强跟踪扩展卡尔曼滤波及其在捷联惯导初始对准中的应用

Strong tracking EKF and its application in SINS alignment

针对在模型失配和观测噪声干扰情况下常规滤波器出现精度下降甚至是发散的问题,提出了基于自适应多重渐消因子的改进方法,即通过计算新息序列协方差的估计值确定多重渐消因子,提高当前观测量对滤波器的修正作用。仿真及实测实验结果表明:该算法可以有效抑制滤波发散,同时保证其快速收敛以及高精度估计。

To solve the problem of filtering divergence in case of strong nonlinearity,large initial error and model distortion etc,an improved algorithm based on adaptive multiple fading factors is proposed.Fading factors are determined by the estimation of the innovation sequence covariance,and the correction of the current measurements is enhanced.Simulation and experiment demonstrate that the proposed algorithm performs better in terms of rapidity,convergence and accuracy.