动基座初始对准中H∞滤波器因子分析研究

H∞ Filter Factors Analysis of Initial Alignment on Swaying Base

惯性导航系统初始对准技术长期以来被作为研究的重点,特别是动基座对准研究。在这种情况下,方位误差角为大失准角,传统的Kalman滤波方法不能得到良好的对准效果;此时,采用鲁棒扩展H∞滤波器来解决这个问题。鲁棒扩展H∞滤波器是建立在鲁棒控制理论基础的次优估计算法,降低了模型的准确性,提高了模型的鲁棒性。同时,只要适当选取滤波器性能因子γ和δ就可以得到很好的对准效果。本文就γ和δ的选取做了理论上的分析和仿真试验,给出了相应的取值范围及准则,为工程中选取γ和δ提供了一定的依据。

The robust extend H∞ filter is a sub-optimal estimation algorithm based on the robust control theory, which reduces the accuracy of the model to improve the robustness. It works better than traditional Kalman filter on swaying base of initial alignment of SINS as the azimuth error is big. The appropriate factors γand δ of the filter are selected to make the filter work better. The range and criteria for selection of γand δ is obtained through theoretical analysis and simulation tests for project.