一种组合导航系统故障检测算法

在研究组合导航系统常用故障检测方法的基础上,将两种χ2故障检验法的检验量相结合,提出了一种针对组合导航系统的二叉树模糊支持向量机故障检测算法.该算法综合应用了Kalman滤波器、二叉树支持向量机等理论,采用模糊因子有效地减弱了噪声样本数据对支持向量机训练结果的影响,采用二叉树支持向量机的多个支持向量机实现对组合导航系统的多级故障进行多级分类和故障检测.最后,针对MINS/GPS组合导航系统,进行了相关仿真研究,仿真结果证明了该方法的有效性.

A Fault Detection Algorithm of Integrated Navigation System

Based on the study of failure detection methods which were used in integrated navigation system before, a kind of failure detection algorithm of integrated navigation system , which is based on Support Vector Machines with Binary Tree Architecture, is creatively proposed. Firstly, this algorithm combines many algorithms such as the Kalman filter algorithm, binary tree support vector machines and so forth. Secondly, the effectiveness of sample data with noise is greatly reduced by using fuzzy factor. Simultaneously, multi-level classification and fault detection for integrated navigation system by using multiple support vector machines of the binary tree support vector machines is carried out. Finally, a kind of computer simulation for MINS/GPS integrated navigation system is designed. And the simulation results show the effectiveness of this algorithm.