorithms on the development and application of Kalman filtering theory, discretization rapid development.Despite the high reliability combined navigation algorithms already have made great progre, but these most studies for the linear system.And many algorithm still exist in performance in theory or blurred, concentrate filtering although in theory can give navigation parameters of the global optimal solution, but there are two fatal limitations, i.e.calculation burden and fault-tolerance poor.And based on the information distribution principle design federal filtering method algorithm, because its design flexibility, small amount of calculation and fault tolerance has good performance and attention.But federal filtering arithmetic ignores every single satellite navigation system filtering the correlation between the output, its fault tolerance is very difficult to guarantee.As the development of artificial intelligence, appeared again using neural network and fuzzy theory nonlinear technology proceing integrated navigation data method, but these method is still in theoretical discuion stage, and the ideal data is used mostly simulation of artificial intelligence technology in the usability of integrated navigation system analysis demonstrated, if applied to the measured integrated navigation system also needs to solve a lot of problems.
组合导航系统的数据处理一般采用Kalman滤波技术。60年代后期Kalman滤波技术在航空航天领域开始应用后，实时动态导航算法的研究得以快速发展，随之出现了非线性Kalman滤波、平方根滤波等集成化的Kalman滤波算法，但当时的研究多局限于解决Kalman滤波精度及计算稳定性方面，从八十年代后期至今，由于计算机技术水平的发展，以自适应Kalman滤波、集中滤波和联邦滤波等为代表，综合考虑精度、可靠性及算法稳健性的新一代导航算法得以发展和应用，离散化Kalman滤波理论得以快速发展。尽管高可靠性组合导航算法研究已取得较大进展，但是，这些研究大多针对线性系统。而且许多算法还存在理论上或性能上的不严密性，如集中滤波虽然在理论上可以给出导航参数的全局最优解，但却有两个致命的局限，即计算负担重和容错性差。而基于信息分配原理设计的联邦滤波法算法，由于其设计灵活、计算量小和容错性能好而备受重视。但联邦滤波算法忽略了各单一卫星导航系统滤波输出量之间的相关性，其容错性很难保证。随着人工智能技术的发展，又出现了采用神经网络、模糊理论等非线性技术处理组合导航数据的方法，但这些方法至今还处于理论探讨阶段，且大多采用模拟的理想数据对人工智能技术在组合导航系统中的可用性进行分析论证,如果应用到实测的组合导航系统还需要解决很多问题。
人人范文网 m.inrrp.com.cn 手机版