摘要
首先应用线性系统的奇异值分解理论对静基座初始对准的可观性及可观度进行了深入地分析。给出了不同奇异值下系统状态可观性状况的直方图,并且计算出不同状态组合时系统奇异值的大小及对可观性的影响。理论分析表明选取N、E、Eε为不可观测时,系统的估计效果最好。其次采用卡尔曼滤波器对可观阵的秩为7的三种状态组合进行最优估计,仿真结果验证这种论断的正确性。最后从误差分析的角度阐述了这种论断的合理性。
First, by use of singular values decomposition method, the observability and degree of observability of INS on stationary base is deeply analysed. The figure is gived,which descripts conditions of the system＇s observability based on different singular values. Also the magnitudes of singular values of different states combination are computed, which affects the system~ s observability. Theory analysis indicates that when choosing,as un-observable states, the system＇s estimations are most optimal. Second, Kalman filter is adopted to estimate those states, whose combinations leading the rank of observability matrix equal to 7. Simulation makes the results valid. Last from error analysis,the author also expatiate the rationality of the result.
出处
《火力与指挥控制》
CSCD
北大核心
2006年第5期22-25,共4页
Fire Control & Command Control
基金
国防预研基金资助项目(413090503)
关键词
初始对准
奇异值分解
可观性
可观度
initial alignment ,singular value decomposition ,observability ,degree of observability
