一种基于UKF改进算法的组合导航系统故障的检测方法

Method for fault detection of integrated navigation system based on improved UKF algorithm

为了缩短组合导航故障检测的时间,提高检测准确性,在残差χ2检验法理论基础上,对无迹卡尔曼滤波(UKF)进行研究。采用一种基于混合滤波思想的UKF,该算法在时间更新步骤分为两个阶段,利用线性方程得到状态的一步预测值并由此构造Sigma点,通过求取Sigma点加权和来实现对量测值的一步预测,降低了计算量。采用改进算法和标准UKF算法对含有故障的GPS/INS组合导航模型进行故障检测。仿真实验表明,改进的算法能够更迅速地检测出故障,有利于提高导航系统的稳定性,在工程应用中有重要意义。

In order to shorten the time of fault detection of integrated navigation system and improve it＇s accuracy, this paper researched unscented Kalman filter（UKF） based on the theory of residual chi-square test, and presented improved UKF based on the hybrid filter. It divided the time update into two stages, to use the linear equation of state to get the one step prediction and generated the Sigma points according to the value, the one step prediction of measured value could be calculated by summing the weights of Sigma points, so it reduced the calculation. It did the fault detection using the improved algorithm and standard UKF in the GPS/INS integrated navigation models which contained faults. Simulation results show that the improved algorithm can detect the fault more quickly and improve the stability of the navigation system. The improved UKF plays an important role in the engineering applications.