CAJ | 학술논문

以自主设计的单轴恒速偏频激光陀螺航姿系统为研究对象,在分析车载晃动基座上初始对准时残差序列特性的基础上,通过加窗估计测量噪声方差,并对测量信息进行去相关处理,进而提出了测量去相关改进的基于残差的测量噪声自适应滤波算法.实验室内静基座和车载晃动基座实验测试结果表明,所设计的自适应滤波算法与标准卡尔曼滤波算法相比,不需要预先精确得到测量噪声统计特性,并且车载晃动基座初始对准在200 s 时间内方位角误差小于40〞.
이자주설계적단축항속편빈격광타라항자계통위연구대상,재분석차재황동기좌상초시대준시잔차서렬특성적기출상,통과가창고계측량조성방차,병대측량신식진행거상관처리,진이제출료측량거상관개진적기우잔차적측량조성자괄응려파산법.실험실내정기좌화차재황동기좌실험측시결과표명,소설계적자괄응려파산법여표준잡이만려파산법상비,불수요예선정학득도측량조성통계특성,병차차재황동기좌초시대준재200 s 시간내방위각오차소우40〞.
The single-axis constant rate biased RLG attitude and heading reference system was studied. By analyzing the characteristics of residual sequences in initial alignment progress on rocking base for land vehicle, the measurement noise covariance was estimated inside a moving window, and the decorrelation treatments to measurement information were made. A measurement decorrelation modified residual based adaptive filtering algorithm was proposed, and then experiments on the postulate system, which was designed independently on stationary base in laboratory and on rocking base for land vehicle, were done to verify the effectiveness of the proposed algorithm. The results show that, compared with standard Kalman filtering, the adaptive filtering algorithm does not need to predetermine the statistics characteristics of measurement noise, and the azimuth error in initial alignment by the adaptive filtering algorithm on rocking base for land vehicle can be less than 40″ in 200 s.